CONTENTS CHAPTER II. MYTH, RITUAL AND COSMIC ORDER 98 I. Myth and nature-myth method—Nathan Söderblom on mana (Brahman) and Supreme Being (Brahma) — K. Th. Preusz on the connection of Supreme Being and Cosmic Order, influenced by rite — Connection of rite with mana (Brahman), cosmic order and Supreme Being (Brahma) — Myth, dramatization of rite as a source of knowledge about this connection — The MBh. as a myth. II. Émile Dürkheim on cosmic order, classificatory system and social organization — Hermann Oldenberg on the classificatory system of the Brahmanas — Naman and rüpa — Epic philosophy and ancient Indian anthropology. III. Cosmic evolution as a rotatory motion — Reincarnation and the theory of Gandharva and Pitri in this line of thought — Cosmic rotatory motion and the ritual of fire-rubbing and churning. CHAPTER III. KRISHNA AND ARJUNA 148 I. Knowing (jnana), cosmic order and ritual act (karman) The „Pantheistic" knowing of the Upanishads — Epic Hinduism as a new form of „Theism", later on harmonized with the Pantheism of the Upanishads? II. Krishna, a complicated being of the Indian Pantheon Classificatory system and the Pantheon — Dualism: Devas and Asuras — Dualism and Krishna — Krishna as the Divine Trickster. III. Krishna as a child — The conflict with Kamsa — Krishna as the Divine Initiate — Is equal to Arjuna. IV. „Variations saisonnières" and rainy season — Rainy season and initation. CHAPTER IV. RUDRA AND THE SABHA «»»»•»• 202 I. The tribal sanctuary and the initiation — E. B. Havell on „symbolism" in Indian architecture. II. Rudra, the central pole of the tribal sanctuary — Rudra and the rainy season — Rudra as the god of initiation Dualism: Rudra-Siva and Vishnu-Krishna — Trimürti and division into three — Rudra and the cult of the dead The chaitya — Rudra's animals. III. Rudra and the sabha — Rudra and the Vratyas — Vratyas and sabha — Club and clan organization — Club and initiation. INTRODUCTION. At the end of the last and the beginning of the present century the Mahabharata was subjected by a number of Sanskrit scholars to a more synthetic examination; but during the last few decades the interest seems more or less to have flagged, at least as far as the wider study of the Epic is concerned. The reasons are by no means self-evident. One of the principal reasons is undoubtedly to be found in the peculiar character of the Epic itself. It is not primarily the magnitude of the work in itself that has been an obstacle to a more synthetic study. Jacobi's concordance and table of contents together with Sörensen's index have fairly well helped us over this difficulty. It is essentially the peculiar composition of the work, consisting of almost 100,000 slokas, that makes it so uncommonly difficult. At first sight it seems to be a bizarre accumulation of all sorts of heterogeneous elements, witnessing only to an almost morbid mania for collecting, at least if it did not sometimes strike us as being so artless and simple. The first impression, however, soon makes place for a second, one of surprise, once one realizes that this vast pile may lay claim to being more or less balanced. For one can hardly say of the Epic that its structure is entirely arbitrary. And one wonders whence the builders of this vast edifice derived their sense of balance to enable them to produce a poem such as that which now lies before us. The Epic is often enough extremely prolix in its teaching in matters of religion and philosophy; of law andethics;ofcosmology and mysticism; of statecraft and the art of war; of logies and the art of healing. There move across its pages beautiful women, devoted spouses, ferocious warriors, austere ascetics and crafty impostors. The dramatis personae walk the stage beneath an everchanging light; the supreme god is likewise a deceiver, the king of justice is given to thoroughly illicit prac- 1 THE MAHABHARATA tices, the Kauravas, wily sharpers at dice, are, on the other hand, equally fair and chivalrous in battle. And yet the figures have not become mere puppets of a playful inventiveness, but are constantly depicted with characteristics that closely connect them with historical reality. How shall we consider this marvellous work as a whole? It would seem to be a most perilous proceeding to consider it as a whole at all, if we bear in mind the anything but enthusiastic reception that other theories of a more synthetic nature have met with. One of the most authoritative students of the Epic, E. Washburn Hopkins, whose ideas seem to be winning almost universal approval, declares in favour of the analytic method, since the synthetic method can get no grip on the Mahabharata at all1). But is seems that this analytic method, which has gradually come to be acknowledged as sound and tested, has in no case stimulated others to study the Epic analytically. One is immediately confronted with the difficulty of deciding where to begin one's analysis and then what to analyse. Here again the multiform character of the Mahabharata proves to be just as serious an obstacle as when the synthetic method is employed. One can understand as matters stand at present that but few feel called to follow Prof. Hopkins upon his weary way through the Epic, not to mention their beating a path for themselves. There were undoubtedly other reasons too. Newer things diverted the attention to other fields of research. There were other questions at issue to be answered, which were considered for the time being of greater importance, such as, for instance, Oldenberg's akhyana theory. Perhaps, too, there were others who thought it advisable to wait until one should be better justified in hazarding an opinion, i.e. until Sanskrit scholars had prepared a new and reliable recension of the original text of the Epic. We shall endeavour, to the best of our ability, to take up the study of the Epic again where it has stuck fast. We trust that, gradually, in the course of this study, the reasons will become evident which have led us to believe, that we have something to say upon this subject which will prove worthy of further consideration. We shall endeavour in a brief survey *) The Great Epic of India. Preface. INTRODUCTION to bring to the fore those points in the history of the study of the Mahabharata which are essential to our purpose. It is quite understandable that the earliest attempts at studying the Epic should have coincided with an endeavour to resolve it into its component parts, to analyse it, with a view to acquinng a general knowledge of its contents. The first thing to look for is, of course a certain nucleus, around which various other elements were later on to be grouped. Indeed the poem itself makes mention of the fact, that Vyasa made a Bharatasamhita, omitting the episodes (upakhyanair vina), and consisting of 24,000 slokas, together with a large Bharata (Mahabharata), consisting of six million slokas. Of the six million three million are recited among the Gods, one and a half million among the manes, one million four hundred thousand among the Gandharvas and only the remaining one hundred thousand slokas among the inhabitants of the earth. (1.1.103). Alongside the work in its complete form there is also talk of a condensed edition, consisting of not more than eight thousand eight hundred slokas. This condensed edition is considered extremely difficult (1.1.81). The reading of the Epic can be started in three different places (1.1.51). Finally, it is said that the poem was recited on three different occasions; the first time by Vyasa in the presence of his pupils; the second time by the Brahman Vaisampayana, Vyasa's pupil, during a solemn sacrifice offered in the name of King Janamejaya; the third time by the Süta Ugrasravas during a sacrifice celebrated by the rishi üsaunaka, a generation later. The history of the critical study of the Mahabharata maybe said to have begun with C. Lassen, who commenced his labours in 1837. The results of his researches may be summarized as follows: the Epic as recited by Saunaka is to be considered a second recension of the poem. There is talk of this second recension in Asv. G.S. 111.4. where a Bharata is made mention of alongside a Mahabharata. Asvalayana is very hypotherically dated 350 B.C. and considering he was a pupil of unaka, whom Lassen just as hypothetically identifies with >) For the history of the study of the MBh. we have consulted A. Opnvr^TV- Mahabharata, IV (1895), p. 165—204; A. BarthUeuvres, IV p. 347 ssq. in which works further literature dealing with the subject is mentioned. the Saunaka of the sacrifice, this second recension must belong to 400 or 460 B.C. From then onwards only interpolations of a Krishnaite nature found their way into the original poem. Omitting such interpolations the Epic in its original shape may be looked upon as being pre-buddhist. „On voit que ses conclusions étaient 1'arbitraire même" 1). Another attempt at reconstructing the poem in its original shape was made byS. Sörensen in 1883. In its oldest form the Epic was, according to him, a saga, of which the unity of the main story presupposes its being the creation of a single mind. In the original poem, therefore, there must be no contradiction, repetition, digression or the like. Going to work in this manner, Sörensen ultimately found himself left with a residue of some seven or eight thousand slokas, which he declared to be genuine. „La méthode de 1'auteur, malgré toutes les pré- cautions possibles, est arbitraire et le problème tel qu'il le pose, est en réalité insoluble" 2). Meanwhile A.Weber,since 1852, hadadopted another course. Turning to the Vedas, he imagined there was some possibility of finding some points in common with the Epic in the gatha narasamsyas (hymns proclaiming the praises of men). Alongside these are the ancient danastuti hymns in the Rig Veda, extolling more especially the generosity and liberality of the ancient rulers. From this early semi-religious literature spring the longer epic songs of a later date. It was principally on the occasion of the great sacrificial feasts that the heroism and liberality of the sacrificing monarch were compared with and glorified and magnified by the tales of the mighty deeds of the noble kings of early times. Prof. Weber seeks for the origins of the Mahabharata in the early epic tales recounting the exploits and tragic end of Kuru's royal race. The poem in its entirety, however, cannot have belonged to the Vedic period, there being, with the exception of a few names that are also met with in the Mahabharata, not a tracé of the Epic to be found in that period. This lack of evidence in the Vedic period can also be partly explained, however, by the fact that in the early ritual, even in a ritual so x) Barth-Oeuvres IV, p. 349- 2) Barth-Oeuvres III. 470 IV. 351 nt. 3. essentially belonging to that of the Kshatriya caste, as the Rajasüya, attention was given not to the narratives of the Kshatriyas but to those of the priests, the result being, naturally enough, that the Brahmins feit no preference for the insertion of purely Kshatriya narratives in their own sacred writings. This for Weber is a partial explanation of the fact that the epic hymns in the period of the Vedas are so closely bound up with the myths of the Gods, the human element not being so evident as it is in the Epic itself. From a hymn in praise of princely liberality to the epic is indeed a far cry; it is a fact that, according to the Mahabharata itself, the poem was recited on the occasion of a Sacrifice at a King's consecration (Rajasüya), but the work has been for ages more than a specimen of laudatory poetry. It is also a popular poem. And it must have been so for an exceedingly long time, there being a passage in Panini, from which it is concluded that in his time there already existed a certain Arjuna and Vasudeva (Krishna) cult.1) Although A. Ludwig (since 1884) contends that the Mahabharata „eine hauptwurzel in den verhaltnissen hat,die auszder vedischen zeit sich fortgesetzt haben" 2), he is really no more successful than Weber in his attempt to establish an organic union between this Vedic root and the Epic trunk. There is undoubtedly a similarity in a number of names, but that is all that can be said. Ludwig sees in the Epic two predominating motives. On one side there is the historical motive, set in an epic poem, which told initsoriginalformofthecaptureofKurukshetra on the river Sarasvati by the Bharatas. But the Pandavas are not historical. Yudhishthira, Arjuna, Nakula and Sahadeva are „einer unanfechtbaren Nachricht zufolge", members not of a single but of different tribes which were only later united under the patronymic or family name of Pandava. Since, then, the purpose of the work is not rendered credible by historical reality, Ludwig concludes that there must have been another motive that actuated the composition of the poem as we now know it, and that a mythical one. Ludwig considers as purely mythical features the blindness of the *) Barth-Religions de 1'Inde. Oeuvres 1. p. 100. 2) Abh. K. Böhm. Ges. Wiss. VI Folge Bnd. 12 p. 1. Kuru king, Dhritarashtra, the polyandrie marriage of the Pandavas, the meaning of their name (pandu = white), the rivalry between the Pandavas and Duryodhana. He sees in the entire story a seasonal myth. The five sons of Pandu, the pale sun, fight against the „difficult-to-be-fought" (Dur yodhana), son of the blind (power less) winter-sun,whose consort, Gandharï, covered her eyes (is wrapped in clouds) *). The five Pandavas are married to Krishna (the dark one = the earth) and loose all their wealth and hoarded gold (= their lustre, splendour) in the fatal game of dice with the base Duryodhana, until, at last, their wife, Krishna, is left in possession of only one garment (the earth becomes bare in winter). He supposes that gradually the human element encroached upon the superhuman in this seasonal myth, until, at last, a poet, realizing its possibilities, seized on the myth in its entirely humanized shape and grafted it on to the story of some intertribal quarrel. „Warum und wodurch veranlas^t ein dichter auf diesen gedanken kam, wird natürlich immer ein ratsel bleiben" 2), says Ludwig more or less artlessly. When Ludwig wrote, this nature-myth method was universally in use. Nowadays there will be but few scholars who have any confidence in such literary acrobatism. We might mention, as an intellectual curiosity, that Ludwig was entirely at a loss what to do with Krishna (black) at all, one of the principal actors in the tale. He wonders whether Krishna might perhaps represent the spring-sun, blackened, as it were, by the ceaseless smoke of sacrificial fires! It sfems to us that insufficiënt value has been attached to the advantages accruing to epic study from Ludwig's researches. We consider it of great importance that Ludwig should have directed attention to the mythical element in the Epic. That this should have been overlooked in criticizing Ludwig's views is quite understandable, but it in no way detracts from the intrinsic value of his labours. In the course of the present study we shall have occasion to return to his work again. To more or less the same time in which Ludwig was expound- *) Ludwig is evidently thinking of i.i 10.14. Gandharï covers her eyes in order to be no better off than her royal spouse. 3) Stzb. K. Böhm. Ges. Wiss. Phil. 1895. IX p. 15. when a close connection existed between Sivaismand Buddhism, and it is none less certainly known that there was once a collision between Buddhism and Brahminism, which knowledge induced Holtzmann to assume that the Sivaism of the Kauravas must have implied a certain partiality for the teachings of Buddha. Finally Holtzmann arrivés at the following historical reconstruction: Right back in the most ancient times there was a guild of courtsingers who extolled in their professional poetry the mighty deeds of their monarchs. Then came a talented poet who made of the original Epic composed in honour of the renowned race of the Kauravas a poem in praise of a great Buddhist ruler, perhaps Asoka. But now the new teaching, coming into conflict with the growing pretensions of the Brahmins, begins to decline, and the priests convert the now popular poem to their own use, but reverse the original purpose of the work as a whole. Now it is no longer the Kauravas who are lauded but their very adversaries, the Pandavas, to whom a decided predilection for Brahminical doctrine is ascribed. TheEpicissubjectedtofurther revision. Buddhism is eliminated altogether, both Vishnu and Krishna are thrust into the foreground, the Epic is assimilated to the ancient and sacred chronicles of the Puranas and portions of a didactic character are interpolated. And in this revised and irrecognizably altered recension the Epic was non-existentuntil the i2th. century A.D. The great drawback is that these alterations can only be inferred from the poem itself. External evidence is entirely lacking. And then it must be borne in mind that there are no traces of Buddhism to be found in the work at all, which Holtzmann himself frankly admits. The zeal and erudition of this scholar have won universal esteem, but nobody will nowadays be found to give his theories serious consideration, however ingenious they may be in them selves. „Quant a sa théorie de la formation du Mahabharata, elle est ruinée de fond en comble" x), that is the judgement passed upon his labours by Auguste Barth, a judgement universally considered final. It was chiefly G. Bühler, who, together with J. Kirste, demon strated the untenability on historical grounds of the theory of Holtzmann. In 1892 Bühler demonstrated in the first place, x) Barth-Oeuvres II p. 150. Oeuvres III p. 434- that Max Müller's idea of the existence of a great gap in Sanskrit literature, which, after being wrapped in the darkness of a sort of eclipse for several centuries, suddenly shone forth again in renewed splendour in about the fifth or sixth century, did not fit in with the facts as they are now known to us. On the subject of the Epic he further proved, that from the xith. century onwards there were no evidences of any additions of an important nature having been made. Moreover, that in the eighth century it was known in the form in which we now possess it, viz. as dharmasastra or book of sacred tradition (smriti). The inscriptions show that it was certainly known as early as the 5th. century as the work of the great rishi, Vyasa, consisting of one hundred thousand slokas, from which he drew the momentous conclusion, that it must have been extant in practically the same shape as we now know it certainly several centuries earlier than 400 A.D. On historical grounds, therefore, Holtzmann's theory must be considered entirely disproved. Bühler's and Sir Bhandarkar's inscriptional researches have shown that conclusively. But, although Holtzmann's far-reaching conclusions left the major ity of scholars unconvinced, there were several who feit more or less attracted towards what Hopkins called his theory of inversion in itself. And a man as L. von Schröder, for instance, was struck by the nature of the problem that the two Holtzmann s had set themselves to solve. „Was bewog nun wohl — so müssen wir fragen — die spateren Bearbeiter des Mahabharata, die Pandusöhne, welche in der alten Gestalt des Epos eine offenbar sehr hassliche Rolle spielten und nur durch Tücke und Verrath zum Siege gelangten, geflissentlich zu rechtfertigen und zu rühmen, die Kuru dagegen, die herrlichen Helden des alten Gedichtes, mit Anschuldigungen und Schmahungen aller Art zu überhau- fen? Dies seltsame Faktum ist nicht leicht zu erklaren, und werden wir bei einem Volke wie die Inder ausser den politischen Motiven von vornherein auch religiöse als mitwirkend vermuthen müssen, — aber welche waren es und wie hing die Sache naher zusammen?" 1). Prof.Von Schröder is of opinion that there was first an epic poem 1) Lit. u. Cult. p. 479. composed by the Kuru bards and that at a time when Brahma was the leading god (7th.—4th. century B.C.). This poem sang of the wars between the Kurus and the neighbouring tribes. These foes of the Kurus had a tribal god, called Krishna. The Kuru bards naturally saw in the god of their enemies an evil being that they could not possibly paint black enough. But ultimately the conflict is decided against the royal race of Kuru, which is humiliated. The new generation was politic enough not simply to scrap the older poem. They retained it, but recast it in their own favour. They also cleverly exploited the contents of the popular Kuru Epic by remodelling it to their own advantage. It was in this manner that all those parts of the work came into existence in which the deeds of the Pandavas are glorified and those of the Kauravas condemned. Thus, too, it was that all those parts originated in which Krishna is honoured and lauded as the highest and holiest of the gods, as the incarnation of Vishnu, and which form such a feil contrast to other passages in which the same god is represented as the auctor intellectualis of all kinds of fraudulent devices. „Der zuerst fast unbegreifliche Widerspruch löst sich aufs Schönste bei unserer Annahme: dem ersten alten Dichter, dem Sohn des Kuru-Landes, muszte Krishna im schwarzesten Lichte erscheinen; den spateren Bearbeitern aber war er der erste, berühmteste Held ihres Stammes, die Gottheit selbst — und in diesem Sinne feierten sie ihn Eine ungezwungenere Erklarung dürfte sich schwerlich für den bisher noch nicht gelösten Widerspruch auffinden lassen". But to us this solution does not seem to be so simple; there is some difficulty in understanding how a laudatory poet or panegyrist can make a poem in honour of his lord in which the party extolled continues to play a villainous part. Von Schröder requests us to consider how differently a Frenchman and a German would judge of a man like Bismarck. But that is not the same thing; the case is different. We should have to imagine a Frenchman and a German, each passing judgement in brotherly agreement, the one in bonam partem, the other in malam partem, without either of them detracting frcm the homogenity of the whole. One is, anyhow, bound in all conscience to admit that the entire: inversion theory is simply an ingenious hypothesis that must be a proof of itself. The inversion theory appears in another shape still here and there.Sir g. grierson,forinstance, is of opinion that once there was a war between the Kauravas of the Madhyadesa and the Panchalas with a struggle for supremacy in the background between the Brahmins and the Kshatriyas. In that case the fons et origo of the epic war would have been an insult offered by Drupada, king of the Panchalas, to a Brahmin, who thereupon took refuge with the Kauravas. Grierson's point of departure is here the identification of Pandavas with Panchalas, as Weber had already done before him. He took these Pandava-Panchalas for a people considered from ancient times by orthodox Brahminism asunorthodox, which contention would find support in certain customs practised by them, such as e.g. polyandry x). A. B. Keith's objection to this theory is that in the first place there is nothing known of a conflict supposed to have raged between Brahmins and enemiesof Brahminisminthetime of the Brahmanas. In the second place — we may consider the relation between Pandavas and Panchalas as proven, but there is nowhere any mention in the more ancient literature of a war between the Panchalas and the Kauravas. On the contrary the relations between the Kurus and the Panchalas are there shown to have been of a friendly nature. One cannot, therefore, possibly change the war between the Kurus and the Pandavas into a war between the Kurus and the Panchalas 2). In spite of the fact that the inversion-theoryhas been severely — but justly — criticised we must still bear in mind, that, in point of fact, the problem posited by the two Holtzmann's, still awaits satisfactory solution. And yet the question of deception in the Epic is, in our opinion, of vast importance. The matter will demand our undivided attention at another stage in the present work. Personally we cannot believe that Hopkins is doing full justice to the deceptive element in the poem, when he alleges in explanation of this ') J.R.A.S. 1908 p. 602, 831, 843, 1139, 1143. 2) J. J. Meyer also thinks the Pandavas to have been a people of nonnoble, aboriginal origin, whose relationship to the Kauravas was artificially constructed: later on. (Das Weib im altindischen Epos — Leipzig (1915) p. 82). astonishing moral outlook, that we have simply to do with two different stages of culture, An earlier age allowed what a later condemned. The peculiar circumstances we are now considering might then be due to the fact that the poets of the Mahabharata were actuated in their labours by two opposing tendencies, viz. by the tendency, on the one hand, to entirely remodel the whole poem as it had been handed down by word of mouth with a view to establishing a new ideal, by the tendency, on the other hand, to keep the original epic with all its sacred traditions, intact, and thus hand it down untouched to future generations. „Is it not reasonable" — asks Hopkins in his well-known essay written in 18861) — „is it not reasonable to suppose that those same priests who framed the fighting code and endeavoured to implant in the brutal warriorkings a moral, not to say a chivalrous sentiment, might have been swayed by two opposing desires in handing down their national Epic?" We think that this question should be answered in the negative. In the first place the literature of the Vedas furnishes us with no facts at all from which to conclude positively that these so-called „brutal warrior-kings" ever really existed. And, moreover, now that our knowledge of primitive cultures has increased and increased considerably, there is no longer any reason at all to picture these more primitive cultures to ourselves as a sort of barbaric state of human society in which „brutal warrior-kings" were the order of the day, although, taking into account the time in which Hopkins wrote, itisthoroughly understandable that this scholar should then have come to look at things as he did. We must be careful not to figure to ourselves those bards of an older day, unshackled in their vision, in the garb of scrupulous philologists, watching with jealous and meticulous care over the unity and the purity of their text. These men were not so concerned with the Epic's being a homogeneous whole as to worry about its showing a lack of unity. Had they had that scrupulous veneration for the poem in its inceptive form inseparable from its preservation intact, things would have !) The social and military position of the ruling caste in ancient India. J. A. O. S. XIII pp. 57—372. found there as it is elsewhere, to the existence of which the singing-contests bear clear witness1). One has only to read Nguyen van Huyen's book on the selfsame practice to be convinced that we have here to do with potlatch customs. And such were not missing in Indian culture, which we ourselves hope to be able to demonstrate.2) Here again, then, the idea of „culture" comes into the limelight. We consider the constant puiling to and fro of cultural elements in itself a fruitless field of study, whether the type of culture concerned be classic and Indian or Austro-Asiatic and Aryan. It is wiser to lay aside for the time being the vexed problem of the introactive influence of the various types of Indian culture the one upon the other, and to try and find out in the first place what there really was then. We believe that the study of the Epic and of the Puranas is adopting a dangerous course in spending its energies endeavouring to find out what influence the Austric or Dravidian part of the population of India has had upon the origin of Hinduism.3) A culture must first be studied in its entirety, after which the problem of cultural influence may be taken up with a greater chance of arriving at satisfactory results. Ethnology, especially in its infancy, has, as a matter of fact, not always taken the idea of „culture" seriously. It has been the habit for years to write volumes on all sorts of things that at first sight resembled one another. The French scholar, Dürkheim and his school, severely condemned that sort of thing. It is not a question of collecting a number of cultural curiosities illustrative of various customs and practices in Indian culture, but of demonstrating, that ancient Indian civilization can be rendered intelligible and in part be explained by attending to what the reality of the present has still to show us among all sorts and conditions of peoples irrespective of race or home. A scholar of Caland's reputation considered the study of ethnography of some importance to philologists because of one s then „being able to point out parallels in the ethnographical data of i) Les Chants alternés des Garpons et des Filles en Annam. Paris 1934. v. Introduction. ') The svayamvara is a contest. MBh. VI. 93, 42. s) Gonda-Austric and Aryan p. 15. To discuss all the merits and demerits of the different theories about the caste system would be out of place here 1). For that matter, only a very moderate part of all the theories about it deserves mention. The investigators have not always succeeded in keeping their imagination from indulging to their heart's content in the even field of prehistory, where facts are no longer impediments. Nor shall we try to concoct a new mixture of these different theories. The truth does not always lie midway, for the truth is not polite. In this case the truth is nearest to Émile Senart in our opinion. This statement is not too bold, for Senart s theory is almost generally accepted as the most felicitous. Let us first just sum it up in a few words. Senart considers caste as „une institution organique qui puise sa sève h des sources trés profondes." This source is the social structure of the old-Aryan population and this is found among the Greeks and Romans as well as among the Indians. „Les termes ici se correspondent trés suffisamment: gens, curie, tribu & Rome; familie, phratrie, phylè en Grèce; familie, gotra, caste dans 1'Inde. L'harmonie générale est frappante."2) In this organization marriage regulations are qualified by the endogamy of the tribe and the exogamy of the gotra. In the Greek and Roman cultures this social organization is totally or partly ousted by the political power of the kings. The lamp of history is not lighted before this ousting process is for a good part accomplished. In India this ancient form of organization is better preserved, not quite intact, it is true, but clearly identifiable. About this gradual modification which the social organization of the Aryans must have undergone before the caste system resulted from it, Senart does not provide much further information, yet he thinks of a fairly strong modification, for, he says, at the outset it is not easy to recognize in the caste the primitive organization. The most ancient Aryan immigrants spread over the just conquered country in numerous villages. Owing to this dispersion the organization became somswhat diffuse. In those villages whose inhabitants were united by a real or fictitious relation- n See: L'Inde Ant. p. 93. Further C. Bougie — Essais sur la regime des castes; Sarat Chandra Roy-Caste, Race and Religion in India-Man in India XIV p. 39 and p. 75. 2) Les Castes dans 1'Inde p. 222. ship, the primitive clan was continued in a modified form. The Aryans — and here the race contrast in Senart's theory comes to the fore — refused to take up certain professions which they thought to be unclean for religous reasons (of an unknown character). The few artisans there were among the Aryan(pastoral) tribes spread over the country and organized themselves after the usual pattern they saw around them. In this way the professional division of the castes came into existence. It seems improbable to Senart that the caste should originate from the tribe of the autochthonous population, because the Brahmins were indeed the keenest champions of the caste system. Still he also finds room for the influence of the native population. For this influence he discovers in the „caste-like" organization of various trades. Now Senart puts this theory to the test by comparing it with the available data. The oldest records, the Vedas, are almost silent about the social organization. But this silence of course need not be a proof of the absence of castes in that period. In the literature of the Brahmanas four castes are always mentioned, usually termed the four colours. The word varna means both colour and caste. According to the representation of this theological literature God created in the beginning the four varnas, the well-known fourfold division of Brahmins, Kshatriyas, Vaisyas, and Südras. According to the Brahmanas this number was so much increased, because people were so infatuated as to scorn caste endogamy, which was ordained by God. In this way a blending of castes was effected, from which in the long run sprang the numerous castes we meet in India later on. In the Jatakas, a literary genre still younger than the Brahmanas, Fick finds a caste system indeed, but the rigid organization of modern times cannot yet be found in this period, in his opinion. The Epic, the legal codeofManu, which were finished still later than the Jatakas, rather show the detailed caste system as this is known in modern times. x) Consequently from the literature itself we can form a very plausible conception of the development of the Indian social organization. The oldest literature knows little about caste ') Die soziale Gliederung im Nordöstlichen [Indien zu Buddha's Zeit, p. 173. and only once in the well-known Purushasükta do we find the usual fourfold division into Brahmins, Rajanyas, Vaisyas and Südras. More recent literature speaks of a blending of the four primary groups, out of which more come into existence. Still more recent literature begins to show the aspect of the modern caste system, which cannot nearly be compressed in a division into four. Still Senart cannot accept this seemingly obvious development in this form, because he believes caste to have sprung directly from the Aryan tribe. It stands to reason that it would be entirely unacceptable to suppose that in the beginning only four tribes invaded India and that from these four tribes the caste system proper should have sprung. According to Senart the castes were given simultaneously with the tribes, also in Vedic times, although the system has been considerably extended since. Caste is called jati, according to Senart, and the word varna which the Brahmanas mention does not really refer to castes, but to four states or classes. The authors of the Brahmanas mention indeed only four castes, but Senart thinks this is only a theoretical simplification of the real state of affairs. In reality there were many castes (jati), just as there were also many tribes. There were, however, only few classes (varna) and this is not distinguished in the Brahmanas. In this number of four the rich variation of reality is not nearly given in its entirety. It is quite unwarrantedtoput full confidence in the Brahmanas. They are much too artificial. With the Vedas it is quite different; the latter are more natural and more reliable. Also the organization into classes, to which the word varna refers, is not limited to India. There is a striking agreement between the four Pishtras of the Iranians and the four Varnas of the Aryans: The Athravas are equivalent to the Brahmins; the Rathaesthas to the Kshatriyas; the Vastriyas-Fsuhyants to the Vaisyas; the Hüitis to the Südras. This class differentiation (division into var nas) introducés into the caste organization (division into jatis) the element Senart lacked for the completion of his theory, viz. the status qualification. For the caste system implies a precise division into several degrees of status qualification, which is not simply given as such in a number of tribes that exist side by side. It is the division into varnas, stimulated by race contrast, that made the old Aryan organization in tribes into the caste system, the system of jatis, which at the same time fixed a person's social status. This is in short Senart's theory, which has almost universally been accepted. On one point it has yet been elaborated, namely a further explanation is given of the fact that the status qualification came to prevail so strongly in the caste. This phenomenon is thought to be explained by the strong corporative feeling that linked the priestly group together. The priests, officiants as they were of an extremely intricate sacrificial ceremonial, were the first to get so far as to develop group exclusivism. It is owing to their influence that also other groups came to cultivate the same exclusivism after their example. C. Bouglé puts it as follows: „C'est 1'habitude du culte fermé des premiers groupes familiaux, qui empêche les castes de se mêler: c'est le respect des effects mystérieux du sacrifice qui finalement les subordonne k la caste des prêtres." *) Prepared by the old family organization and instituted by the Brahmins, — in this corrected formula Senart's theory is accepted by many. This influence of the Brahmins was so much put in the forefront, in order to evade the objection to Senart's theory: That in his theory of family organization he was thinking too much of typically Indian phenomena, which, owing to local, historical circumstances only occurred there. For a family organization in the sense Senart took it, has long since proved not to be restricted to India. It may even be stated that the status qualification through the membership of a group, to which a person belonged by birth, is rather generally found in primitive societies. Some kind of status qualification is given in every clan organization. So there must be another factor to account for the fact that the status qualification could grow into such an elaborate caste system. So it was not the intention to say that the Brahmins simply invented the caste, (which has also been asserted) but that it was the influence of their corporation that conferred on the Indian caste system its more intensive application. On the one hand this is an improvement on Senart's theory, for the racial contrast now gains a religious *) op. cit. p. 82. aspect besides. But on the other hand this is again a disadvantage. Indeed Senart considered caste to be a typically Indian phenomenon. This is not, however, the proper consequence of his theory as such, but is, in our opinion, rather due to the data that were available in his time (1896) with regard to the organization of the ancient peoples. Also Senart himself attached some value to a certain status qualification, for such is the division into varnas to him. Indeed this weak form of status qualification alone does not account for the exclusivism of the caste, but the somewhat mysterious influence, ascribed to the priestly group, is not only difficult to understand, but again locally Indian. Senart's theory, which has at any rate the advantage of being clear as a hypothesis, now becomes more or less vague again. The correction introduced in this way can only account for caste phenomena in India. Now it should not be said that a caste system is in fact only distinguished from a clan system by a more intensive application of status qualifications. No, we must stick to the fact that clan organization is especially characterized by the exogamy of the clan, and caste organization particularly by the endogamy of the caste. Caste and clan are not only distinguished from one another by a greater or smaller degree of exclusivism. If this were the case the degree might indeed be explained by local factors. For not only a certain status qualification is found in the clan organization outside India. But also certain caste phenomena are found there, endogamy (different from that of the tribe) and functional groups (certain family groups that have functions in society of an economie, religious or social character). If it became evident in course of time (as it has already become, to some extent) that also caste phenomena are not specifically Indian, the acceptance of the influence of the priests in India, as was suggested above, would have led to deprive Senart s theory of some of its suppleness (the possibility to apply it elsewhere with such modifications as are required by circumstances). Before we shall further expatiate on this we should like to discuss some objections to Senart's theory. His theory fully deserves the appreciation it was generally received with. It was not an insignificant matter to understand as early as in 1896 the character and implications of the social organization. At that time only little was known about primitive social structures. The regulation of marriage (which is connected with endo- and exogamy) has proved to determine the social structure for no small part. There are, however, objections as well. Not in the first place that Senart based the social structure so much on Aryan phenomena. The question of cultural influence should not be made prominent in India, nor will it perhaps become so in future, as far as the oldest periods are concerned, simply because of the lack of data. But a more serious objection is that Senart found himself obliged to contradict the after all rather positive data provided by Indian literature. As has been stated, the literature gives the impression that the caste system is not so old as Senart wants us to believe. In the Jatakas Fick could not yet discover a well developed caste system. „Ein Zerfallen der Brahmanenkaste in verschiedenen Unterkasten, ein Sichvereinigen der aus ihrer Kaste Ausgestossenen zu neuen Kasten, wie es im heutigen Indien besteht, ist, glaube ich, für die altere buddhistische Periode nicht anzunehmen, da wir nirgends in den Palitexten eine Spur davon finden. Auch dass die brahmanischen Gesetzbücher nichts davon wissen, möchte ich nicht mit Senart.... auf das Bestreben der Verfasser die Kaste in ihrer idealen Integritat darzustellen zurückführen, vielmehr daraus schliessen, dass erst in neuerer Zeit, wo die Kasten mehr und mehr den Charakter von Berufsgemeinschaften angenommen haben, die alte, wenn auch nur in der Idee bestehende Einheit der Brahmanenkaste zerstückelt worden ist." x) Also Oldenberg has some objections to Senart's hypothesis. That the caste of the Brahmins should have split up into a number of sub-castes at a rather early date, Oldenberg does not consider to be proved from the appearance of thieves, butchers a.s.o. in that caste. At best it proves that of old from early times not all Brahmins were ideal Brahmins.2) Questionable is also the difference Senart had to make between Varna and Jati. The use of these words does not justify 1) op. cit. p. 125 n. 1. 2) Z. D. M. G. Ll, p. 277. a supposed difference between Varna (class) and Jati (caste). Fick has clearly pointed out that the two terms may refer to one and the same group. Then not all experts agree with Senart on the point that the Brahmanas are to the Vedas as an artificial to a spontaneous genre of literature, a contrast Senart was in want of, in order to explain why the Brahmanas know only of four castes. P. Masson-Oursel puts it to Senart's credit that he indicated precisely „la différence entre les cadres sociaux aryens et brahmaniques." *) But Oldenberg cannot discover the gulf between Vedas and Brahmanas indicated by Senart.2) Senart's philippic against the philologists,whosimplyacceptthewording of their text, he thinks to be entirely out of place. It is, however, as Macdonell remarks, inconceivable why the Brahmins should so doggedly stick to a rigid division into four if this had been completely at variance with reality as they saw it daily around them. 3) What weighs most heavily with us is the objection, loosely formulated by Joseph Dahlmann, that the clan is exogamous and the caste endogamous. This formulation is loose because Senart did not parallel caste with clan, but with tribe. Over against the clan he puts the gotra, which is indeed exogamous. Senart's theory cannot so easily be reconciled with what is known about clan organization at present. The conception he had of the bilateral family group, the exogamous clan and the endogamous tribe is not quite correct. The exogamy of the Indian bilateral family group, the sapinda exogamy, which extends as far as the fifth degree of the ascending line, is of a quite different character and has an influence on marriage regulations totally different from the real family nucleus in a clan system.4) It would betray a wrong conception of a clan system to hold that tribal endogamy may be taken as marriage restriction and on the other hand clan exogamy as marriage enlargement. Every clan often stands in a fixed connubial relation to some other clan (sometimes to more than one). The members of those clans that are connubially coordinated, x) L'Inde Ant. p. 96. 2) Z. D. M. G. L1 p. 375 and p. 288. 3) Am. Hist. Rev. XIX p. 235. 4) See p. 46. Krishna is usually explained euhemeristically; now and then a mythical figure is started from or a combination of both. As the result of a kird of progressive movement the existing religion is then supposed to have caused the development of new elements, or the other way round these new elements are supposed to have attracted the ancient religion and given it new life. Auguste Barth, having to make a choice between the two possibilities: 1 .Vishnu is identified with Krishna, thereby rising in importance; or 2. Krishna is identified with Vishnu, thereby rising in importance — chooses the former, Krishna thus becoming the central point about which epical investigation revolves. Sir Bhandarkar, however, is of opinion, that in the Upanishads Vishnu himself is already on the way to becoming the supreme deity. Bhandarkar endeavours to see in the person of Krishna the confluence of four „streams of religious thought"; viz. one having its source in Vishnu, the Vedic god; one in Narayana, the cosmic and philosophicgodjonein Vasudeva, the historical god; and one in Krishna-Gopala, the popular lover-god of the Abhïras. Bhandarkar sees the power that made these four streams flow into one basin in a rising theistic persuasion born of the urgent need of ,,an adorable object with a more distinct personality than that which the theistic portions of the Upanishads attributed to God for ordinary people". *) One of the sources here mentioned, then, is the belief in Narayana, who is spoken of in the Narayanïya fragment in the twelfth book of the MBh. In this fragment it is related that Narada once journeyed to the White Island (Svetadvïpa), where Narayana reveals himself as the All-Being that manifests himself to all whose aim He alone is (Ekantin), Narayana, as the name itself says, is the one and only goal of the gods or of men; the word, Nara, especially in the Vedas, means both gods and men. Narayana is likewise associated with the waters, it being said of the waters, that they are Naras. Among his avatars Narayana also mentions him who shall be born of woman to slay Kamsa in Mathura, and who, after having defeated many demons will establish his rule in Dvaraka. *) R. G. Bhandarkar-Vaisnavism, Saivism and minor religious Systems — Gr. I. A. Phil III. 6. p. 2. for himself and his brothers a beautiful woman, whom later on they lost again." *) Initiation, being an important tribal ritual which is met with in a form of human society where clan-organization prevails and one which obtained in India certainly not later than several centuries B.C., it is improbable, that the worship of the Childgod, Krishna, should ever have originated and developed under the influence of early Christian thought and religious belief. The hero of „The City of the Gates" stands out clearly before us in the shape of the Divine Trickster and is particularly connected here with the ritual of initiation, as in other cultures he is associated with demons or with the Supreme Being. By a lucky chance the worship of the male pair, Narayana-Nara, men's refuge - man, Krishna-Arjuna, black-white, is known to us from Panini. That fits in perfectly with our own idea of the figure of Krishna; and a cult, which is inexplainable in itself at once becomes explainable. It also accords with our idea, that the ritual of initiation must have obtained in the social structure of ancient India as far back as several centuries before Christ. By a no less happy coincidence the fact of the oldest data bearing upon the drama, the Kamsavadha, being found as early as in the Mahabhashya of Patanjali is a second string to our bow. Rassers having demonstrated the close connection obtaining between ritual and drama in its primitive form we are fully warranted in seeing in this Kamsavadha a dramatization of the primitive initiation ritual. Krishna's doubling the röles of initiator and initiate is further admirably illustrated by the undecided manner in which family relationship is determined, Kamsa being at one moment Krishna's uncle and at another his cousin on his mother's side. We shall not in this place enter into the question of the value of all this for epical studies. In order to avoid anything that might tend to make the trend of our argument less clear we shall only venture some observations upon the point in question after certain other points have come under discussion. * * * When the initiation rites together with the rest of the cere') W. H. Rassers — De Pandji Roman p. 286. monial constituting the tribe's great religious ritual are to be celebrated, it is essential that the entire community shall come together. This coming together of the various dispersed units of a tribe, their contraction or concentration about a central idea of cosmic unity encourages and promotes a more intensive religious and economie life. Now when we compare this concentration of the dispersed units with the contrary normal period of deconcentration, during which the various members live apart, each within his own domain, we become aware that in the tribe's social intercourse there obtains a kind of seasonal variation, a constant succession of two phases in the tribal existence, one of concentration and one of dispersion. These „variations saisonnières", as this succession of phases in tribal intercourse is termed by the French scholars, Mauss and Beuchat, are apparently to be met with in almost all states of primitive society.x) We come up against them not only among the Eskimos but also among the American Indians and the aborigines of Australia. Obviously these periods of concentration are closely bound up with the habits and the mode of life and thought as well as with the entire economie life of the tribe, in a word, with its ecological environment. When, aware of the fact that the whole idea of initiation in the Indian mind was of such paramount importance, we start off investigating whether similar phases of tribal intercourse are also to be found in Indian culture, we are immediately struck by the practice of dividing the year into two parts, the Devayana (the path of the gods) and the Pitriyana (the path of the fathers), also called the Uttarayana or Northern path and Dakshinayana or Southern path. The connection between this division of the year and the succession of the seasons immediately becomes clear once it is known that the two „paths" have their starting-point at the winter solstice and the summer solstice respectively. Parallel to the division of the year into two astronomical halves there is also a climatological division of the year into three seasons, each of four months: the rainy season which sets in immediately after the summer solstice, followed by the cold season in turn succeeded by the hot season. The ') M. Mauss et H. Beuchat — Essais sur les variations saisonnières des sociétés Esquimaux. Ann. Sociol. IX — 1906. serpent, Sesha; during this season no marriages are contracted; now is there „nor vital breath no- potency in Nature". This critical period at an end, the merchant balances his books and closes his financial year. Now Vishnu wakes again. The new year begins.1) What we here have to say concerning the Indian calendar will certainly not lay claim to being in any way complete. 2) Our observations have been made more or less tentatively. Our concern is merely to draw attention to the fact that the rainy season in India shows a very distinct line of demarcation separating it from the rest of the year. The rainy season is a time of ill omen, of war, antagonism, risk, danger, mischance and general unrest; everything seems to be upside-down; the remaining period is one of peace, of safety and security. This line of demarcation running between the rainy season and the ') We should not be surprised if the Pitriyana, i.e. the rainy season, were once thought of as being the time in which the Asuras were puiling the churning-rope in the cosmic ritual of the Amritamanthana, causing the churn to turn from right to left. Now we have seen in the description of the cosmic ritual of the churning of the sea that, the gods having had their pull at the rope (serpent), it is the Asuras' turn to pull likewise but in the opposite direction. There is every possibility of this ritualistic churning having been conceived of not as having taken place and been completed once and for all but as still being performed by the cosmic participants. In that case it is quite probable that the Devayana is thought of as being the pravritti of the cosmic churning process, and the Pitriyana the counter-movement, nivritti. In the Devayana, then, the gods are supposed to be puiling the churning-rope — the churn-staff turning from left to right — whereas in the Pitriyana it is the Asuras who are thought to be pulling — the churn-staff turning from right to left. The idea would then be the same as Vishnu's sleeping upon the serpent, Sesha, during the rainy season, which we associate with the pralaya followed by a srishti. We imagine the idea contained in the word, srishti, to correspond to the idea expressed by the word, pravritti, and pralaya to nivritti. Not having succeeded, however, in gathering any positive evidence to justify us in crediting our conjecture with any measure of correctness, we submit it to the reader's attention for what it is worth — a mere conjecture, nothing more. 2) It is not always an easy thing to estimate the data at one's disposal at their true value. Investigators sometimes attach too slight a measure of importance to these feasts, or else they bestow undue attention — even frequently enough to the exclusion of practically everything else — to the „nature-elements" in such ritual feasts (summer solstice and the like), whereby other customs and ideas are not infrequently ignored as being „foolish and absurd". The employment, too, of terms such as Spring, Summer, Autumn with reference to Indian seasonal conditions is more often than not an awkward and indefensible habit, leading to unnecessary confusion, such terms possessing a significance in Europe with its entirely different climatic peculiarities not applicable to the seasonal demarcations of the Indian year. remaining part of the year being even nowadays still more or less distinctly discernible, we are with reason justified in assuming that things were not otherwise at an earlier period of Indian civilization, which means to say, that it can by no means be deemed inadmissible to consider the paucity of data at our disposal concerning the older civilization of India in this light. In the first place mention must be made of the S a r p a b a 1 i, the annual Serpent Ceremonial, described by the Austrian scholar, Winternitz. The serpent, this amphibious creature of the waters, the born rival of the monster-bird, Garuda, was invoked and sacrificed to, according to this ceremonial usage, throughout the entire rainy season, and especially at the beginning and at the end. *) The three seasons (the wet, the cold and the hot period) were separated from one another by a Chaturmasya sacrifice. The ritual at the beginning of the rainy season was decidedly of the nature of a propitiatory sacrifice, everyone making confession of his shortcomings. Upon this occasion it was customary to make a barley-meal effigy of a ram or a sheep provided with huge genital organs. The second of the Chaturmasya sacrifices was accompanied by a great Feast of the Dead together with a sacrifice to Rudra Tryambaka. Maybe this festival was not the beginning of the period following upon the rainy season, but a provisional termination of the rainy season itself, a desacralization rite, therefore. Another important festival is that of the Winter Solstice. Then takes place the ceremonial we have already spoken of. It was customary for a Brahmin and a Südra to contend with one another for the possession of a round white piece of leather, the Brahmin ultimately becoming the possessor. The earth-drum was beaten (a hole in the ground with bull-hide stretched across it); an arrow was shot into a tightly stretched cowhide; girls danced round the fire with pitchers of water, the water afterwards being thrown into the flames. Ritual cohabitation was practised. A Brahmin had to sit upon i) The feast of the Sarpabali together with the Indian Serpent cult in its entirety is frequently mentioned as being of Dravidic origin. We are reminded, however, by Prof. Vogel (Serpent Lore p. 6) that we have no reason to assume that anything unmentioned in the Rigveda could not on that account have existed in that distant period. a swing. It will not be considered too hazardous an assumption to see in this feast at anyrate a feastof good omen.The,,auspicious" ritual is here at its height of „auspiciousness" betokening all that is good: there is life and procreation in fullness; fireworship and cohabitation. The darkness of the Pitriyana, the gloom that htmg over „the path of the fathers", is now entirely gone. Professor Hillebrandt, the first to realize the influence of the climatological division of the year upon the development of mythical thought and the elaboration of ritual, contended that Agni in the waters is a mythological representation of the sacrificial fire being extinguished, the which was kindled anew only after the expiration of the rainy season. If further, more convincing evidence be required of the rainy season being „serpent-time", a time of ill omen and not of „good-omened" ritual, it will only be necessary to give one's consideration to the great Soma sacrifice, which, it was believed, should always be celebrated by preference in Springtime, although apparently no special time was fixed for the celebration of this most important rite. The rainy season was as much a time of taboos in the good old days of yore as it is now. The members of the Buddhist orders were also forbidden to undertake a journey in this season of ill omen.x) There may be still more of such religious prohibitional sanctions to be found for the varsha. But even more important for our present purpose is the fact, that it is at the commencement of the rainy season that the Upakarman takes place, the inception of the probationary period for the Brahmacharin, the youthful novice or initiate. This probationary period terminates at the end of from four to six months, the young neophyte's discipleship thus covering the rainy season only or the whole of the Pitriyana.2) Now it is a very remarkable thing, that it is required of him who studies the sacred scriptures upon his own account, the private student of holy lore (svadhyaya), that he shall pursue his studies beyond the boundaries of the village, to the East or North thereof or in between, and in a spot from which the roofs of the houses are no longer visible. It behoves *) Rhys Davids — Buddhist India p. 112. *) Hillebrandt — Ritualliteratur. p. 58. connection is to be found between the smasana and this central sanctuary. That this relation exists may be expected with good reason. In the sacral ritual in the house of meeting one sees the ancestors of the clan appear. Not only is the younger generation admonished by its elders, the latter also turn with respect to the ancestors, who, in their lifetime, maintained the ritual as a divine institution. The ceremonies attending this relation with the ancestors may of cotirse differ locally. To further elaborate the supposed relation between the sacral house of meeting and the smasana, we call special attention to some points from a remarkable essay by Heine-Geldern.1) In this essay he points to „die für das ganze Zentralindischsüdost-asiatische Megalithgebiet so charakterisusch enge Verbindung von Totendenkmal, als Geistersitz, und Sitz- und Versammlungsplatz der Menschen." In various places it is the custom to erect to the deceased whom, for some reason or other one would like to remember more permanently, a stone memorial, often consisting in a number of upstanding stones, which are masculine, and a number of lying stones which are feminine A little pond often forms part of it. It is not necessary that this memorial should be erected by surviving relations. One can also erect it to oneself, a prerogative which is among the Angami Nagas connected with the most expensive of the feasts through which one can enhance one's social importance. These memorials often also serve as seat to the man who erected them during his life time and as seats for the spirit, and as general sitting- and resting-places. The memorials not only closely resemble burial-places, it is also well-known, that sometimes the skulls of the deceased relatives are kept underneath. Some Naga-tribes erect memorials to their chiefs, which are arranged in a large circle. Presumably connected with this is the large circle of big stones, inside which the youths wrestle at the annual feast to the dead. In the villages of the AngamiNagas a dancing- and feasting-ground lies in front of the house of the hereditary priest, in which the former hereditary priests are buried, and which is also used for meetings and councils. 1) Die Megalithen Süd-Ost Asiens und ihre Bedeutung für die Klarung der Megalithenfrage in Europa und Polynesien. Anthropos XXIII/1928. p. 276 seq. 15 Sometimes these sacral places of the Angami-Nagas are dug out a little in the ground. Sometimes they are circular, with cube-shaped boulders round them, on which the men are sitting during the council. Oftener yet a semi-circular stand has been erected in the style of a classical theatre. These meetingplaces too, are, according to old traditions, memorials to the departed, under which deceased people are indeed still buried. At the annual feast to the dead wrestling-matches are held here. This ground is also used as a dancing-place. The whole complex of memorial, seat for the chief, burial-place, meeting-place of the men, dancing- and wrestling-place Heine-Geldern finds as far as East-Indonesia.x) Returning, after this digression, to the material put at our disposal by lndian archeology, we may try to answer the question whether the data mentioned so far teil us anything concerning the function of the ancient structures which time has spared for us in India. We are thinking of the chaitya. In Fergusson we read: „The word Chaitya, like Stüpa, means primarily a mound or tumulus, but it also means a place of sacrifice or religious worship, an altar — from chita, a heap, an assemblage etc." The chaityas are not only monuments for the dead, but also „assembly halls". „They correspond with the churches of the Christian religion: their plans, the position of the altar or relic casket, the aisles, and other peculiarities are the same in both, and their uses are identical in so far as ritual forms of the one religion correspond to those of the other. 2) It is generally thought, that the chaitya was gradually built around the tumulus proper, more or less in the same way as a museum might be said to be built around the collection. Our point of view, however, does not imply that a tumulus was the only starting-point of importance of the chaitya. Let us inspect for instance the oldest chaityas that have been preserved, the stüpas proper, therefore. The specimens found, dating from the third century B.C. consist of a hemispherical dome usually raised on a terrace. On this hemisphere (anda) we notice a sort of kiosk (harmika) which is again surmounted by some 1) On men's home, centre of cult of the dead, ev. burial-p consult H. Schurtz-Altersklassen und Mannerbunde p. 208. 2) History of lndian and East. Arch. ed. 1910 I. P- 55- times may also have been present in the chaitya. Crooke mentions at any rate the fact that in the cult of the Bhümiya (also known as the Kshetrapal, „the guardian of the mound on which the village stands") the cult of the earth, of the sacred buil, and of the deceased ancestor are united. The Bhümiya is sometimes identical with the common ancestor of the village. „When a village is founded a mound is raised and near it a Jand tree (Prosopis spicigera) is planted; the first man who dies, whatever his caste may be, is buried or burnt on this mound, and a shrine is dedicated to him as the village guardian." x) Even in its name the modern temple has an obvious reminder of the ancient sabha. The assembly hall of the big temple of Mïra Baï (Chitor) is called sabha-mandapa by Havell. 2) Moreover every village contains a tree which is regarded with special reverence and serves as trysting-place for the villagers. 3) The relation between village-temple and tree appears to be beyond dispute. „The village shrine in which the collective godlings abide is, in the first place, closely associated with the sacred or guardian tree of the community." 4) Such a sabhavriksha stood on a crossroad. 5) Now it will already nave been observed that the sacral village-house is situated inside the settlement, whereas in our argument it has constantly been assumed that the sacral festive building had been assigned a place outside the settlement, in the woods. In this respect, however, India again does not constitute an exception. In this connection we consider the village as a social unit in the same sense as has been done in the case of economy and art. Also in other civilizations the sacral meeting-place is sometimes found within the settlement. In the case of exceptionally important rituals a seperate festive building will often be erected, outside the settlement. The temple in the village, which according to its origin is still doing duty, as of old, as men's house, clubhouse, guest-house or *) Rel. and Folkl. of North.-India p. 95. 2) Handbook etc. p. 70. s) This „trysting tree of the village", „tree of wisdom", Havel connects with the tree, so well-known from Buddhism and Jainism, „for all Indian art derives ultimately from the life of the Aryan village". Anc. and Med. Arch. of India p. 49. *) Crooke-Religion etc. p. 88. 5) Dutt-Townplanning in a. India p. 155. something of the same kind, yet stands, in the vilhgers' opinion, in a place, not quite suitable for especially sacral rituals. Dr. Rassers illustrates this with the following example: „In Seran the ordinary men's house, the baileo, and besides also the so-called priests' baileo (baileo anakota) stands in the village; but when they have decided to celebrate the religious feast par excellence, i.e. the ceremony of initiation (this is called with a very telling expression: „to enlargen the baileo anakota"), the villagers build this really sacral men's house in the woods and close it in with a „wood of rumbiapalm-leaves strung together" and moreover fence it in with some additional sereens, likewise plaited of leaves. Where two buildings are unknown and the house inside the settlement is used both for the more profane purposes (meetings, guesthouse etc.) and for the purely sacral (initiation), this house, on the occasion of the great, really religious feasts is artificially transformed into a wood." J) How exactly the situation has been in India, where the data concerning the older periods are not too numerous, is difficult to settle conclusively. At any rate the wood, in olden times, was regarded as the place, indicated for ceremonies with outspoken taboo-features. The vrata for a mystery-lore or those for the Vedarecital, had to be held outside the village.2) Whoever applied himself to self-tuition of the Veda, had to go outside the village, actually to the North-East of it, to a spot whence the roofs of the village-houses were invisible. An entire literary genre (aranyakas) reminds one of that particular significance of the wood. As to the sabha Ap. II. 25. 2 prescribes, that it must stand outside the village, indeed South of it. 3) This decermination of its place is of course important for our subject. To the South, outside the village, lies also, according to some, the smasana. One might therefore say that 1) Contr. Geog. and Ethnog. Mag. Neth. East Indies vol. 88 p. 410. *) Hillebrandt-Rituallit. p. 57. . . 3) Referring to Rassers' essay on the drama, which in the Island ot Java appeared to be founded on the dramatization of the myth, we may state here, that the man, at whose request a drama was performed, is called sabhapati. And further „The auditorium is marked off by pillars, in front a white pillar for the seats for the Brahmins, then a red pillar for the Ksatriyas, in the norht-west a yellow pillar marks the seats for the Vaifyas, while the (Jüdras have a blue-black pillar in the north-east" (Keith — The Sanskrit Drama p. 370, 359)- the circle: chaitya-burial ground-sabha, has almost come full turn, so that a connection between chaitya and sabha becomes likely. We do not wish to assert, however, that the chaitya should be quite identical with the sabha. That there exists a close connection between the two we think beyond dispute. But the chaitya, owing to its association with Buddhism, has gone through a process of development of its own, which we cannot examine further here. It might be said that the ancient sabha has been the prototype of the chaitya. As regards the sabha we shall leave undecided, whether Apastamba's information should be taken as proving that the sabha always had to stand South of the village, so outside the settlement, or whether this only applies to a special sabha for special purposes. The latter theory may perhaps seem the more plausible one, as the village meeting-place is usually to be sought in the centre of the Indian village, e.g. under the big assembly tree. That the sabha, together with the samiti is called the daughter of Prajapati, is, considering the great significance of both, quite conceivable. Some scholars (Zimmer, Ludwig, MassonOursel) distinguish the sabha-assembly from the samiti, a distinction which Keith thinks untenable. x) Masson-Oursel calls the ancient sabha a „conseil politique" or „une assemblée judiciaire", besides which the samiti would have to be taken in the sense of „conseil de guerre". Partly, however, the two are identical, for when the samiti has chosen the king, a priest further utters the wish, that the sabha may remain loyal. Now, in Ath. Veda XV. 8 the sabha is mentioned in the same breath with the samiti, the army (sena) and ardent spirits (sura). The whole XVth book of the Ath. Veda discusses the Vratyas. As the Vratyas are again closely associated with Rudra-Siva, we have here quite a complex whose interrelation does not seem clear at first. As ihis is, as far as we know, the only place where Rudra and the sabha are more clearly related with each other, wé must consider this complex more closely. It is to be regretted that our knowledge of the various parts of the complex is in many respects incomplete so that we must first try to determine each component part separately. The first and chief *) Cambr. Hist. I. p. 96. question to be considered is: Who and what are the Vratyas discussed here? *) According to the Petersburg Dictionary vratya is: „einer schweifenden Bande angehöriger Landstreicher; Mitglied einer Genossenschaft, welche ausserhalb der brahmanischen Ordnung steht". The word is derived from vrata-Schaar, Haufen, Trupp, Abtheilung, Gilde, Genossenschaft. As to the question as to the exact nature of the association of the Vratyas, scholars do not agree, hesitating between a group of people excommunicated from the Brahmin religion, and people belonging to a non-Aryan part of the population. At any rate it is certain that the Vratyas stood outside the orthodox Brahmin community, and could with the aid of a certain ritual, that of the well-known Vratyastomas, be received into the orthodox belief. In various places Vratyas are talked about, where the meaning remains rather obscure. In the P.V.Br, is told that the divine Vratyas (divya or daivya vratyah) led a Vratya-life and performed a ritual of 61 days duration, which, however, for some reason or other did not turn out well. The most elaborate on this subject is T.M.Br. XVII. 1. 1 sqq. There the gods are said to have gone to heaven, but the daivas to have remained behind, so that they became Vratyas. Thus those remain behind who perform no brahmacharya, do not occupy themselves with agriculture (krishi) or commerce (vanijya). 1. Deva vai svargam lokamayamstesham daiva ahïyanta vratyam pravasantah. 2. Hïna va ete hlyante ye vratyam pravasanti na hi brahmacharyam charanti na krishim na vanijyam shodaso va etat stomah samaptum arhati. The Vratyas, therefore, are people that „do not adhere to che Vaidik rules of brahmacharyya which principally bears to the idea of prosecuting studies and residing in the house of their teachers, and begging of alms and following other rules prescribed by the Vedas for the students. 2) Put briefly: Heaven is closed to the Vratyas according to this Brahmana; they remain behind, because they do not perform initiation in accordance with Brahminical conception, and pursue neither agriculture nor commerce. 1) The writings of Hauer on the Vratyas wer> not at our disposal. 2) Braja Lal Mukherjee—The Vratyas and their sacnfices — J.K.A.b. Bengal 1925 p. 154* After this Vratyastoma, which is said to be a Marutstoma, has been further expatiated upon in 3-8 of the same Br., 9 contains the additional information: Garagiro va ete ye brahmadyam janyam annam adanti. The Vratyas, the Brahmana says here, are people who consume food which is suitable for Brahmins, but poison for them. Braja Lal Mukherjee suggests here the sura, the ardent spirits, which, when opposed to the amrita, is poison, but after a consecration becomes suitable for consumption. Aduruktavakyam duruktam ahur. We agree with Mukherjee that this cannot refer to a bad pronunciation of Sanskrit words, as Albrecht Weber and Jarl Charpentier have explained. Ahur is not „pronounce". Mukherjee suggests obscene words, which the Vratya does not want to employ, because he was a heretic. We would suggest a more spacious and more general explanation: Terms that are not taboo (durukta-dangerous to say), they do call so. The sense of this translation will become clear, when first a few other matters have been discussed. Adandyam dandena ghnantas charanti. Mukherjee suggests the chastisement of a Brahmin, who is adandya by direction. The Brahmana therefore states a certain hostility between Brahmins and Vratyas. Adïkshita dïkshitavacham vadanti: Without being consecrated they pronounce the consecration-formulae. So chey know ritual formulae, but these are not acknowledged as such by the Brahmins. There are altogether four different Vratyastomas: 1. The stoma for those Vratyas who instruct other Vratyas in the use of arms, in dancing, singing and games. 2. One for those who are nindita nrisamsah („Beschwörer die in üblem Rufe stehen" — Charpentier)*). 3. Those for the youngest Vratyas (kanishthah). 4. Those for the oldest Vratyas (jyeshthah). Mukherjee after all sees the Vratyas as people „originally belonging to the Vaidik community but becoming alienated and neglecting or defying Vaidik precepts and creating disorder." Now the Vratya is at any rate connected with Rudra. As Bhava, the archer, the latter is the patron of the Vratyas. Rudra is the patron of the Vratyas in all quarters. This close relation of Rudra with the Vratyas has, in the opinion of many author- ') Bemerkungen über die Vratyas — W. Z. K. M. G. XXV p. 363. ities, been expounded in the enigmatic XVth Book of the Ath. Veda, where a Vratya is dealt with, who travels in alle directions of the wind. Various scholars (e.g. Macdonell, Oldenberg, Güntert, Arbman) agree, that this Vratya is nobody but Rudra, venerated here as the heavenly Vratya. On the ground of all this information concerning the Vratyas Professor Charpentier came to the conclusion, that the Vratyas were worshippers of Rudra-Siva, who occupies a position somewhat outside stnctly orthodox religion. This theory has called forth some senous criticism. It must be conceded that the evidence in support of Charpentier's argument is rather thin. x) Although Rudra had of old been characterised by a certain ambiguity, yet it will not do to place him outside the orthodox religion. That the Vratyas should originally have been members of despised and impure castes Charpentier justly stigmatizes as Unsinn". On the contrary people of great distinction could belong to the Vratyas, as Manu X. 28 counts the well-known ruling families of the Lichchhavis and of the Mallas among them. Of particular importance is the fact that the Kurus too were Vratyas.2) Manu calls the Vratyas simply savitripatitas, i.e. people who do not observe the initiation, at least not in the Br'ahmin sense. 3) For the rest the Vratya seems to have had a deep respect for initiation (a non-Brahmin vanety, presumably), as Maurice Bloomfield remarks that the Vratya seems to have been a kind of Brahmacharin. 4) The Vratya is therefore a member of a society which stood more or less outside Brahmin religion, having both a ritual and initiation of lts own, which, however, were not acknowledged as such by the Brahmins. The heavenly Vratya is Rudra, who lives in contact with the earthly Vratya. On this heavenly Vratya, who represents totality (he fills all quarters), follow, according to Ath. Veda XV. 7, the sabha, the samiti, the army and the sura. The initiated (ya evam veda) becomes the persona grata of these four. Combining all these data we think ourselves justified to draw the conclusion: the Vratyas formed the sabha-society, which had grouped itself round Rudra and enjoyed a doubtfu 1) Keith — J.R.A.S. 1913 P- !55. 3) Thepoem with°which study of the Veda opens, is called Savitri. 4) The Atarvaveda p. 94* reputation among Brahmins. In this modified form Professor Charpentier's view appears acceptable to us. Now the existence of societies may be expected in a social structure, whose clan-organization is falling into decay, In our chapter on social organization we have expounded, how, owing to its steadily expanding status indication, class-exogamy is bound at last to decay, and caste with its endogamy comes into existence. Just as the whole tribal organization does not disappear all at once, when caste springs into existence (it is indeed seen to survive in caste), the entire tribal cult does not immediately disappear. It is the very societies which continue the tribal cult. They, indeed, carry on the ancient clan-myths and rituals. It is impossible to define this society-ritual better than by calling it tribal ritual, which only lacks the clan. As early as 1904 H. Schurtz recognized the sabha as a society. At the same time he draws attention to the remarkable etymology of the word sabha, which, according to Zimmer, Schrader and Uhlenbeck is related to Gothic sibja, German Sippe. So the connection between society and clan is still apparent from the etymology of the name.1) The period of the Brahmanas was, in our opinion, the period in which clan-organization was decaying and castes coming into existence. With regard to this it is quite possible that the society existed at the same time, for in the very Brahmanas the Vratyas are most frequently discussed. In a stage of civilization as that now under discussion, the initiation can no longer be an affair of the whole tribe, because together with the clan-organization the tribal cult in its ancient form is bound to change. Initiation then becomes a prerogative, a honorific distinction to be attained by means of money and high social position, which gives the initiated access to more or less exclusive, often secret societies. This phenomenon, we should think, is sufficiently indicated in India by the symptom, that initiation was a privilege there, denied to the Südra. The J) Altersklassen und Mannerbünde p. 282 ssq. The meaning of the word sabha shows the same modulations as the word club. It means both the building where the club meets, and the club itself. See the work of Schurtz on the Men's house, as gambling house, council house, arsenal, clubhouse, fortress, place for the cult of the dead, residence for travellers, workshop for certain trades (blacksmith, occasionally weaver). 16 latter could not be a twice-born (dvija) and consequently did not undergo the sacral initiation-death. We agree with Max Müller in thinking it important, that in the Grihya Sütras the Südra is never expressly denied the right to initiation. It follows from our argument that we think it probable that there has been a time, when the Südra could also be initiated. This must have been the time when initiation was still the sacrament to which every grown-up man was entitled and not a distinction of caste. That writing about the initiation of a Südra was simply not thought of, if merely to forbid it, as Dr. Ghurye adduces against Professor Müller, seems improbable. For does not the very term ekajati denote that the Südra was not entitled to initiation? This was the chief characteristic of a Südra. That the Sütras do not even lay down this, is truly remarkable.1) * * * In the rainy season, the period of the initiation-demon Rudra, the sabha-society of the Vratyas held its ritual assembly. Why the Brahmins were so little pleased with this society will be discussed in the next chapter. Several questions that still await discussion, and which may at the same time afford further evidence in support of the summarizing conclusion given above, will then be more closely studied. l) Ghurye-Caste and Race in India p. 56. V GAMBLING In the preceding chapters we have been discussing the time when, and the place where the great tribal feasts were celebrated. Moreover it appeared that the persons who, in those places and at those times celebrated their ritual feasts, were members of some club or society. The question which arises is: — What were these societies practically doing there? What was the ritual they were performing there? Clubs of this kind spring from a clan-organization. They are especially found in communities in which the old clan-organization is decaying, so that the clan is losing its character of an exogamous group. Circumstances are favourable to the arising of castes, a new arrangement in which the status-indication which is never quite lacking in clan-organization either, is systematized. In the chapter on social organization we expounded that the status-indicati°n then gives rise to an entire scale of various ranks. It stands to reason that now the clan can no longer preserve its religious character. More and more clan-indications and clan-dignities are being concentrated on individuals belonging to the higher classes of society. For the clan-organization in so far as it is based on exogamy, falls indeed into decay, but it does not necessarily follow that everything relating to the clan, vanishes as ïf by magie. Where totemism exists, the initiation can no longer be a ceremony, through which one is identified with one's totem, simply because the group which owned the totem-ritual as such, no longer exists. The initiation, the right to the totemname assumes the nature of an introduction into the club possessing the totem. Especially candidates of standing are introduced, novices who have inherited the totem, but also those who possess sufficiënt means to reconcile club-exclusivism to their presence. In the club itself a hierarchy arises at the top of which the head of the society takes his place, the highest bearer of club-dignity. In the sabha-society one should not directly expect to see a clan practically, but probably a group of men sprung from one or more of the chief clans. Where we come into contact with societies of this kind we also find, entirely in keeping with the development depicted here, an all-pervading craving for distinction and priority. Each group — whether it be a group of aristocratie people from a certain part of the country, or an aristocratie family — does its utmost to maintain its position or to rise. This attempt to put the achievements of one's own club in the comparative degree with regard to those of other clubs, and if possible in the superlative degree with regard to all clubs of the community put together, is the most striking phenomenon about these societies. Rivalry is now no longer curbed so much by the mutual rendering of services which the old clans owed to each other. Marriage-relations have been shifted. Other criteria are showing. The process of specialization by status-indications, which is also never lacking in a clan-organization, is becoming more and more intensive in higher circles. This phenomenon is called „potlatch". The clans changed into societies, get the best chance of showing their high position and their great wealth at ceremonial meetings, e.g. at the initiation. But besides these there are many more opportunities at which the club can demonstrate its wealth. We refer not only to all those cases in which the club as a whole enters the field against other clubs in all sorts of tourneys, but also when representative members of a society accomplish something, e.g. when a prominent man in the society builds a house. When the society organizes some ceremony, it is sometimes impossible to determine whether it is aiming at performing this ritual or whether the centre of gravity has been shifted toanostentatiousdisplayforwhichtheperform- ance of the ritual offers an opportunity. Potlatch, festivals, rivalry, these are the three sides of the same thing. A characteristic feature of the potlatch is always the custom of giving feasts which the organizer cannot afford. Now one might betempted to say that potlatch in that case is identical with what is observed in the exclusive clubs in our modern community; with what is rather frequently observed in upstarts who often try their utmost to put their coats of arms, the paint of which is not quite dry yet, in a gilt frame. This is an error. For potlatch is not merely an opportunity for snobs to vent their snobbism, but bears a pronounced sacred character. Potlatch is directly connected with the old clan-ritual. The society indeed possesses original myths just as well as the clan; it possesses masks, dances, and sacred hymns. The tendency to rival, which is not foreign to clan-organization either, (one need only remember the well-known phratry-rivalry) has come more and more to the fore. If anyone wants togive a potlatch partyhe makes the necessary preparations. Just before the guests may be expected to arrivé large quantities of provisions hare been stored. Trees around the grounds where the fête is to be given are hung with choice victuals. Round this tree which is a wish-tree, in the literal sense of the word, (only think of the Indian kalpavriksha) food is heaped up. All bespeaks superabundance and wealth, and rank, due to the mercyof God. Meanwhile special messengers have been dispatched to invite the guests who are deemed worthy to come to the feast, and to compare their own wealth with that of their hosts, at the date fixed. For, as we have said before, the potlatch is a rivalry. It may be that the receiving party had to perform a ceremonial itself, and wishes on this occasion to give its guests a splendid reception. In this case the capital which the receiving party loses is not really lost, owing to the fact that it gives credit, so to say. He who, on occasion, cannot give an equally splendid reception to his hosts, is sure to sink in the estimation of his fellow-beings,and consequentlyin importance. The painful consequences of this will not fail to come, for the number of people who are willing to bestow gifts and tribute on him will decrease accordingly. Once it has come to this that one is in danger of no longer being considered one of the most distinguished, there is no other means to regain one's lost dignity than a splendid potlatch. Thus this institute becomes a means of repairing loss of standing and consideration both in a material and in a moral sense. Thus the potlatch is a means to obtain credit, and as moreover an important shifting of wealth forms part of it, the study of the potlatch touches economy. Hence it is that in the introduction we said that we would study the social, the religious, and the economical aspect of Indian culture. But as it is not easy to associate modern economy with primitive cultures, our views may also be considered to regard religion that is to say its ritual element. The potlatch may also possess the character of an „exchange-duel". In one potlatch it is decided who may practically be called most prominent and richest. This may be effected by means of the sale of ritual property of high value, or by simply comparing a certain kind of property with another, or by making the gods speak through some oracle. In the presence of the entire assembled crowd the names of the parties and the nature of the transaction are publicly proclaimed. The contracting parties express their approval of the proclamation either by showing their blazon, or by giving a performance of the myths to which they are entitled. In this way the potlatch may become a regular competition, an element which, more or less seperately, is always to be found in the matches attending it. „Especially owing to the excellent ethnographical research by Boas, Malinowski and others, and the profound theoretical studies of Davy and Mauss, we know now that the ceremonial barter performs an important social function, especially where, as one of the systematized utterances of the relations of rivalry and reciprocity between certain groups, it has not only created manners in social intercourse, in which the structure of the later economical intercourse lies contained, but forms at the same time one of the most important links between religious ritual and the social feast, consequently between the sacred and the profane walk of life. The kolekole-feasts in Melanesia, the kula-system in the Trobiands, the potlatch in Northwestern America — to mention some only — are all, as Mauss has convincingly shown, to be classed under the same complex of phenomena. All these feasts bear the character of secret or open warfare, not for property or possessions, but for social influence, prestige; sometimes between individuals, mostly between groups or both simultaneously. The weapon is the gift, which commits to an at least equivalant counter-gift, and redounds to the giver's credit. With feasts of the potlatch type the competition in liberality has demoralized to boundless extra- vagance; destruction of goods on a large scale had replaced the competition of making presents. Considered in this light these feasts present themselves as purely social phenomena. When ,however, we consider that these antagonistic groups, whose peculiar relation also manifests itself in ceremonial matches or even in fights, at the same time stand for certain parts of the cosmos, and that consequently their strife demonstrates the antagonism of primeval cosmic forces, we recognize in these feasts not only the ceremonial sediment of social rivalry, but also a religious rite in the shape of a cosmologic drama".1) Thus the potlatch may become a regular strife, an element which, more or less apert, is yet always to be found in the attending matches. The other aspect shown by the potlatch, is that of a sacrificial ritual. The potlatch may indeed mean not merely a lavish distribution of wealth, but also destruction of it. The headman of the society simply destroys wealth, or breaks to pieces ritual objects of great value. Also in this case the element of strife has not disappeared, of the rivalling party it is expected that he shall evince a no less boasting sense of sacrifice. That it here concerns a real sacrifice, a ritual, appears not only from the sacral sphere of the whole, but also from the fact that among the Kwakiutl, the object which is sacrificed is the sacred piece of copper, often adorned with clan-symbols. The symbol of the clan, which is the keystone of the entire classification, contains a great deal of mana. If, moreover, the clan-symbols of the piece of copper adorn the owner's Standard, we may justly conclude that „as long as the idea of the unity of totem and totemowner was still living, the copper piece was thought to be a part of the being of its possessor".2) In his remarkable studyon the serpent, G. W. Locher has also proved that the serpent, which was the possessor of wealth, is the central animal of the potlatch. There is a close relation between copper piece and serpent. The serpent, the bisexual being of the underworld, occupies a very important place in these societies. The serpent not only appeared to be the central animal of strife, but also of the potlatch. Sacral property, strife, serpent, underworld and societies were the various elements which appeared to belong together. n t' PÜ B- de i°sseLin de Jon8 — The divine Trickster p. 26. ) Locher — The Serpent in Kwakiutl Religion. p. 82. descend from the clouds which are plainly visible over his head. We may perhaps assume that the pose of the Chhargaon Naga has the same significance" 1). This supposition would correspond entirely with our conception. The chakravartin (preserver of cosmic ritual) causes a shower of square coins to descend. In the square coin we recognize the Karshapana (the aksha, which is very probably the potlatch-emblem). Parallel with the King stands the serpent, the potlatchdeity, which causes potlatch-emblems to shower down. In this connection the hypothesis of Professor Vogel becomes very plausible. Behind the name „pasaka" one ought perhaps to look for the idea of serpent. Pasaka2) means „rope", a name which on the face of it, does not look very plausible for a die. Yet serpents are not unfrequently given similar names. In the A. V. the serpent is called datvatï rajjuh (the toothed rope), pütirajju (the putrid rope) or „the biting rope, the rotten rope" 3). To this very day the serpent is given the name of „maternal uncle", or „the rope 4). Apparently this name finds its origin in vernacular taboo. As we have stated above, the serpent functioned in the ritual of the Amritamanthana as the rope with which the churnplunger was pulled. B. Renz will probably be right in taking the ropes with which the linga in the Indradhvaja is set in motion, for snakes. It is of course not easy to state exactly what all the characters on the Karshapana stand for. We should like, however, to draw attention once more to the little dominofigures mentioned above, as further evidence of the view, laid before our readers. A mark like this is called a „rüpa" 5). We have given this word as a term and expression of the relation between the things within the classification-system. ^Ve would here connect the more theological-philosophical signification which this word appeared to possess in the Brahmanas, with the meaning: ') Vogel-Indian Serpent Lore p. 42, 282. 2) That pasaka should be derived from pra-as is an altogether unfounded supposition (Lüders — op. cit. p. 16). Weber (Ind. Streifen I. p. 274) thinks of the „bestrickende Gewalt des Spieles". Culin also considers the pasakas substitutes of the arrow. 3) Vogel — op. cit. p. 12. ') Crooke — Rel. and Folkl. of North. — India p. 392. 6) Bhandarkar — op. cit. p. 68. Rüpaka is the name of a certain coin. „representation of the animal or object with which one is specially related, emblem, blazon", just as a totem mark is for the member of the totemistic clan. These emblems are not lacking in India either. The prominent warriors take their dhvajas with them into the battle, provided e. g. with these emblems: — a palm, a jewelled naga, a buil, a monkey, a boar, a bird. The standards are said to give off smoke, and to tremble again and again. When the dhvaja toppled over, great confusion was the result, so that the dhvaja was often the first object of attack. The dhvaja brought luck or disaster (amangalyadhvaja)1). When Sanjaya in the MBh. has to describe the dhvajas, he does so in accordance with their rüpa,their naman,and their varna. (VII. 105.2) Rüpa and naman have already been described as the great criteria for classification. These rüpas naturally remind us of the emblem. We should like to draw attention to two words. In the first place to the terms visvarüpa or bahurüpa, which are often given to gods. These terms are generally considered to express the power to assume any shape. In our opinion this translation gives only a part of the real meaning.With the indication visva-or bahurüpa one wants to express that a god is entitled to bear emblems of all clans, and that consequently the things classified with all those clans can, in a mythical sense, be substituted for him. Visvarüpa is consequently a distinction for the gods, emprising the entire system of classification, or for whom one wants to pay homage to in this manner. For though one may be prepared to believe a god to be endowed with the power of assuming all shapes, and therefore being called, „visvarüpa", it is difficult to imagine, reading of Rudra's having a „visvarüpa nishka", that the same power may be attributed to a necklace (nishka)2). The translation „multiform or multicoloured neck- " M Hopkins - J. A. O. S. XIII. p. 243 ssq. Havell says that these dhvajastambhas or „tribal ensigns" were, during the A^amedhaplaced in the ground outside the entrance to the place of sacrdice. A cunou5 dhvaja is e. g. depicted in Havell-Anc. and Mediev. Arch. of India PI. XII. In the name of the hero of the Pandjinovel, Rassers professes to see the memory of the emblem, in which the totemistic ancestor of the tribe is presenting himself mystically, and thinks of the parallel with the Australian bullroarer, „liturgical Instruments, considered as a demonstration of the totemistic ancestor . Pandjinovel p. 341 SS<1- The word nishka with lts doublé meaning of „coin and „necmce strengthens us in the opinion, that ancient Indian money was what consacré", as de la Vallée-Poussin calls him, in histurn receives the power to irtvest the Brahmin with Kshatriya-power (rajno niyogad yoddhavyam brahmanena viseshatah) ^.This phenomenon is the background of the struggle between the Kshatriyas and the Brahmins, reflected by the Epic in the story of the king's daughter Sarmishtha and the Brahmin-girl Devayanï. ,,You are", the king's daughter says, „the daughter of a Brahmin, whose work it is to praise and to accept gifts (stuvatah pratigrihnatah). I am the daughter of the man who is praised and does not accept (stüyamanasya dadato 'pratigrihnatah)". The king is more conservative than his daughter. He is afraid of the threats of the Brahmin and confesses that all wealth of his race is practically the property of Devayanï's father2). This story of girlish vanity contains quite a piece of cultural history. It is all very well for the Brahmin to admonish his daughter that forgivingness is a beautiful virtue, and that one ought not to believe all a king's daughter says, but after all it is a hard truth, for the Brahmin's daughter surely knew that, „blessed is he who gives", had a pregnant meaning in India. It is obvious that the Brahmins as a caste cannot but have regarded the great prosperity of the potlatch with ever-increasing suspicion. It cannot have been pleasing to them to see clubs of nobles make this one, but therefore not less important, side of the great tribal feast, the potlatch, an object in itself, a ritual in which the prominence of the Brahmins hardly took part, unless as donees, as mannequins at a pageant. Wealth was power, a proof of the blessing of God, an exponent of the exercise of dharma. The Brahmins wore themselves out in exhorting the kings that a better work was not to be found, than largesse to the Brahmins, but at the same time, in doing this they subjected themselves implicitly to the kings. „Receiving means: „undertaking the obligation to return more than one received". He who receives, is bound. Acting in the fight proper of the potlatch is, according to the dharma, forbidden to the Brahmin. He cannot give his property away, for giving away means accepting strife. „La propriété du brahmane, c est le brahmane lui-même" 3). ') MBhflX. 45, 41. 2) MBh. I. 78, ssq. s) Mauss — loc. cit. p. 147- It is therefore not surprising that the Brahmin more and more turned away from dicing, and that he scornfully calls the rainy season, the time of the potlatch and of the initiation, mobtime. How far this went we cannot ascertain. Be it sufficiënt ,however, that we have explained how it is that the potlatch came to be condemned so severely later on x). A certain contrast arises between the period in which the societies perform their ritual, and an other time chosen by the Brahmins for religious ceremonies 2). The sabha, the club-house, was not a place where the Brahmin was fond of staying. Doubtlessly, when he visited it, his voice was listened to, but above all the sabha was always the club of the Kshatriyas. There are perhaps a few more facts which enable us to tracé this development as it proceeds. For instance the lighting of the sacred fire in the sabha, a rite, which, according to various texts, was only performed by the Kshatriyas, and which, according to Hillebrandt, was „eine Prarogative adliger Geschlechter". „Nur von einigen Lehrern wird sie gefordert, von andern, wie Salïki verboten, von manchen wieder ins Belieben gestellt. Charakteristisch für die Feier der Sabhya-Anlegung ist das Würfelspiel. (In a note follows:) Die Frage, wer spielten, Ksatriyas oder Priester, war controvers" 3). Weber says about this game of dice: „Der Schol. Mahadeva will nicht, dasz die Priester selbst spielten, offenbar ist dies aber ein alter, heiliger Brauch, durch welchen die sabha zu ihrer künftigen Bestimmung geweiht wird" 4). In our opinion, we see the Brahmin standing here on the threshold of the sabha, which he will ') In our essay we have more than once strongly stressed the importance of the rainy season, through which the impression may have been made, that we deny all sacral significance to the rest of the year. This ,however, is not the case. It is most remarkable that Hillebrandt should have defended the thesis that the RigVeda served for the Devayana, and the Atharva Veda for the Pitriyana. It remains to be seen whether in this way „der kalendarische Gesichtspunkt" is brought into prominence too much. Oldenberg — R. d. V. p. II nt. 1. ) The same division is found among the Red Indians, whose cult alsoin other respects, bears typical points of resemblance with that of India. Among the Indians the organization in societies preponderates in winter, the clan-organization in summer. See further Davy-La Foi Jurée p. 349 ssq. — The Pachisi-game so closely resembles a RedIndian lotgame, that E. B. Tylor wrote an essay: „On American Lotgames, as evidence of Asiatic Intercourse before the time of ColumbusInt. Arch. f. Ethn* IX Suppl. p* 550 ssq» 3) Hillebrandt — V. M. II. p. 20. 4) Ind. Stud. X. p. 327 nt. 5. 19 allow a single subordinate tone to predominate in the picture he was painting, would be to spoil the effect of the entire work. It is, moreover, most improbable that the Pandavas were the successors of the Kauravas. The sons of Pandu can hardly have feit themselves flattered at being represented in the Epic as the deceitful brothers of the conquered Kauravas. We must also bear in mind that the predecessors of the Pandavas, the Panchalas, were not the foes but the friends of the Kauravas. There is nothing in the Epic to point to anything resembling a definite succession of a number of different royal houses which, in the course of years, conquered one another and usurped the supreme authority. We, therefore, think it wiser to assume another motive in the mind of the poet for representing the contending parties as brothers. There is no reason to doubt that in his narrative of the warring brothers the poet depicted the everyday reality round about him. He does so in such an accurate and systematic manner, right through the entire poem, that anything in the way of an artificial straining after effect must be considered out of the question. The relation between the two parties is undoubtedly pictured by the poet as a phratry-relation. It is not difficult to marshal the evidence necessary to the confirmation of our assertion. The Kauravas are the elder, the Pandavas the younger brothers. The Asuras are on the side of the Kauravas, the Devas on the side of the Pandavas. The Kauravas lose, the Pandavas gain in power. During the actual battle the Kauravas are drawn up in the East, their faces turned to the West; so they fought with their faces backward (paschanmukha) like the Daityas, the Pandavas fighting with their faces front (pranmukha) like the Devas. „Siva is the Kurumaker, lives with the Kurus." Krishna-Vishnu belongs to the Pandavas. x) There is no mention either of sects or of a succession of royal dynasties, but simply of two separate groups related to each other as are the twin phratries of a tribe. Both groups together constitute a complete unity, seemingly divided but forming in effect a totality, the oneness of the tribe. How shall we otherwise explain Duryodhana s sacrificing to Vishnu and Krishna's regularly making nocturnal sacrifice 1) Hopkins — Ep. Myth. p. 220. together with Arjuna to Rudra Tryambaka and their worshipping Éiva's consort, Durga. Although we are of the opinion that there is but little fear of one's being mistaken on this point, there is also, of course, a possibility of historical motives underlying the phratryrelationship in question. There is always the possibility of an actual battle underlying the idea of phratry-rivalry, a battle between two peoples or between an invading force and the native population fought upon the field of Kurukshetra. But the MBh. as it is now known to us does not testify to the actuality of such a battle. May be other historical motives have incidentally found their way into the poem. We are not at present justified, however, in advancing any such theory, although perhaps the future will shed more light upon the subject. Each of the two groups by itself constitutes the whole. lt is this idea of totality which it will behove us to see in the number „five of the Pandavas, in which Ludwig and Dahlmann rightly saw what they themselves called a „mythical" number. Dahlmann understood it as being expressive of the undivided and indivisible unity of the family, which it indeed is. The number „five is in turn subdivided into the number „three" of the sons of Kuntï and the number „two" of the twins of Madrï. These numbers likewise strike us as being remarkable just as the number „hundred" of the Kauravas. In the meantime we shall do well to bear in mind that he who would prove too much will prove nothing. There is also the name of the sons of Pandu (pandu — white). A phratry-contrast denoted by colours is of very common occurrence. An indication by colour has also been seen in the name of the Kurus, the word according to Ludwig signifying red,x) red and white then being opposed to one another. That the appellation, Pandava, more than any other, should have first made its appearance at a later time cannot, in our opinion, be adduced as evidence of a later dating of the possible rule of the Pandavas. We see in the name nothing but the fixation of a phratry-indication by colour. That the appellation later on acquired the fixity of a proper name does not, of course, prove that the phratry-relation from *) Abh. Kön. Böhm. Ges. Wiss. Cl. f. Phil. 1895/IX p. 4. Ludwig mentions kuruvaka, kuruvinda, kuruvilva, kurumba etc. Pandus and the Kurus themselves — the plot of the whole play. The question to the Hindus is exceedingly complicated. Dhritarashtra, the natural heir in the first generation, was blind and excluded from the throne; his younger brother, Pandu, reigned, but resigned the throne to become a hermit. This obliged Dhritarashtra to rule (the other possible heirs we can overlook in this connection). The sons of each grow up together. Dhritarashtra installs his nephew as heir-apparent, either from a sense of right or through fear of the people (I. 139. 1—2). And what happens? His own sons conspire to get the throne. The king changes his mind in regard to the heir. Then the people murmured against the king (Dhritarashtra), and said that his son was not so brave or so good as the nephew; and, therefore, they would have the nephew for king. They feared a plot, and became even bolder. „They met in courtyards and on the streets and in assemblies", and demanded that the king should be dethroned, and his nephew not only installed as future king but made king at once. „For how," they asked, „can this Dhritarashtra now have a right to the throne, when he failed to obtain it before on account of his blindness?" (I. 141. 23 ff.)- That these plans would have succeeded is to be inferred from the fact that the king seduced his nephews to a distant town and tried to burn them. The people, belie/ing in their death, sank into passive acquiescence to the yoke. But again, far later in the drama, when this nephew has become king of a neighboring city, and returns to his paternal home, and is cheated out of the kingdom he has made, and banished into the woods by the wiles of Dhritarashtra s son, even then the people cry out as before against the Kurus, and the priests become enraged and refuse to perform their duties. In the earlier scene the king is supported in his schemes by his sons, and the ministers give no sign of disapproval. But for the people, he would not have found it necessary to send his nephew away, before putting his son upon the throne. The question of legal right is but lightly touched upon. The people know that it is a disputed point which prince has a technical right to the throne. They insist, however, upon the moral superiority of their choice, and their right to chose. When the technical point comes up for discussion, Dhritarashtra himself acknowledges that the nephew ought to have had the throne, and throws the whole blame upon his son. His queen also says: „It is the custom in our family that the crown shall descend from father to son;" and, therefore, argues that the son of the actually reigning monarch had the rignt tosucceed."1) It is, of course, from the very nature of the case impossible to adduce anything like a mathem^tical proof for a thesis of this kind, but we may be permitted in this connection to draw attention to the curious conclusion of the MBh. After it has been demonstrated in full detail by the vast array of facts that it was dharma (the Pandavas) that prevented the ultimate victory of adharma (the Kauravas), the reader is dumbfounded by the Epic's describing at the end how the Kauravas revel in the ineffable joys of heaven, whereas the Pandavas are doomed to endure the terrible tortures of heil. It is easy to understand as a perfectly human reaction that the dharmaraja Yudhishthira, is not particularly pleased with this token of appreciation of a virtuous life. Later on his disappointment is tempered by his being informed that he is only doomed to contemplate the infernal horrors for a certain period of time, at the expiration of wbich the „retributive" tables will be turned. If it be not permissible to explain this curious passage by the tendentious design of the poet arising from his „partidity for both parties alike , we shall have to conclude that this appalling prospect for the devotees of dharma can hardly have had a very stimulating influence. It is this unstable partiality, centred upon a doublé object which in numbers of instances urges him to ascribe a victory to actions and dealings of a deceitful character. Bhïshma, Drona, Karna and Duryodhana, the commanders of the Kuru host — they are all levelled to the dust, but by unfair tricks. It is in the figure of the double-dealing Krishna, that this deceit is hypostasized, personalized, as it were. The reader will now understand the end and purpose of our argument. The Epic, the Sabha Parvan especially, is the story of a vast potlatch. In connection with the question which now has our interest, it is exceedingly remarkable that the word, kitava means both „gambler" and „deceiver", and kaitava both „stakes" and „deceit". The entire gambling scene ») Hopkins — J.A.O.S. XIII p. 142. is founded according to the representation of things in the Epic upon deceit and Sakuni hardly hazards a single throw without its being told that he does so in reliance upon nikriti (nikritim samupasritah). This is certainly not a common instance of fraudulent piay in a game of hazard; it is very much more than that. There were other people following the game besides the Pandavas. „Aber beim Spiele sehen Bhïshma, Drona, Vidura und viele Könige zu. (II. 95- 1 )* Kann da £akuni unbemerkt fortwahrend betrügen?" The question thus put by Holtzmann clamours for a negative answer.This element of deceit has apparently cost later epic poets, too, a great deal of trouble. The Dharmaraja admonishes Sakuni most urgently not to play false. Sakuni then urges the defence that it may likewise be termed deceit when a learned person tries conclusions with an unlearned person, a skilied warrior with an unskilled one, a strong man with a weaker, if a game of dice between a practised and an unpractised player is to be given the name of deceit. But however that may be, one thing is certain: the reader is continually being told, world without end, that the Pandavas lost the game and thereby all their worldly possessions by sheer trickery.2) . Now let us parallel the instances: in the battle it is from the side of the Pandavas that deceit comes; in the dyüta the Kauravas are the deceivers. The purpose of the Pandavas practising deceit is to show that the Kauravas were really invincible in the yuddha. The deceit of the Kauravas — it follows immediately — is intended to show that the Pandavas were really invincible in the dyüta. This is a weighty conclusion n Holtzmann was again thinking here of an inversion. It is supposed that in the original Epic the gambling match was staged by the Patidavas and that the Rajasüya sacrifice was not performed by Yudhishthira but by2)DWhatdthe poet is thinking of when he employs the word nikriti can only be conjectured. Nikriti is also famihar as being one of the Vasus, identified with Nirriti (destruction), hkewise personalized as one of the Rudras. It is likely that in dice-playing nikriti is to be under stood as an attempt at making it impossible for the opponent to execute the krita-cast (to throw krita), the dicer playing with nikriti, therefore, not playing merely to win himself, but rather to prevent his oppone from winning. In this manner he would be making a magical P/adice of the eambling-rite. There is no doubt that magie can be performed bySLSfrfffi. This is evident from the eyl-averting formula^gain the power of magie (kritya) quoted above (p. 271)- "he reader must take this as a mere conjecture however. to arrivé at, for it is now established a priori that this epical game of dice had nothing at all to do with an innocent game of hazard; it was a veritable tournament, the stakes being the prestige of the gambling antagonists. The poet wishes to say that the Pandavas, it is true, were the losers, but — that actually they could never lose. The Kauravas were the winners, indeed, but — by means of deceit. If our representation of the facts be correct, it must follow that the game of dice, at the time of the Epic's assuming its literary shape, had not yet lost its ancient significance of a giving-contest, in other words, of a potlatch ritual. The stakes, just as in the battle, were the prestige of the contending parties, otherwise the poet would not have introduced the element of deceit into his work in such a systematic manner. The poet has still more strings to his bow than the employment of this element of deceit alone. As we have already seen he is fond of calling into being dubious cases of a technicojuridical nature as a result of which it becomes extremely difficult to establish the superiority of either of the two groups, as e.g. proved definitely to be the case on the occasion of the two houses laying claim to the right of rule. Now he does the same thing in the dyüta; here, too, he creates the selfsame atmosphere of uncertainty. The man who has forfeited all his worldly goods, all his rights, even the right to his own person and that of his wife, such a one is definitely lost in the dyüta. Now the situation in the epical game of hazard is as follows: Yudhishthira gambles first himself away and then his consort, Draupadi. Now are the Pandavas done for after the loss of this stake? No, says one, for Yudhishthira had already gambled himself away when he staked and lost his wife, Draupadi. After having played himself away, Yudhishthira was no longer in a position to assert his rights. On the one hand: according to the dharma, Yudhishthira could not gamble away that which he had no right to i.e. his wife, in the present instance; on the other hand, according to the dharma, a man always has a right to his wife. It is a problem that cannot be solved, as Bhïma says to Draupadi, by reason of the subtility of the dharma. *) •) II. 67. 46. 20 Na dharmasaukshmat subhage vivektum Saknomi te prasnam imam yathavat Asvamyasaktah panitum parasvam Striyascha bhartur vasatam samikshya. In this manner the poet succeeds in making the question as to whether the Pandavas are definitively the losers in the dyüta a dubious one. Things are not different in the world of the gods to what they are in the world of men. The Asuras, the elder brothers of the Devas, were deprived of the Amrita in the Amntamanthana by fraud.1) We can hardly expect anything else. Hopkins, who can scarcely be accused of bias in this respect, writes: „It is noticeable that the death of the great Asuras is effected by deceit." One of the Asuras even bears the curious name of „Kapata" (deceit).2) The poet even hesitates to hazard a definite statement to the effect that the Asuras did not partake at all in the drinking of the Amrita, thus be coming mere mortals for ever. Rahu's head, at any rate, enjoyed the life-preserving nectar, thereby winning immortality. It may, therefore, be maintained that the Asuras also partook of the immortalizing elixir. The Devas, on the other hand, in the Indian mind, were also not thought of as being immortal from of old. Deceit is also the determining factor in the conflict between Kadru, the Mother of Serpents, and Vinata, the mother of the bird, Garuda. This element of deceit does not, therefore, place us before a psychological or historical problem. It occurs whenever the poet feels called upon, with a certain reservation, to credit either of two groups, related to one another as are two phratries respectively, with enjoying a position of superiority m respect of the other. , . And now the question arises whether a phratry-relationship such as that existing between the two contending parties of the Epic may be considered to fit in with the social structure of ancient India during the epical period of its history. The indications we have collected as pointing to such a possibility are, in our opinion, undoubtedly symptoms in which one cannot 1) See p. 144- , 2) Hopkins — Ep. Myth. p. 51, be mistaken, but it might, perhaps, very well be imagined that all the data that strike us as testifying to a phratry-relationship are to be explained from the point of view of local historical fact. It is, after all, not our purpose to maintain that in every single instance where contrasts such as black and white, male and female etc. occur, one is of necessity bound to detect evidences of a phratry-relationship. Phratries in the real sense of the word, that is to say, two exogamous moieties, the parties of the Pandavas and the Kauravas are certainly not. Evidences of the caste-spirit are to be found for that matter, in as early a literature as that of the Brahmanas. If the two parties were real phratries in the narrow sense of the word, we should then find ourselves compelled to assume that the Epic must have been in existence prior to the Brahmanas. And nobody will venture to defend such an assumption. The potlatch does not, moreover, evolve first and foremost in communities in which a clan-organization obtains. The striving after a higher status only begins to assume the proportions of a consuming passion when there are rival societies. In a clan-organization the potlatch is always more or less an attendant circumstance in the great tribal feasts. The veritable „sociétés a potlatch" are those types of culture in which the potlatch ritual has become the end in itself; it is to be found where there are societies, and where clan-exogamy gives way to caste-endogamy. Now it has been our endeavour, in the space at our disposal, to make it clear that the society is in point of fact a continuation of the clan. „The men's house," says Rassers, writing of the clubhouses of the societies, which, as the word implies, are open mainly to men, „the men's house obtains in those communities.... the peculiar organization of which.... is only to be understood when contemplated from the standpoint of totemism." And further: „In the ritual of the sacral men's house (often denoted by the term, potlatch) opportunities present themselves of bringing about alterations in the social hierarchy, resulting in the latent spirit of rivalry bursting out into open antagonism. On such occasions the rival parties encounter one another in ceremonial contests and even in hostile combat; frequently they attempt to outrival one another and to make one another „feel small" by displaying an excessive liberality and revelling in a forced extravagance. ) The study of primitive societies has brought to light that their genetic connection with a clan-organization still continues to exercise its influence even after exogamy has given way to endogamy. The society continues to exercise the function of the clan. That means to say that the important social structure, the tribal dualism, is also met with in the structure of the society in another shape, functionally taking the place of the older phratry-dualism. In every community where the society system obtains they are to be divided into two great societies, sometimes into two groups of societies. Such, in our opimon, is the relationship existing between the two parties of the hpic. They are opposed to each other as phratries, but are not real phratries for all that. They are the two groups of the society just spoken of. If our own conception of the nature of the Vratyas be correct, it must needs follow that the parties in the Epic were parties of Vratyas. As long as such renowned families as those of the Mallas and the Lichchhavis were considered to belong to the Vratyas, there is no reason why it should have been a disgrace to the Pandavas and the Kauravas to belong to the same society. It will only have been so in the eyes of the Brahmins who, for reasons of weight, did not take very kindly to the Vratyas, always granting that we rightly see in the Vratyas members of t e secret society. We need not be surprised, therefore, that the Epic itself does not explicitly mention the Vratyaship of the Bharatas. If such has ever been the case (which, of course, is by no means strictly necessary), the Brahmins will have left no stone unturned to obliterate all vestiges of it. For the rest it is not our intention to link up our idea of the dualism of the two parties in the Epic once and for all with that other hypothesis, viz. that the Vratyas constituted the members of the secret society. Should our own idea of the Vratyas prove to have been erroneous, - we deern it far less probable that that of the relationship between the Pandavas and the Kauravas will also prove to be erroneous. But on the other hand there is no reason why anybody, ~~i) Bijdragen Taal-, Land- en Volkenkunde Ned. Oost-Indië vol. 88 p. 378, 369- merel y to do us a pleasure, should believe that the Bharatas were really Vratyas. There is a text in which it is explicitly laid down that such was, indeed, the case. Prof. Caland records the following: „Der in mancher Hinsicht hochwichtigen Darstellung der sogenannten Vratyastomas wird die folgende Erzahlung historisch-legendarischer Art angeknüpft. „Diesen Ekaha verrichteten einst die Maruts, Visnu war ihr Sthapati; denselben Ekaha verrichteten einst die Göttervratyas (daivya vratyah), ihr Sthapati war Budha, der Sohn des Soma; denselben Ekaha verrichteten einst die Söhne der Kurus, ihr Sthapati war Aupoditi Gaupalayana, Sohn des Vyaghrapad. Als sie denselben verrichtet hatten, gingen sie (d.h. wohl: wolken sie gehen?) als Vratyas zu den Pancalas. Ihre Vater aber sprachen zu ihnen: „Weiiet, o Söhne, nicht unter den Pancalas; die Pancalas sind ja feindlich gesinnt (upavadinah), die werden euch beschreien (berufen, schadigen). „Sie kehrten sich aber nicht an diese Warnung und gingen hin. Da kamen sie in die Wohnung des Kegin Dalbhya. Am nachstem Tage fassten sie diese (d.h. wohl Ke^in und die seinigen), als sie zum Bahispavamana hinschritten, von hinten (sich in dieser Weise bei ihrem Opfer beteiligend) an, also denkend: „das Bahispavamana ist eine Reinigung, in dieser Weise werden wir uns reinigen." Nun befand sich unter den Pancalas ein Vedakundiger Brahmane, Gandharvayana Baleya Agnive^ya genannt; dieser fragte sie als sie sich beim Somaopfer in der beschriebenen Weise beteiligten (wörtlich: „als sie mit hinausschlichen"): „Wer sind es, die dort fortschreiten?" — „Die Maruts sind wir", antworteten jene. „Wer ist euer Sthapati?" „Ich, Visnu", entgegnete der Aupoditi Gaupalayana, der Sohn des Vyaghrapad. „Dann habe ich durch die Fragen:was hast du da gethan, wer hat dies unternommen? implicite euer Vratyatum geleugnet (d.h. meiner Meinung Ausserung gegeben, dasz ihr, die beteiligten Kurusöhne, nicht Vratya, also nicht zum Somaopfer berechtigt seit"; mit diesen Worten wahrscheinlich seine Reue über diesen Zweifel aussernd), sprach er zu ihnen. „Wir haben euch, die ihr unwissend (d.h. unschuldig?) waret, beschrieen; wir thun Abbitte" (Lesart und Übersetzung unsicher). Jene antworteten: „ (verdorben), dessen Söhne hast du beschrieen; deine Nachkommen werden sich verschlechtern". So ging es denn auch (fragliche Übersetzung): seit jener Zeit leben im Pancalalande die Gandharvayana Baleya Agnive?ya in verschlechtertem Zustande, so zu sagen; vormals war es ein machtiges Geschlecht. Wer ein Vratya beschreien will, der soll es so machen, und der Vratya, welcher im Stande ist zu erwidern, soll ihm (dem Beschreier) so erwidern." 1) We shall not venture to make any comment upon this passage. We would only draw attention to the fact that the friendship between the Kurus and the Panchalas was eviaently of a rather unstable character. Here, too, the phratry-relationship explains the different attitude of the Panchalas and the Pandavas to the Kauravas, which does not seem at first sight to correspond with their identity. The mutual relationship between the two phratries is always fluctuating between friendship and enmity. The relationship existmg between the two parties of the Epic may be more specified when observed in the light of this phratry-connection. We remarked above that there is a certain differentiation in status and in function to be discerned between the two phratries. One phratry is frequently more closely connected with the religious life, t e other with the social life, the activities of the one more directly concern the sacral plane, the activities of the other the profane plane of the tribal existence. This social and functiona discrimination is to be seen in the great tribal festive gatherings. The same differentiation is still further sustained. The more social side of the tribal life runs in the direction of the potlatch, while the more religious side is concentrated about the rites of initiation. The same thing frequently applies to the societies cilso» Writing of the Melanesian societies de Josselin de Jong says: „On the strength of what the last ten years have taught us concerning the character of the secret society in Melanesia we are fully justified in seeing an intimate connection between this ceremonial dualism and the social groupmg or classification, even though particulars as to the orgamzation of such societies are not fully known. And also that the appellations Kwat and " n W. Caland - Ueber das rituelle Sütra des Baudhayana-Abh. Kunde d. Morgenlandes XII/i9°3 P- 21 • Sukwe, by which the two groups of societies in various islands (of Melanesia, namely, H.) were designated, correspond to the places allotted to them in the cosmological system. The group designated by the name Kwat is associated with heaven, the group denoted by the appellation Sukwe with the earth. If we were to attempt to define the two separate functions as succinctly as possible, we might say that, the Kwat is the permanent form given to the religious ritual of initiation, and the Sukwe the permanent form given to the social exchange- and giftgiving feasts." x) Such, too, in our opinion, must have been the state of things in ancient India. In the society of the Pandavas we have the permanent form of the ritual of initiation, and in that of the Kauravas the permanent form of the systematized potlatch feasts. Rassers has already seen initiates in the Pandavas. In Krishna-Arjuna, in our opinion, the idea of initiation is concentrated. They are on the side of the Pandavas. On the side of the Kauravas there is Rudra-Siva, the heavenly gambler. The two functions are not mutually exclusive, are also not diametrically opposed to each other, but are complements the one of the other just as are the rival moieties of a tribe. That is the reason why in the preceding chapters we have continually placed the ritual of initiation over against the feast of the potlatch, or in other words Krishna over against Rudra. Now that we have at last returned to the Epic itself from which we started, we shall select two passages in which potlatch and initiation are treated respectively and then proceed to consider these two passages more closely. We shall first take a passage in which the potlatch is spoken of. That is the second, so the Sabha Parvan. ♦ * * „Hearken, o King, how the quarrel about Kingdom began between Kuru's royal race and the noble sons of Pandu, the fruit of a game of dice, and likewise how Pandu's sons came to live in exile in the forest." a) Srinu rajan yatha bhedah Kuru-Pandavayor abhüt Rajyarthe dyütasambhüto vanavasas tathaivacha. *) De oorsprong van den Goddelijken Bedrieger p. 20. 2) I. 61. 4. Thus began Vaisampayana his recital of the great Epic of the Bharatas before the King Janamejaya. It is related in the Sabha Parvan how the sons of Pandu attain to the summit of their glory. At Arjuna's request Maya, the celestial architect, builds a sabha for Krishna, not inferior in magnificence to the splendid sabha's even of Indra, Yama, Varuna, Kubera and Brahma. In fourteen months this palace of delight was completed, made of heavenly materials, and beautiful as a celestial mansion. In the centre thereof the ingenious Asura had built a pond teeming with all manner of fish, tortoises and other denizens of water, and bright and fragrant with the loveliest of lotus-flowers. The odours of sweet-smelling flowers filled the pleasant air, wafted upon a mild, cool breeze, delectable to the sense. Encircling this court was an enclosure of beautiful workmanship (mahamanisilapattabaddhaparyantavedika). On the occasion of the inauguration of this edifice a magnificent feast was celebrated. Yudhishthira made presents of food, jewels and cows to a myriad Brahmins (II. 4. seq.) Still higher ascends the Pandavas' lucky star: Yudhishthira acquires pre-eminence among the Bharatas by his performance of the Imperial Sacrifice (Rajasüya). (II. 12. ssq.). The Kauravas were also invited to attend the ceremony. Yudhishthira requests them to be of service to him by assisting in a right distribution of the gifts collected on the occasion of this sacrifice. Thus it is that the phratries, too, are expected to render one another mutual service upon the occasion of important tribal ceremonies. All the guests invited to the great Sacrifice vied with one another in the giving of sumptuous presents, Duryodhana, whose task it was to accept the gifts in his royal brother's name, not having hands enough to gather in what was given. Everywhere learned Brahmins could be heard disput ing about sophistries in the sacred scriptures. Narratives were recited by others. Natas and Nartakas demonstrated their dramatic accomplishments.1) There was a great multitude of talking, laughing and eating Brahmins, and the continued 1) According to Weber (Ind. Stud. XIII. p. 487 sscl-) ,natf, r?.ust be understood as meaning not only „dancer" but also „reciter . It is told of the Yadavas, Krishna's people, that they „danced a dramatic representation of the Ramayatia (natakïkritam). Hopkins (Great Epic p. 56) thinks of an interpolation. reiteration of „Give, Give!" and of „Eat, Eat!" re-echoed everywhere. The guest-halls, high as Kailasa's sacred summit, were well-stocked with all thmgs. Gold and silver were there in rich profusion, for the king had ordained that his treasurechests should be well-filled before the great sacrifice took place. And when all the princes and Brahmins were gathered together, the proud assemblage was like to the coming together of the immortal gods in the celestial Courts of Indra.x) According to ancient customs presents of great price were distributed among the assembled guests in the sequence of their rank and station. At the advice of Bhishma Yudhishthira resolves to favour the universally beloved Krishna by presenting him with the gift of honour, to the intense displeasure of the envious Sisupala, who does not esteem his mother's brother's son, Krishna, worthy of so marked an expression of honour. Krishna, now raised to the eminence of Supreme Being by the Pandavas, his wratb at last kindled against his cousin, (he had shown exemPlary patience, as the result of a promise made to the king's mother that he would never allow himself to be incensed by her son, in the face of continued insult) declares that Sisupala shall die according to an old prophecy by the hand of Mahakala. Sisupala is then slain by Krishna, who is here, thercfore, identical with Mahakala. This jarring incident over, the Rajasüya can now go on undisturbed. Then begins the Dyüta Parvan proper. It is related that Yudhishthira sees Rudra in a vasion, his countenance turned towards the South. The vision is interpreted thus: the downfall of the warrior-caste of the Kshatriyas is at hand, and Duryodhana is the instrument of the ruin. The soul of the Dharmaraja is very sorrowful. In the meantime it appears that Duryodhana has not yet returned to Hastinapura, the native city of the Kauravas. After the guests had departed to their several homes, Duryodhana and his uncle Sakuni, Gandharï's brother, were the only ones to remain in the palace erected by the Pandavas for the pleasure and convenience of those attending the great ceremony. Full of astonishment and admiration, he is dumb- tn \\ ll 1S related in XIV. 3. that Yudhishthira menticns lack of money n,,Y™Hh 3S t? !7aS,0n fPr not «lebrating an Asvamedha. He says that Duryodhana had already won all his wealth, and that he no longer possesses the money requisite for a great sacrifice. struck at the sight of the cunningly designed works of art presented to his view. But he was not aware of its ingenious inventions being the instrument by which he was to be made a fooi of. He comes upon with what he imagines to be an ornamental pool in the centre. He decides to bathe in it, and takes off his clothes for that purpose; but the transparent pool proves upon his coming into contact with it to be of a less yielding substance than water. It is a crystal floor. He is greatly discomiited and the Pandavas burst out laughing. Then he sees what he takes to be another pond, but this time he wil' defeat those that have made him a laughing-stock; he keeps his clothes on, wa'kes calmly into what proves to be real water after all and not crystal as he had supposed, and gets a ducking. Another outburst of laughter from the Pandavas. He is furnished with a new suit of clothes belonging to Yudhishthira. Thus decked out in another's finery he walks about as if nothing untoward had happened to upset him. Loud and long was the laughter of the Pandavas, when they saw him strutting about thus clad in another's garb; loud was the laugh of Bhima, loud the laughter of Arjuna and the divine twins, the Asvins, and of all the serving-men loud and long was their derisive laughter. But Duryodhana acts as one that has not been baited and does all he can to save his face. He scents a new pitfall in the shape of another pool of pellucid water. He makes a jump, sure of landing in the enveloping coolness of a pool, and - crashes with a heavy thud on to the solid ground. Again the Pandavas give vent to their feelings in an outburst of renewed laughter. He sees a door of crystal which he believes to be open, but is soon persuaded of the contrary by striking his head against it. He sees another door where there is none, and falls through with hands snatching at the empty air. No wonder that, having thus played the röle of the stupid elder brother, the butt of jests and derision (Evam pralambhan vividhan prapya ), no wonder he should have returned home not in the best of tempers. The whole scene is described by the poet in a certain laconic stateliness of style which heightens in no small degree the effect produced upon the reader. Dhritarashtra's son returns home, but his visit had left a deep-rooted envy behind. He was tormented by the thought of the Pandavas' Good Fortune (Pandavasrïtapta) and became possessed of an evil inclination. All his devisings to bring the proud Pandavas to their knees (Book I relates of this) had come to nothing. Of what avail is it for his uncle Sakuni to talk of their own power and possessions and of the impotence of men when the very gods take a special delight in bestowing their divine favours upon his adversaries. But now Sakuni gets a brilliant idea. Yudhishthira was a born gambler and at the same time an inferior player. One might take advantage of both. Gambling is his besetting sin, but he plays atrociously (dyütapriyascha Kaunteyo na sa janati devitum). What if they were to challenge him to a friendly contest. He would be bound to lose but would not be able in point of honour to refuse to play (Samahütascha rajendro na sakshyati nivartitum). Sakuni is an experienced player (devane kusalah) and Yudhishthira is not. If the consent of the old Dhritarashtra could be won, and Yudhishthira were to be sent a challenge, he would be standing face to face with inevitable defeat from the very first throw. It was easy enough to persuade the blind old king and his consent was very soon obtained. He commanded that a magnificent pavilion (sabha) should be built, containing a hundred doors and supported by a thousand pillars, irradiant with the hues of all manner of precious stones. Vidura, the trusty friend of the Pandavas, knows that now the evil hour has come, that Kali is standing at the door. He endeavours to persuade his brother Dhritarashtra so to act, that there shall be no gulf fixed to separate his children (Kauravas and Pandavas), for the sake of a game of dice (Putrair bhedo yatha na syat dyütahetos tatha kuru); but the old king assures him that nothing undesirable can happen in the presence of so many excellent men. Moreover — be it good or evil — a friendly dicing contest must take place; Fate has willed it so (Pravartatam suhriddyütam dishtam etanna samsayah). There is evidently something wrong with the text here, the description of „the brothers' gambling contest, that monstrousabomination"(dyütam bhratrinamtan mahatyayam), is now taken up again at the mention of Duryodhana's complaining of the wealth and splendour of the sons of Pandu. Duryodhana relates how many peoples brought their tribute to the king, his brother. All princes and peoples paid tribute; two peoples alone were not laid under tribute by his brother, the king, Krishna's people, and the people of Drupada, the former because of kinship (sambandhikena) and the latter because of friendship (sakhyena). The jealous Duryodhana cannot bear to think that the elder brothers are becoming less, the younger brothers more (kanlyamso vivardhante jyeshtha hiyanta eva cha). From the first under the influence of Vidura the old king was opposed to Sakuni's suggestion. But Duryodhana argues with his old father until he is convinced. The dyüta has been lawful from time immemorial, and there is neither sin nor hand-tohand fight to it. (Dyüte puranair vyavaharah pranïtas tatra'tyayo nasti na sampraharah). Dhritarashtra reminds him th2t the sons of Pandu are his brothers, but Duryodhana is totally possessed by hatred. The true king, he says, shall strive after power and supremacy, and nothing shall hold him back. When supreme authority is the aim, all means, whether good or bad, shall be employed to attain it. Again there is mention of the building of a sabha. Thousands of artisans were employed. Gorgeous was the decoration thereof; numberless gems radiated their light; seats of gold were set there. No notice was taken of the objections to the playing of the game brought forward by the wise Vidura, such as, e.g. the possibility of the friendly contest ending in contention. Man is entirely dependent upon the will of the gods, is Dhritarashtra s argument. He says, therefore, in other words, that the game of dice destined to take place, is not simply a fraudulent game of hazard arranged with the definite intention of plundering the Pandavas; it was a game to be played under divine supervision; it was a ritual contest of which the final decision was left in God's hands, A messenger is now sent forth to invite the Pandavas to join in the game. (II. 57)- The messenger is none other than the noble Vidura, the very man who was so opposed to the dyüta being arranged between the two parties. His being chosen to perform the delicate task shows that it must have been considered a mission of honour which was not to be entrusted to the first man that happened to come along. He informs the Pandavas that a sabha has been built and fitted up, in which a catch-question, for the Pandavas give no answer. Perhaps the import is this: should it be denied that Yudhishthira is lord and master, it would be the same as admitting that his relationship to Draupadï and the other Pandavas is not that of a lawful head. If Yudhishthira's claims are upheld, Draupadï is lost and the Pandavas have suffered a definitive defeat in the dyüta. At this critical moment, however, Dhritarashtra is terrified by alarming portents, and all at once shrinking from the consequences of this dyüta he saves the situation by unexpectedly bringing about a solutioi) of the difficulty. Three times was Draupadï allowed to demand a favour of the old king. Her only wish is for the freedom of the Pandavas. This unexpected deliverance elicits from Karna the sardonic remark that the Pandavas are indebted to a woman for their good fortune. Now follows that episode in which it is related how Yudhishthira betakes himself to king Dhritarashtra in order to give him an assurance of his attachment to his person, and to ask him what his will is regarding himself, his brothers and Draupadï. (Nityam hi sthatum ichchhamas tava Bharata sasane). Again, then, it is evident that even Yudhishthira does not think of fraud. The blind king explains to him that the match was played with his approval, that it was arranged with a view to his seeing his friends and at the same time to ascertain whether his children (the Kauravas and the Pandavas) were weak or strong. Prekshapürvam maya dyütam idam asïd upekshitam Mitrani drashtum kamena putranamcha balabalam. The Pandavas mount their chariots and, taking Draupadï with them, returned to Indraprastha. And that with all their possessions (II. 73). Nor did Duryodhana like this peculiar result. He prevailed upon his father to have the Pandavas summoned back again. A messenger is sent out to acquaint them with the old king's desire, and catches them up when they are well on the way home. Yudhishthira is informed by the messenger that he has been challenged to another game of dice by the Kauravas with the old king's approval. Obedient to the calling and the rules of conduct of the Kshatriyas king Yudhishthira commands that they shall return, for a challenge to a gambling-competition, and that by an older man's in- 21 struction, can under no circumstances be ignored though Yudhishthira knows that it would bring him rum. Akshadyüte samahvanam niyogat sthavirasya cha Janannapi kshayakaram natikramitum utsahe. The game of dice is resumed (Anudyüta). This time there is only one stake: the entire realm and besides the loser will also have to go into exile for a period of thirteen years. The stake was a rather risky one for an experienced gambler like Sakuni, a player reputed an expert in staking. Had he not already won everything in the dyüta? Why now, then, stake everything upon one single throw? Again the Pandavas lose. They are ruined, and thirteen years of exile await them. Here, too, we read that Dhritarashtra is very much concerned with the'lot of the sons of Pandu, inquiring after them and sympathizing with them in their misfortune. It is clear, therefore, that Dhritarashtra was in real earnest when he invited the Pandavas to come and play a friendly game. The Pandavas are defeated by his own party, but after this daivam, this judgement of the gods, Dhritarashtra still continues to entertain a warm interest in them," for whom this suhriddyüta ended so sadly. The Epic helps us but little as to the techmcal aspects of the game. We know that the akshas were thrown. It is worth noticing that Sakuni played while Yudhishthira only put in the stakes.x) Perhaps the game was played in the same manner in ancient India as it is played among the aboriginal races of the American continent. 2) An essential element of the game was undoubtedly the ostentatious recommendation of the valuables to be staked. The love of pretentieus display in the players got so much the better of decency and delicacy that even Draupadi was sacrificed to it. Sakuni brags that all his riches mean nothing to him. 3) Santi me manayaschaiva dhanani subahüni cha Matsaram cha na me'rtheshu. Yudhishthira announces one of his stakes e.g. as follows: „A thousand rutting elephants have I, o son of Subala; i) Lüders — op. cit. p. 57* !) See p. 251. 3) II. 58- 9- elephants with golden girdles, their heads beautiful with ornaments, the sign of the lotos upon them; hung with chains of gold; well tamed; mounts fit for a king; calm and easy to handle even in the roar of battle; strong tusks they have and mighty bodies; each having eight cow-elephants; all of them disturbers of cities, beautiful like to new clouds. That, o king, is my wealth. For that will I throw with thee." This boasting is part of the game. It is potlatch etiquette. The Kauravas would not appropriate the wealth they had won to their own use either. Presents were made of it to other princes with the purpose of ensuring their friendship (Te vayam (sc. Kauravas) Pandavadhanaih sarvan sampüjya parthivan ) *) This potlatch aspect of the dyüta ceased to be understood later on. We imagine at least that we have discovered things that point to it. We must naturally rely to a large extent upon our subjective insight in a case like this. But perhaps our observations will to a small extent be instrumental in helping others to arrivé at a right understanding of the Epic. There is e.g. Draupadi's appearance in the sabha, a great disgrace to the Pandavas. The messenger who, later on, has to challenge them to battle, begins his invective for that reason with these words: „Once ye were defeated at dice and Draupadi was dragged into the sabha" (Parajito' si dyütena Krishna chanayita sabham). This is even supposed to be the primary cause that led to the horrible war. Asaditam idam ghoram tumulam lomaharshanam. Panchalïm apakarshadbhih sabhamadhye tapasvinim.2) Recounting his misfortunes Yudhishthira mentions among them, without any further explanation, that his wife was compelled to come into the sabha (Bharya cha me sabham nita pranebhyo 'pi gariyasï). 3) ') II- 73- 9* 2) V. 161. 7. • A2- 49 — In tile Vir5ta Parvan Draupadi flies to her husband m the sabha in order to escape from the importunate attentions of Kïchaka. KIchaka brags of his doings later on, however, merely in order to show that he stops at nothing. Nala sojourns with his wife, Damayantï in the sabha. Was the sabha only strictly taboo when there were ceremonies in progress? Or was Nala together with his wife the personification of the dice-playing Rudra-Siva and his consort Lakshmï? Or is the tale in its present shape of a later date? It is, of course, not possible to indicate such a thing witn any precision, but we can imagine that the poet, by adding the dramatic circumstance of Draupadï's having her menses, when she was brought into the sabha, really wanted to explain why Draupadï could not appear in the sabha. In this manner Draupadï's being forbidden to come into the sabha acquires in the eyes of the poet a certain plausibility. It is, of course, all pure conjecture. If we are to have anything like certainty in the matter we should need other MSS. than those we at present possess. It is possible that the interest in the fact of Draupadï's appearing in the sabha then became diverted and was more concentrated upon the circumstances under which she came there. The episode of Draupadï s being continually clad in a new robe by Krishna to save her from the shame of nakedness to which Duhsasana tried to expose her, might also then be seen as a further elaboration of the shifting of the interest in the fact itself to the circumstances under which the event took place. Wmtermtz also takes this miraculous event for a later interpolation. ) To put it shortly, then, the recasting of the text on this pomt is supposed to have taken place something like this: in order to motive Draupadï's not being permitted to appear in the sabha, the poet mentioned her being in this physiological condition. As a further elucidation of the fact it was added that she was clad in only one garment, which Duhsasana moreover tried to tear off. To the greater honour and glory of Krishna this episode was later on submitted to a chastening recast by the introduction of the miraculous re-clothing of the affronted queen. Let us repeat, that this is nothing but a conjecture on our part. But there is a possibility of this conjecture having been the reality. . . It is especially the figure of the king, Yudhishthira, that is thrown into a clearer light, revealing him in vivid colours, once the accumulations added by later generations and hiding his intrinsic character and essential personahty from view, are eliminated. It is self-evident, that Yudhishthira has come to occupy an exceptionally high place in the Epic, the highest place, in fact, of all the multitude of characters that move i) Winternitz — Festschrift Kuhn p. 299. stood, directs its shafts only against a formalistic conception of ritual, not against the practice of ritual in itself.2) Indeed, it is from another source than that of Brahminism that the Song of Krishna springs. It is a remark, fraught with great significance, that Garbe thus makes. It is grounded i.a. upon that passage (IV. i.) in which it is explicitly stated that the Song is an ancient song of the warrior-noble caste of the Kshatriyas.2) That such should be the case fits in with our own conception of the Pandavas as constituting a society. It is a fact that the society of the Pandavas as such can not have been regarded by the Brahmins with a sympathetic eye. It is, therefore, very well understandable that in this Song there should be no very great expression of sympathy with the Brahminic conception of religion, although the mutual feeling of antipathy existing between them will have been very much less bitter than that which drove in a splitting wedge of distrust between thepotlatch society and the caste of the Brahmins. On this point it is more to the seeking heart that the Song of Krishna speaks. Krishna explains in the seventeenth adhyaya that it is faith (sraddha) that counts more than anything else. It is his reply to a question put to him by Arjuna, who desires to know how it shall fare with the man who offers sacrifice without giving heed to the ritual prescriptions of the Sastras. The prescribed course of ritual action may not be deviated from; the true sacrifice, however, must be performed without its efficacy being counted upon and with a believing heart (XVII). There are those who declare that it is meet that Yajna, Dana and Tapas be disregarded, but Krishna declares explicitly that Karman shall not be abandoned (XVIII). Gradually the Song begins to speak in the soft whisperings of the language of mysticism. Jnana is better than Karman, but Yoga is better even than either of these (VI. 46). Yoga is the complete surrender of the human heart in its unwearying quest of God, all its energies concentrated upon the one and only goal of its seeking. It is required of the initiate that he submit to all sorts of taboo-regulations, in order that he may be able to obtain a x) R. Garbe — Die Bhagavadgïta. Introduction. ') op. cit. p. 33. vision of God. *) The initiale must seek by means of a strenuous psychic training to make himself fit to penetrate into the world of the sacred. In like manner must the Yogin, in strict adherence to the prescriptions laid down for the Brahmacharin during the period of his probation (Brahmacharivrate sthitah), concentrate all his thinking and doing, his entire being, upon Krishna (VI. 14). And now the tone of the mystic song becomes ferventer still, more ardent, more intense. It sings of the love of God is the one and only source of knowledge and of action (bhakti). Much has already been written concerning Bhakti. Some scholars are of opinion that the religion of Bhakti was conceived in the intellectualistic circles of the Upanishads. De la Vallée Poussin looks for the origins of the Bhakti faith elsewhere, at all events beyond the pale of the formalistic religion of Brahmimsm proper. Weber did the same, thinking of early Christian mfluences. De la Vallée Poussin also agrees with Barth, citing the French scholar in the following passage: „Ces cultes (i.e. those of the Bhakti religion) n'étaient pas seulement étrangers au vrai Brahmanisme; au début, ïls lui ont été plus ou moins hostiles, hostilité dont la réconciliation postérieure n'a jamais effacé entièrement le souvenir."2) Might we not assume that Bhakti belief became so intimately associated with Krishna worship exactly because it was upon him (Krishna), the divine initiate and culture-hero, the mediator between go s and men, that the love of the seekers after God could most easily be concentrated? The hostile disposition of the Bhakti cults towards Brahminism of which Barth speaks would be sufficiently explained by the anti-Brahminic atmosphere prevailing in the society. It would then be explained whyt e Bhagavadgïta attachés greater vilue to the Bhaktimarga than to the Jnanamarga and the Karmamarga, which with a certam reserve are also allotted a place in its theology. It is the Bhakti belief of the Lord's Song that speaks to us in such clear and eloquent language of the eternal unrest of the human heart in its seeking. Nowhere but in God can that heart find rest. "TTThT^tiate among the American Indians has to give himself up to t^p?acticfof aü manner of exercises until he obtaxns a viS1on of god (R. H. Lowie — Pnmitive Religion). 2) Indo-Européens etc. p. 3H- STELLINGEN I Er is reden om aan te nemen, dat Arabisch sadak betrekking had op het goed, dat bij huwelijkssluiting door de vrouwsfamilie werd gegeven in ruil voor de mahr, het goed van de mansfamilie. Vgl. bijv. Lisan al-'arab Vol. XII p. 65 laatste regel. II De mogelijkheden, die Nieuw-Guinea biedt, dienen niet alleen gezocht te worden in beinvloeding van buiten af door kolonisatie e.d., maar ook in de inheemsche cultuur. III De beteekenis van Javaansch sajembara past evenals Sanskrit svayamvara in de contractueele opvatting van het huwelijk, die men aantreft in maatschappijen met potlatch. IV Javaansch ringgit („een bepaalde munt", tevens „wajangpop") wijst op de beteekenis van het Javaansche geld als „Renommiergeld" en is een zijdelingsche bevestiging van de juistheid der opvatting van W. H. Rassers betreffende de wajang. V Hoewel het werk van E. B. Havell dikwijls belangrijke resultaten oplevert, verliest het aan beteekenis door een te beknopte documentatie. VI Doordat in de etymologie de beteekenis der woorden dikwijls belangrijker wordt geacht dan de klank, kan de ethnologie vaker als hulpwetenschap voor de etymologie fungeeren dan omgekeerd. VII Het is in overeenstemming met de wijziging van het Hoofd stuk van Defensie bij de Grondwetsherziening van 1887 om art. 182 Gw. in verband met de uitzending der Mariniers naar het Saargebied soepel te interpreteeren. THE MAHABHARATA AN ETHNOLOGICAL STUDY G. J. HELD 1 THE MAHABHARATA THE MAHABHARATA AN ETHNOLOGICAL STUDY PROEFSCHRIFT TER VERKRIJGING VAN DEN GRAAD VAN DOCTOR IN DE LETTEREN EN WIJSBEGEERTE AAN DE RIJKSUNIVERSITEIT TE LEIDEN OP GEZAG VAN DEN RECTOR MAGNIFICUS Dr. W. VAN DER WOUDE, HOOGLEERAAR IN DE FACULTEIT DER WIS- EN NATUURKUNDE, VOOR DE VEREENIGDE FACULTEITEN DER RECHTEN EN DER LETTEREN EN WIJSBEGEERTE TE VERDEDIGEN OP DONDERDAG 11 APRIL DES NAMIDDAGS TE 3 UUR DOOR GERRIT JAN HELD, GEBOREN TE KAMPEN UITGEVERSMAATSCHAPPIJ HOLLAND AMSTERDAM MCMXXXV PROMOTOR PROF. Dr. J. P. B. DE JOSSELIN DE JONG The author of this study considering that there might be readers other than professed Sanskritists preferred to apply not the more scientific system of transcription. He expects for instance the transcription Krishna to be easier for an outsider than Krsna. CONTENTS INTRODUCTION r L Short historical survey of epic research — The Epic seen analytically: Christian Lassen and Sören Sörensen seek for a kernei — Albrecht Weber on Veda and Epic; danastuti and Epic Alfred Ludwig on history and myth — Adolf Holtzmann and the inversion theory — The Epic seen synthetically: Joseph Dahlmann S. J. and „die einheitliche Diaskeuase Sylvain Lévi and the Bhagavadgïta — E. Washburn Hopkins, the outcome of his many-sided analytic researches accepted in wide circles. II. The idea of „epic" — The problem: investigation is captured in a vicious circle — Solution: Reconstruction of „epic" culture by means of ethnology — Cultural influence in India. CHAPTER L SOCIAL ORGANIZATION 36 I. Émile Senart's theory on caste — Caste and tribe — Objections to this equalization — Proposed emendation of Senart's theory: Equalization not of caste and tribe, but of caste and clan. II. Social organization in modern India, especially in the Central Provinces — Doublé unilateral tracing of kinship — Circulatory marriage system — Kinship exogamy in the North of India — Hypergamy and circulatory system — Dualism and phratry. III. Social organization of ancient India — Sapinda relationship Gotra and matrihneal clan — Circulatory marriage system. IV. Clan exogamy and caste endogamy — Stagnation in the circulatory system — Date of this stagnation — Clan and functional grouping. APPENDIX 8g Südra and pre-Aryan — The idea of „Aryan" — Division into four and four-clan-system. CONTENTS CHAPTER V. GAMBLING 243 I. Club and status indication — Potlatch — Potlatch and conflict — Potlatch and ritual — Potlatch and game of cnance. II. Indian game of dice — Die and punch-marked coin — Rüpa and emblem. III. Diagram of dicing and classification — The „course" of the die and the movement of the netherworld (serpent) — Dyüta and conflict — Dyüta and ritual — Dyüta and sabha. IV. Dyüta and Rudra — Si va, Kali and the risk of dicing — Marcel Mauss on gift-giving — The Brahmin and the gift — Brahmin and conflict — Brahmin caste and Kshatriya club — Vratyas and Brahmin orthodoxy — Hinduism and organization of clubs. FINAL CHAPTER. THE MAHABHARATA 294 I. The two parties of the Bharatas and dualism — Phratry rivalry and preference for either party — Deceit of the Kauravas in the dicing scene — Deceit of the Pandavas in the fighting scene — Club and dualism — Bharatas and Vratyas — Pandavas and initiation — Kauravas and potlatch. II. Sabhaparvan and potlatch — Ostentation — Draupadï's question — Yudhishthira, the ideal dicer and Kshatriya — Adolf Holtzmann's inversion theory and phratry relation — Alfred Ludwig on myth in the MBh. — Albrecht Weber and the danastuti. III. Bhagavadgïta and initiation — SylvainLéviontheBhagavadgïta — Arjuna as initiate — Jnana and Karman — The effect of ritual — Yoga and initiation — Bhakti and Krishna, the Divine Initiate — Joseph Dahlmann on the unity of the MBh. — E. Washburn Hopkins on the date of the Epic. IV. W. H. Rassers on the Javanese drama — Contact between the East-Indies and India — Marcel Mauss calls the MBh. the history of a potlatch. ing his ideas of the poem belong the labours of Adolph Holtzmann, following in the footsteps of his paternal uncle. At first briefly, later on more exhaustively, this scholar made public the results of his profound investigations. As already said, the Mahabharata consists not only, according to our modern taste, of a number of heterogeneous motives, but those motives also seem to have become merged one into the other. The monarch by reason of piety and virtue, the dharmaraja Yudhishthira, has, in spite of all his righteousness, one great weakness, an inordinate love of gambling, a passion he does not hesitate seriously to condemn in himself. The Kauravas win the game by fraud. Even Krishna, the supreme god himself ,is not only a giver of good but also a shameless deceiver. This was the younger Holtzmann's starting-point in his enquiries in 1884. It was to explain this element of contradiction that he thought out the ingenious theory, which Hopkins later called the inversion theory. The inversion consists herein, that originally not the Pandavas but the Kauravas are supposed to have been the righteous party. A reminiscence of that older partiality for the Kauravas is, according to Holtzmann, to be found in the fact, that the Pandavas are only able to conquer by fraud. Hence all those features in the Epic are to be considered old which tend to exalt the Kauravas at the expense of the Pandavas; and conversely, all those portions of the poem that redound to the glory of the Pandavas and testify to the infamous practices of the Kauravas, are to be looked upon as being comparatively young. In order to motivate this theory on the historical side Holtzmann started from the idea, suggested in the Epic itself, that the Kauravas were the older party. He sees in the rivalry between the two parties the collision between the ancient ideal of chivalry with the growing spirit of a less disinterested practice of statecraft. Now the Pandavas are spoken of as being Vishnuites and the Kauravas as being Sivaites. We might, therefore, expect that the struggle between the two parties in the Epic would be the recoil of a collision between an older Sivaism and a subsequent rise of Vishnuism. There are, however, no traces of such a collision to be found. But there was certainly a time INTRODUCTION been different and they would have acted otherwise. Nothing easier than simply to have omitted the objectionable parts or to have subjected it to a radical revision. But the important thing is this. whether such drastic uprooting of the offensive elements, the elimination of the features in question would not have altered the whole complexion of the Epic so as to render it unrecognizable. Krishna shorn of his wiles, for instance, would be Krishna no longer. Yet another solution was suggested by J. Hertel *). In his opini°n it is not at all unusual to find in manuals concerning the conduct of princes that the employment of artifice in cases of emergency is permitted. But be that as it may, our difficulties concerning the divine Krishna remain just the same. „Aber die Sache steht doch hier ganz anders," says Winternitz. „Das eigentliche Epos ist kein Lehrbuch der Politik, sondern ein Heldengedicht. Es erzahlt von Heldenkampfen, bei denen auf ehrliche Kriegsführung Gewicht gelegt wird. Ausdrücklich wird die Kampfweise der Pandavas als unehrlich bezeichnet. Wenn trotzdem Yudhisthira als Dharmaraja, als Musterkönig vom Standpunkt des Dharma, gefeiert wird, so besteht hier tatsachlich ein nicht ausgeglichener Widerspruch, der sich, wie ich noch immer glaube, am Besten durch die Annahme erklaren laszt, dasz im Epos Lieder verschiedener Barden vereinigt sind, und dasz die letzte Redaktion (nicht Umarbeitung) des eigentlichen Epos von Barden herrührt, die den Pandavas oder deren Nachkommen nahe standen" 2). It is evident, therefore, that the lack of unity spoken of by Holtzmann in his treatment of the Epic is an insufficiënt mainstay on which to tack any sort of reconstruction of the poem as a whole. Nor does the inversion theory assist us in getting back to the poem in its original shape. So far the Epic, however loosely constructed, has proved of firm enough texture to be able successfully to defy all attempts at solving the problems it presents by subjecting it to an examination according to the analytical method. Nowhere has the stratographical method of literary research yielded fossils of a trustworthy nature to guide us in marking off from one another the various layers which W. Z. K. M. G. XXIV p. 421. !) Lit. Gesch. III p. 62a. might have become superimposed upon an original base. This it was that first suggested to J. Dahlmann the idea, that to speak of different layers at all is entirely out of the question, for the simple reason that to all intents and purposes the Epic in its entirety belongs to one single stratum. To speak without the use of metaphors: the Epic, according to Dahlmann, as it is now known to us, is not the work of generations of poets, but of a single diaskeuast, who welded various older elements of the poem into a single whole and produced a work that more or less complied with the requirements of a definite unity: the Epic in its present shape. For however prolix the Epic may be, however often the action of the story may be broken off by tedious didactic poems and irrelevant episodes, the argument is never devoid of a steady purpose, the characters are even drawn with a steady hand. To prove the existence of this unity Dahlmann develops an argument not entirely free from a certain prolixity (under the influence of the Epic, one might venture to say), with the result that the learned author is inclined often enough to repeat himself. According to Dahlmann, then, the elements that go to make up this unity are twofold: epic and didactic; and the problem he sets himself to solve is how both these elements have come to be welded into a homogeneous whole. „Tm Problem des Doppelcharakters von Epos und Lehrbuch liegt das Problem der Genesis des Mahabharata. Warum ist die Dichtung in ihrer vorwaltenden Richtung belehrend, die epische Kunst Tragerin des religiösen Elements?" 1). In answer to this question he assumes the existence of a diaskeuast, who, with the purpose of telling a dramatic and stirring story of the war between Good and Evil and the triumph of Good, recast the extant Epics of the war and defeat of the Kurus by making the defeated Kauravas the representatives of Evil, Adharma, and the conquering Pandavas the champions of Good, Dharma. So the antagonists are designated in the Epic itself. Kali in the person of Duryodhana is opposed to Dharma personified in Yudhishthira. No one need be surprisedat the absence of all traces to be found in history of a battle between Pandavas and Kauravas; this battle only took place in x) Genesis des Mahabharata — 1899 p. 75. the imagination of the poet. Had that titanic struggle ever taken place in reality, there would certainly have been some historical traces of it to be found somewhere or other. We must, therefore, assume, that, prior to the Epic in its present form, there must have existed a twofold literature by which it was influenced; in the first place a narrative literature, in the second place a moralizing. With a view to popularizing the Dharmasastra among the masses, the minstrels simply combined the two kinds. For this purpose they chose with partiality the narrative parts telling of the fall of the mighty Kurus and the rise of the Panchalas. It was in the mind of the diaskeuast that the two parties first became the personifications of Dharma and Adharma. The Panchalas, therefore, may be ranked with the Pandavas, in spite of the fact that the Panchalas were the friends and the Pandavas the enemies of the Kurus, the Panchalas being historically the successors of the Kauravas, but in the didactic literature the champions of Law and Order over against the Kauravas. As early as Panini we have evidences of a cult of Krishna and Arjuna, which shows that even in his time the Epic was not merely a story, but a sacred book, a sastra. This leads us to date the poem in the mould into which it was cast by the diaskeuast certainly not later than the 5th. century B.C. Briefly: „1. Das Epos ist eine zu Recht bestehende Einheit. 2. Die Theile sind in zweckbewuszter und zweckmasziger Weise aneinander gegliedert. 3. Diese Einheit ist von einem einzigen Individuum concipiert und verwirklicht worden. 4. Eine successive Erweiterung oder eine oder mehrere Umarbeitungen sind daher ausgeschlossen" 1). Dahlmann therefore found himself compelled to synchronize the whole Epic together with all the philosophical ideas of which it is the vehicle, with its system of Castes and Guilds and everything relating thereto, with the economical conditions described in it, with the same state of culture of wich the Jatakas are the representation. After a study of the Jatakas, however, R. Fick came to the conviction that centuries must have elapsed between the type of culture as contained in the birthstories of the Buddha and that *) Genesis p. 108. represented in the Laws of Manu and the Epic of Vyasa.1) Dahlmann is also naturally puzzled by Holtzmann's problem of contradiction and lack of unity, but which he only observes in passing. He does, indeed, explain some points, making them less objectionable. In imitation of Ludwig e.g. he looks upon the polyandrie marriage of the Sons of Pandu as a purely mythical element. But that was not the way the diaskeuast looked at it; he does all he can to justify and to motivate the polyandrie marriage e.g. by representing it as a misunderstanding, Kuntï giving her daughter-in-law in marriage to the five brothers, under the impression that her sons were concerned with the eating of a food, of wich all of them had to partake. The dual character of the Dharmaraja Yudhishthira being accounted for by his blind obedience to the precepts of the Dharma is difficult to reconcile with the character of the Epic as a people's book. One can hardly expect to find in a book destined for the use of the people a rule of conduct such as summum jus summa injuria, and all those prolix argumentations on every conceivable subject can only be considered as out of place and irrelevant. The diaskeuast must certainly have had an uncommonly high opinion of his fellow-countrymen's intellectual abilities, for the Mahabharata can hardly be termed a popular work in the more literal sense of the word. It is almost impossible to conceive that Dahlmann should have been induced to connect the didactic and the epic motive so closely with the „einheitliche Diaskeuase". For one is bound to believe with Winternitz that Dahlmann's diaskeuast „zu gleicher Zeit ein groszer Dichter und ein erbarmlicher Stümper, ein Weiser und ein Schwachkopf, ein genialer Künstler und ein lacherlicher Pedant gewesen sei—abgesehen davon, dasz dieser Wundermann ein Kenner und Bekenner der entgegengesetztestenreligiösen Anschauungen und der widersprechendsten philosophischen Lehren gewesen sein müszte"2). And that is, indeed, asking a little too much. And whatever reason this diaskeuast can have had for playing the preacher *) Die Soziale Gliederung im Nord-Östlichen Indien zu Buddha's Zeit. 1897. p. 173 and 174. 2) Lit. Gesch. 1. p. 392. INTRODUCTION with such dogged persistency throughout the whole eighteen books of the Mahabharata — we cannot teil, and Dahlmann cannot teil either: „Wir sehen zunachst keinen Grund aus dem sich die umbildende Tendenz ableiten liesze"1). It is remarkable that Th. Goldstücker, writing yearsbefore Dahlmann, and more immediately concerned with the Epic in its juridical and ethical aspects, had already come to the conclusion, that, viewed chronologically, it could not lay claim to being a uniform production, such being to him most startlingly evident from the obvious discrepancy therein to be found between the author's conceptions of Law and Ethics as they are treated of in pure theory on the intellectual plane and as they are applied in simple practice to persons and events in the reality of the actual narrative.2) Dahlmann is violently opposed to what he terms the proton pseudos of literary criticism: the making a distinction between earlier and later elements in the Epic, the same distinction characterized by Holtzmann as „den obersten Grundsatz der inneren Kritik" 3), and aheady formulated by Franz Bopp in 1829. Such textual criticism is based, in his opinion, „auf die reinste Willkür eines subjectiven Ermessens," the reconstruction of an Epic together with a pseudo Epic founded „auf einem System von willkürlichen Interpretationen." For the rest, Dahlmann's assault upon this sort of criticism does nor demand any Herculean effort; it has been torn in tatters before it has had time to offer any resistance; for it has laid down no reliable Standard by which a justifiable separation can be effected between the older and later portions of the MB h. Andfurther, A. Barth—whose opinion we have here given—is at one with Dahlmann, that the Epic can lay claim to being an unmistakably uniform work. It is not a literary mosaic work made up of fragments and extracts of other works successively added and made to fit into one another as best they could. „C'est un remaniement complet, fait avec une vue d'ensemble aussi conséquente qu'on peut 1'attendre des exigences faciles a contester en pareille matière, de 1'esprit hindou, et qui, selon 1) Genesis p. 92. 2) Barth-Oeuvres III. p. 351. 3) Hoitzmann-Mah. IV. p. i77. 2 toute apparence a été exécutée d'un seul coup ou, du moins, dans des limites de temps trés rapprochés" x). But let us now consider the theory of the „einheitliche Diaskeuase" itself. It is now, says Barth, exceedingly probable if not absolutely certain, that the legendary part of the Epic came into being in a very distant age (perhaps farther back than Weber imagined, who dated it after the age of the Vedas) and certainly in a milieu different from that in which Vedic literature originated. It is certain, too, that there was a Mahabharata in existence several centuries prior to our own era, but it is equally certain, that, judging it as one judges an archaeological site according to the latest remains, the Epic is comparatively young. And Dahlmann neglected most of the facts upon which the latter conclusion is based. Dahlmann's views met with opposition from various quarters. A formal Dahlmann literature was the result. But now that the clamour of opposition, at times rather vehement, has quieted down, the final judgement seems to be: „weighed in the balance and found wanting." But in the present instance, too, we must take care not to reject the good with the bad. There is in the first place always the question raised by Dahlmann: in what manner are we to establish the connection between the epic and the didactic element; and secondly: granting that there is an element of unity in this vast and complex poem, in what manner was that unification brought about. That the presence of such a unity in the poem must be taken seriously is evident from a short article by such an authoritative scholar as Sylvain Levi, who stated as his opinion a number of years ago, that the Epic was written to instruct the knightly order of the Kshatriyas in the service of Krishna as a warrant of their welfare and success. The idea was suggested to him by the continual identification of Aijuna, the ideal Kshatriya, with the divine Krishna, an identification which ca lis up before the mind a different pronunciation: yatah Krishm s tato jay; h, with the frequent complemental analogous formula: yato dharmas tatzh Krishnah, the two together resulting in the for mula: yato dharmas tato jayah; unde Krishna, *) Barth-Oeuvres IV. p. 354. INTRODUCTION inde jus; unde jus, inde Victoria, ^^hen we consider what an important, not to say a central place Krishna occupies in the Epic, every parvan of which commences with an invocation to Nara and Narayana, i.e. to Arjuna and Krishna, it is very clearly evident what importance Lévi attachés to that unity in the Epic so distinctly emphasized by Dahlmann. With all itsfantastic extravagance, its vast accumulation of richly varied detail the foundations of the poem are to be sought, according to Lévi, in a Kshatriya manual based upon the moral code and system of conduct of the Bhagavatas.1) As for Lévi's own opinion (Winternitz, exaggerating, calls him „Dahlmann redivivus )2) we must agree with Winternitz, that, thus considered, there is but little room left for the action proper of the poem. The Epic not only teaches but also tells; it tells a tale in the first place and teaches a lesson only in the second, which in Lévi s brief article is not done full justice to. For the rest Lévi has only touched lightly on the subject, and to discuss it in detail here and now would be out of place. Inasmuch then as it is evident from the foregoing observations that the synthetic method has given us but little to h°ld on to, scholars have come to abide by the more analytic method, championed in the person of the scholar, already mentioned, E. Washburn Hopkins. He started from the idea, as sound as can be in itself, that it is no use trying to synthetize things unless one knows what one is going to synthetize, An attempt at combining things and ideas must be preceded by a definite knowledge of the things or ideas to be combined. In a number of essays, long and short, he sets forth what is to be found in the Epic in different domains, and leaves the subject of its origin, as a point of premature speculation, for the time being more or less untouched. In his book „The great Epic of India" (1902) Hopkins comes to the conclusion that it is evident from a comparison with the other literary contents, that the Epic itself, examined from the synthetic side, must be dated after the youngest Vedic works. Whether the Puranas were written before or after the appear- in Ln?B"anïSInTi!°pfiB3h-2o!rkar'IQI?' PP' "~I06; in English 2) Lit» Gesch. III. p. 627. ance of the Epic is a question that cannot be unconditionally answered. Hopkins looks for the origins of the Epicinthepriestly hymns accompanying the annual cycle and in the songs in praise of the liberality of princes, the narasamsï gatha, which had already been pointed out by Weber. This ancient poetry must be dated perhaps 700 B.C., perhaps 1700 B.C. But talking about the Epic in its present shape, we must think of a time, „when the sixty-four kalas were known, when continuous iambic padas were written, when the latest systems of philosophy were recognized, when the trimürti was acknowledged, when there were one hundred and one Yajur Veda schools, when the sun was called Mihira, when Greek words had become familiar and the Greeks known as wise men, when the eighteen islands and the eighteen Puranas were known, when was known the whole literature down to grammars, commentaries, Dharma^astras, granthas, pustakas, written Vedas and complete MSS. of the Mahabharata, including the Harivanga."1) By thus arranging large numbers of heterogeneous details in their mutual connection one with another and each with all, Hopkins finally succeeded in constructing the developmental history of the Epic as follows: 400 B.C. There is a collection of Bharata lays, in which the Pandavas are as yet unknown. 400—200 B.C. There springs up a Mahabharata tale, in which the Pandavas are the heroes. Krishna is a demi-god. 200 B.C. to 100 or 200 A.D. Krishna now becomes the all-God. Interpolations of a didactic nature; new episodes added. 200—400 A.D. The introduction and later books are added. Hopkins, whose „Great Epic" is not unaffected by his rejection of the theory of Dahlmann, is convinced that four points may be considered absolutely certain: that the Epic in its present form or even free from its didactic matter was composed or compiled after the invasion of Alexander; that all this didactic matter was inserted later on; that the Mahabharata was not essentially a book of Krishnaite belief until somewhere near the first century B.C.; that the Epic was practically completed about 200 A.D.; that no date can be found to cover the entire 1) Great Epic, p. 386. Epic, although, broadly speaking, it can be said to have existed in the second century B.C. In pursuing the analytical method a great deal is naturally left to the individual insight and imagination of the investigator. Special attention, says Ludwig, for instance, must be given to the way in which the various episodes have been joined together, whether they have been welded into an harmonious whole or whether they have been put together clumsily, awkwardly. One must be on the watch for „auffallige mühseligkeit" and „unnatürliche gesuchtheit der motivierung", for „absolute überflüssigkeit, widerholung des themas, anknüpfung an misverstandenes, incongruenz zwischen dem zu exemplificierenden und der exemplification"; one must further pay attention to metre, language and style1). And then, as a result of such a thorough examination of the Poem, seen and studied from various angles, one can hope, in the end, to arrivé at a reliable analysis. It certainly does not behove us after all that Dahlmann has said on the subject to dweil again upon all the shortcomings of this oftentimes too subjective method of study. As a matter of fact no theory can ever be built up without subjective views, however often people may assure us that they are going to work purely objectively and that they will only allow the facts of the case themselves to influence their conclusions. The facts, as a rule, do not of themselves speak a very plain language, and, therefore the subjective insight of a competent expert is to us of more value than the objective judgement of an outsider; but there are, of course, certain limits. For if we are really to take into account what the great Sanskrit scholars have said concerning the study and the contents of the Mahabharata, the only thing now left to us to do is to hope that the truth will one day emerge from the present confusion of conflicting opinions and ideas. And even though the subjective method of study above-mentioned may be the cause for the greater part of the contradictory character of the results of expert research work, there is still so much contradiction, so much inconsistency and so little agreement, so little harmony of opinion left, that one feels the whole Epic melting away in one's *) Stzb. K. Böhm. Ges. Wiss. Cl. f. Phil. 1896. V. p. 28. hands beneath the high light of this method of critical inquiry like snow beneath the sun. And it is certainly not concerning trifles that agreement is wanting. Take, for instance, the figure of Krishna as starting-point. It is the opinion of a large number of scholars all those portions of the poem must be taken for interpolations of a later date, in which divine homage in an especial manner is rendered to Krishna. But these portions are so closely related to, so essentially interwoven with the rest of the work that they do not admit of simply being separated from the context without more ado. The text itself does not offer any justification for such wholesale mutilation, hence we must take the dual character of Krishna himself as our starting-point. That the Epic itself declares that not everybody knows the divine form of Krishna and that in his divine form he is hard to comprehend, simply proves that he did not always appeal equally to everyone. „Man müszte" — says Ludwig, speaking of those characteristics in which emphasis is laid upon Krishna's divinity —" die behauptung aufstellen dasz alle dergleichen anspielungen, auch wo gar nichts darauf hinwiese, doch nur spatere einschiebungen seien, die den zweck gehabt hatten, die göttlichkeit Krsna's überall durch zu führen. Das hiesze aber das durch willkürliches verfaren erst hervorbringen, was man als das ursprüngliche vorausz gesetzt hat. Man bewegt sich hierbei in einem zirkel"x). We should say that is clear enough. And that such critical inter pretation is to be taken ser iously we have already seen in the case of another scholar, whose authoritativeness nobody will venture to call in question,Lévi,who considers Krishna the pe'rsona sine qua non of the entire Epic. So it is not Dahlmann only who considers Krishna as belonging to the older portions of the poem, but also Barth and Oldenberg together with Lévi are of the same opinion. According to Winternitz it is essentially the Sivaite features, which „überall leicht als spatere Zusatse erkennbar sind". But according to Holtzmann it is possible to eliminate both the Sivaite and Vishnuite portions of the poem as interpolations of a purely sectarian nature. Let us consider another important point: the Bhagavadgïta. This work occurs in the Epic immediately before the descripi) Abh. Kön. Böhm. Ges. Wiss. VI F.Bd. 12. p. 15- tion of the great battle. In the opinion of many scholars, therefore, it must be relegated to the division of „clumsy insertions", and looked upon as a later addition. It is really an interpolation, says Hopkins, and it is even „unnecessary to prove it".1) Not at all, says Lévi; on the contrary: „Cet incomparable dialogue, souvent considéré comme un hors d'oeuvre sublime, est tout au contraire le coeur et le noyau de 1'ouvrage"2). And now to lay stress upon the aesthetic criticism of the work: Winternitz is not at all impressed as Lévi is by the sublime nature of the Bhagavadgïta. He is even almost surprised that this Song could have aroused somuchenthusiasm in East and West, which he ultimately ascribes not to the profound thoughts and the inscrutable wisdom it contains in the opinion of most Indian and a great many European scholars, but to its purely poetic qualities, the sublime language and the splendour of its poetic imagery. Let us leave it at that. It must be admitted that there is yet every reason to devote time and energy to the study of the Epic. Many of those reliability certificates which have been issued for various portions of the poem by famous Sanskrit scholars are all but worthless, even though one have the greatest respect for the illustrious names by which these declarations are supported. Adding all these various and contradictory statements together, we find that there is practically nothing left of the Epic at all and everything of the points at issue. We trust that we have not thus sketched the actual state of affairs in which the critical investigation now finds itself — more or less in a cul-de-sac — in all too tendentious a manner. It was in no way our intention to be able thus to extend a more hearty welcome to the method of study which we ourselves desire to pursue as a sort of deus ex machina arriving in the nick of time to solve the difficulty and with whose assistance we shall conjure up what has really been for a long time common knowledge. One need only read what Oldenberg wrote in his posthumous work on the origin, form. and contents of the Mahabharata. „Das Mahabharata begann seine Existenz als einfache epische Erzahlung. Es wuchs im x) J. A. O. S. XIII. p. 204 nt. 2) Comm. Ess. pres. to Sir Bhandarkar p. gg seq. Laufe der Jahrhunderte zum ungeheuerlichsten Chaos"x). And the Epic is still virtually a chaos. Critical investigation has settled that the poem existed in part some centuries B.C. and that some centuries after Christ it must have been practically complete. What are called „epic elements" are to be found as far back as the age of the Vedas, whether the sacred songs of that age have been given a more essentially human character or whether, on the contrary, they have become more theologized (we might almost say „sacerdotalized"). We may also consider the existence of mythical elements in the Mahabharata established, though perhaps not all investigators will feel inclined to follow Ludwig in this respect, at least not straight away. Further that the Epic is a Dharmasastra and a continuous production, in which epic and didactic elements are interwoven, even though we may not yet know how it was that this combination was brought about and nobody is willing to acknowledge a unity in the poem as Dahlmann does, at least not in the same sense. It has further been established that the poem is characterized by a peculiar contradictoriness, a remarkable inconsistency in the representation of all kinds of important facts and events and in the depicturing of the protagonists in the drama, even though one may not believe in an inversion as conceived of by Holtzmann. Let us bear in mind in the face of all criticism of the work done by Ludwig, Holtzmann and Dahlmann, that these scholars have so far received no answer to the questions proposed by them2). Unity and yet chaos — the two points of view are irreconciliable, however one may puzzle one's brains to arrivé at a satisfactory solution of the difficulty. There seems to us to be but one possible way out of the impasse; we are at a deadlock, but the fault may very well be in the scholars themselves and their methods of investigation and not in the Epic. Now when one queries why so little success has attended the indefatigable labours of famous scholars and critical investigators during several decades, it occurs to us that the primary answer must be: Because one has taken as a self- *) Das Mahabharata. Göttingen 1922. p. 1. 2) We would again draw the reader's attention to the fact that we are here only concerned with studies of a more synthetic nature as to the origin and character of the Epic. evident truth the statement, that the Epic consists of a nucleus of an essentially narrative nature around which have collected so many different threads that this kernei has ultimately become indistinguishable.1) This has come to be considered as being so obvious, that scholars have not found it at all necessary to even take the trouble to determine the purport of the axiom so unconditionally laid down. Thus, without more ado, was the path pointed out which the investigators were to follow in the first place: the path that was to lead to the discovery of the original central element or kernei; whether the kernei was to be found by removing the surrounding husk, according to the course of procedure laid down by the originators of the analytical method; or whether the form and composition of the kernei was to be determined by a consideration of the nature of kernei and husk together, i.e. of the original narrative together with its supplementary episodes, according to the synthetic method of inquiry. No one will think of denying the existence of a kernei to the Epic. Nor shall we. But, all the same, we doubt whether the image, employed by Hopkins, of a kernei enveloped in a tangle of well-nigh unravellable threads is, after all, quite a happy one. Why not threads only, as a matter of fact, wound, if you like, round about a kernei, but about a kernei made by the threads themselves ? If the latter image comes anywhere near the facts of the case themselves, the words analysis and synthesis lose their narrow limitations. A tangle can only be straightened out by the happy use of analysis and synthesis together. It is, of course, not our intention to consider the difficulty as settled with the employment of this metaphor. But we may be permitted to remark at once, that it is of no importance at all as regards the Epic itself whethersomeepisodeorotherhasproved to be a later interpolation, or not, speaking, of course, with a view to our knowledge of the state of culture, represented therein. With regard to the language it may, of course, be of some importance to know whether a certain part is lying before us in its earlier or later shape. But it cannot help us to arrivé at conclusions as to the cultural aspects presented for our consideration. Irrespective of its value (certainly not very 1) Hopkins-Great Epic. p,|3Ö3. great) for philology as such, we can say without fear of exaggeration, that the only value of the „original Epic" lies in its being a relic of antiquity. In whatever way considered, the interpolation also requires a place, which means, that, when occupied with an interpolation, one should really be able to explain the why and wherefore as well as the sources of the inserted part. It is a simple enough thing in itself merely to remove those parts which one does not know what to do with as interpolations pure and simple, but it is just then that the actual difficulty begins. It is a pity, but it is impossible from philological data to get to know very much of the culture that developed the language and in which it was a living strength. This rage for eliminating is for a great part due to that distinction between husk and kernei which has led almost of itself to kernei and husk being pulled to pieces. The unsatisfactory results of this method of investigation are a proof that there must be something wrong somewhere. It is not without reason, that Lévi warns students against a too zealous application of the analytical method of criticism: „L'existence d'un „Mahabharata primitif" est une question singulièrement obscure, et peut-être décevante"1). What is now to be done ? Shall we have to acquiesce in what Mme. de Willman-Grabowska wrote last year concerning Dahlmann and Holtzmann: „Des savants ont tenté de construire le Mahabharata primitif. Cette tentative était vouée a 1'insuccès, car 1'histoire même de la lutte fratricide nous est arrivée bien altérée, et il est difficile d'opter pour une version, encore moins pour une version épurée selon nos goüts modernes". But we cannot get rid of Holtzmann and Dahlmann2) so summarily. The study of literature is, after all, dependent on other things than the taste of the student. And yet there is something in the words we have quoted that gives food for thought. The study of epic and perhaps, also, of other branches of literature is largely, indeed too largely dependent on the *) Journ. As. 1929 vol. 265, p. 248. 2) Dahlmann really ought not to be mentioned in tbis connection, his studies not being directed towards the unravelling of the primitive Epic. L'Inde antique, Henri Berr — L'Ëvolution de 1'Humanité. Vol. XXVI. p. 299- subjective taste of the student. Now if something else could be found to replace such purely individual taste, something of a somewhat more objective nature, we should have found a way out of our difficulties. * * * We are accustomed to call the Mahabharata an Epic, further specified as an Epic with a mass of didactic interpolations. The word itself does not say much. In a general sense we think of a kind of literature which first saw the light at the courts of the nobles, a literature that became the more fabulous and fantastic the more liberally the noble was able to reward the poet. The aesthetes or professed appreciators of the beautiful are wont to put the legendary element aside with a gesture of obliging friendliness as the product of a childlike imagination, belonging to the department of the psychologist; the historians do the same, but often more forcefully, all these by-products being looked upon as considerably adulterating the pure residue of history. That does not mean to say, of course, that it is our intention to be at all ungracious to the scholars above-mentioned; we merely wish to point out, that the Indian himself looked at his great national Epic from an entirely different angle. It is a book for which he has great veneration, not directly acknowledged as a book of divine revelation, but all the same cherished and revered as a work of great religious significance. It is called the fifth Veda. It tells him of his gods and extols the feats of brave and noble men, the heroes of the Epic moulded of common clay, who come to be revered as highly as the gods themselves, as is evident from a well-known passage in Panini, in which there is mention of a certain Arjuna cult alongside that of Krishna. Now nobody will doubt the importance of deciding for oneself, when subjecting a text to critical investigation, the nature of the text with which one is occupied, what sort of text it really is that one is studying. It cannot be the same thing whether the book lying before one is a theological treatise or a volume of poetry. And this is of all the more importance when the work in question has to be critically studied with a view to discovering on aesthetic grounds the genuine basis of it. And however surprising it may sound, no one, as far as we know, has ever formed to himself a clear idea of what is really meant by the word „Epic". It seems of itself to be so clear that any attempt at defining it would be like forcing an open door. And yet this apparently clear-cut idea is so vague, so hazy, that one can hardly be said, by examining it more closely, to be running the risk of making a show of being afflicted with that sort of critical mania which would even reduce God himself to a mere formula. For itis exactly the „legendary element"intheEpicthat puts us on our guard. If all this must simply be put down to the peculiar spirit of Indian mysticism, we have assumed too many unknowns ever to be able to cherish any hope of ultimately finding a solution to the problem. For it is quite possible — and we hope to be able to show that it is not only possible but a literal fact — that all this has not welled up from out the unfathomable depths of a purely mystical mind, but that there is a satisfactory explanation of the legendary elements to be found in Indian culture itself which will furnish us with a key to their correct interpretation and lend them, at last, significance. Now it is a fact that Indianists have so far given but little attention to the idea of „culture" in their studies. It was in the nature of things, that precedence should be given to philological research. For the rest, as far, for instance, as religion was concerned, scholars simply adhered to the results yielded by the nature-myth method of explanation, which method of procedure is still far from being played out in Indological study. It is comprehensible, that, in a time when mainly the language itself was being studied, the word „Epic" called up in the mind of scholars ideas analogous to those connected with the heroic literature of mediaeval Europe or of ancient Greece in the Homeric poems. But it has become more and more evident of late that a culture should be studied with the same accuracy as a language, and that not only in the case of our modern culture but also where it concerns the culture of ancient India. It has become entirely inadmissible, for example, to look upon a primitive religion simply as a collection of singular but „superstitious" practices. One cannot get to know Indian culture simply by entering into the spirit — as it is so often called — of that culture. Sanskrit is not learned by entering into the spirit of that language. The being brought up in the practice of a language or the being educated in the spirit of a certain culture as a member of the linguistic or cultural community is, of course a very different thing. And even then — when it is a question of being able to comprehend a language or a culture with a critical insight, and with historical understanding, it is of little use having been brought up in their use and practice. Language and culture study in their entirety are adequate evidence of the truth of our assertion. It does not do after having produced a scrupulously careful translation of some work or other to simply further hazard a guess as to the significance of the text translated. It is in this peculiar state of things that, in our opinion, the reason is to be sought, for a large part, why the critical study of the Mahabharata has hitherto yielded such meagre results. The Epic has been studied overmuch from the purely philological side, with the typical result, that the „original Epic" loomed up in the imagination of scholars as the one object of their scientific quest, whereas its discovery can only be considered of any great importance from the point of view of its value to the science of philology proper, and cannot be made to serve as a criterion, a Standard by which to judge the nature of the cultural elements therein described. Philology has yielded splendid results in Indian studies. But that science has, in our opinion, been now and then too heavily burdened, and mainly on the etymological side. Obviously it is of importance to know, for example, that the words, Dyaus and Zeus are etymologically related to each other, as long as one does not presume on the strength of their etymological relationship to be able to determine the position and the significance of either of these gods. And we generally see, too, that this sort of thing is, more often than not, attended with a vast deal of speculation, of the so-called „naturemythological" type. For it is obvious that words etymologically related to one another may have entirely different meanings in various languages. Words are seen to wander far afield. Take the word „flirt", for instance; if this word could not be historically traced, it would not be so easy to determine that „flirt" and „fleurette" in French are the same, „fleurette" having found its way into English and, in its altered spelling of „flirt", found its way back again into the land of its origin.1) We cannot do better than stick to the fundamental principal of all etymological research, contained in the words: Verba valent usu; this „usus" is continually leading us back to the culture in which the words are living and functioning. It is by no means our intention to belittle the labours of philologists in their own peculiar domain (such conduct would simply recoil upon ourselves), but merely to press the point, that the study of language and of culture should each be more narrowly restricted to its own particular domain. The idea that a people can be understood by a knowledge of its language is an exaggeration, rendered all the more dangerous by its universal acceptance, and that, not in the least, as regards the study of language itself. One meets with all kinds of information regarding the text and language of the Epic without all the time knowing what the Epic — we shall continue to employ the term for the sake of convenience — in reality is supposed to be. To know what an Epic is, we must know what sort of culture it was that produced it and likewise what place it occupied in that culture. Now in the study of the Mahabharata we are up against a peculiar kind of difficulty, all data concerning the epical culture having to be gathered for the greater part from the Epic itself. The so-called epic period, fixed by Hopkins at 300—100 B.C. must be extracted from, and got acquainted with, through the Epic itself. The attempts of scholars to arrivé at reliable data from a study of the Epic's connection with the older vedic culture have so far met with but little success, the nature of that connection not yet being clearly defined; we only know that it was there once. A knowledge of the culture is essential to an understanding of the Epic and an understanding of the Epic is indispensable to a knowledge of the culture. One need only call to mind the figure of Krishna, whom we know mainly through the Epic, and our knowledge of whom must be further employed as a basis for the reconstruction of the Epic's history. We must start from A to demonstrate the existence of B and then retrace our steps again, this time 1) J. Vendryes-Le Langage, p. 227. in order to prove the existence of A with the help of B. To escape from the pernicious influences of the vicious circle, in which students of the Mahabharata now find themselves, we shall have to start our investigations anew from the outside; we must break through the meshes we have collected around us; and the position we wish to choose as our coign of vantage for further examination is that of the ethnologist with the latest facts at his disposal yielded by modern research. It is not part of our plan to submit all manner of disconnected elements of culture from various areas to a comparison with cultural data as supplied by the study of Indian civilization. Isolated elements of culture cannot be compared unless something definite is known concerning the state of culture in which those elements are to be found. The comparison of various tales one with another is doomed to remain ineffectual unless one is acquainted with that particular form of culture in which the tales to be compared are to be found. That explains, in our opinion, why, hitherto, the comparative study of the epic poems of India and Greece has yielded such meagre results. We cannot express our purpose better than by turning, mutatis mutandis, to our own use, what W. H. Rassers, of the Leyden Ethn. Museum, wrote in one of his essays: „My purpose here is to do only what is primarily necessary, viz. to determine in the first instance with what form of human society (the Mahabharata) is genetically connected" x). We may, however, be permitted to make a few introductory remarks by way of elucidation with a view to obviating misunderstandings. We have already said that the Epic is usually admitted to be a conglomeration of decidedly heterogeneous elements, which scholars, during the last few years, have given up studying as a whole. They have now confined themselves to a detailed investigation of portions of the poem with a view to determining their origin. A similar course of action is followed in the study of Indian society in general. Fragments of a miscellaneous character are torn from the whole with a view to investigating their origins, apparently in the hope of thus being able to determine the root (Aryan?) elements proper. The choice of such fragments *) Contr. Geog. and Ethnog. Mag. Neth. East Indies. vol. 88. p. 320. is often influenced by a vague sense of preference or by some traditional idea. A good example of one of these traditional ideas is the belief, that the oldest immigrant nomads must have, resembled the Germanic tribes described by Tacitus in his Germania.1) Ifweareto avoid confusing language, race and culturet we must take care not to confound the problem of race with thae of culture.lt is, in our opinion, erroneous to maintain, that the Mahabharata in its primitive shape is non-Aryan byreasonofthpolyandric union of the sons of Pandu and the grim bloodthirss tiness of the warlike Bhïma, and such-like. The erroneousnes, of such a wide-sweeping assertion is not even, in the first placeg to be found in the fact of only a vague sense of partiality beino made to serve as evidence. Even though such features were to be considered demonstrably Dravidian, that does not mean to say that the difficulty is thereby solved. It is not enough to have demonstrated that the work harbours a number of borrowings. The difficulty is not then done with; that is just where it in all seriousness begins: the why and the wherefore, the manner of the borrowings and the place they occupied both in their new and in their old environment. It must not, therefore, be imagined, that in treating the object of our inquiry as an ethnological problem we shall by preference concern ourselves with the study of those elements which are deemed of Dravidian origin. We shall first take Indian culture as a whole. The borrowings of themselves only seldom suggest an explanation; they must more often themselves be explained. There was a time when attempts were made to explain the religion of India — at least, partly — by means of classic culture. By so doing the ideas of culture and language become confused. For it is quite possible that two peoples may speak related languages and yet have different cultures. It is with good reason, therefore, that of late emphasis has been laid from another side upon the influence of Austro-Asiatic culture, which the language itself is supposed to be able to show. But here again the selfsame danger menaces; the elements that prove or seem to be Austro- Asiatic run the risk of being eliminated as foreign to the culture proper (i.e. the Aryan culture). „There do indeed seem to be sound reasons for assuming the ') v. criticism on this in Camb. Anc. Hist. Vol. i. p. 66. existence in many philological, ethnological and cultural data of the rudiments and influence of an Austro-Asiatic substratum", says Professor Gonda of Utrecht1). It is the use of the word „rudiments" that worries us. Reading J. Przyluski's interesting article on the svayamvara, for instance, one cannot get rid of the impression, that here that dangerous word „syncretism" is lying in wait,biding its time2). Przyluski is uncertain whether that form of bridal, in which the bride herself selects her husband from the suitors for her hand, is an Aryan or non-Aryan practice, but finally comes to the conclusion that it is to be considered as a non-Aryan custom. In Further India it is the custom for the boys and girls to play a game of ball, the girl becoming betrothed to the boy who throws the ball she catches. Of the svayamvaras in India only that of Damayanti is genuine, Damayanti herself giving a garland to the man of her choice. In other cases (cf. Sïta and Draupadï) this practice has been replaced by a physical test of skill or endurance submitted to by the man, that man being chosen whose skill or endurance is greater than that of the others. The first case of the girl herself making known her choice is accounted for by Przyluski by assuming the existence of a matriarchal organization, and so, matriarchy prevailing in Further India, he scents Austro-Asiatic influences. The second case shows us how the peculiar practice of a matriarchal form of human society has found its way into the sp'here of patriarchal, i.e. Aryan, culture, and become a trial of skill or test of endurance to be undergone by the man instead of a choice to be made by the woman. Although we are in every respect inclined to assume that Aryan culture on this point came under the influence of AustroAsiatic cultural elements, still we feel bound to say, that this method seems to us to be a dangerous one, and that not in the first place because of the peculiar use of the terms „matriarchy" and „patriarchy", but especially because we are not made to see how the non-Aryan practice of the svayamvara was grafted upon the stock of Aryan culture. The girl is not free in her choice in Further India either, and the trial of skill is to be *) Gonda, Inaug. Addr. Utrecht 1932. 2) Journ. As. 1924. Vol. 205 p. 101 seq. 3 various peoples"1). That was as far back as 1906. We hope to be able to show in the present study that it was not without reason that Prof. Caland even then directed the attention of philologists to the desirability of taking up the study of ethnology, and, further, that this branch of knowledge, thirty years later, has a still stronger claim upon our interest. Let it not be imagined that we are intending to demonstrate our own particular method as the only certain and infallible course of critical investigation. Ages of profound study and patiënt plodding will act as a gentle reminder to the student who would consider himself the exclusive possessor of a purely objective judgement to be modest in delivering his opinions. We only hope to draw the attention of philologists to the vital importance of the science of ethnology to a clearer understanding of the nature of the peculiar difficulties continually encountered by them in their own particular branch of study. The nature of the idea of „culture" as developed by de Josselin de Jong in his lectures at the University of Leyden presupposes and demands our considering, the religious, social and economie manifestations of Indian culture in particular alongside one another as three aspects of that culture. The purpose of the present study, therefore, is in the first place to endeavour to determine what form of human society it is which is represented in the Epic and then to return again in our final expositions to a consideration of the Epic proper. ') W. Caland — The Study of Sanskrit in connection with Ethnology and classic Philology. Utrecht 1906. I. SOCIAL ORGANIZATION It may at first sight seem somewhat strange that it should be necessary, in order to put the difficult problem of the MBh. to begin at once with the other, equally difficult, problem of the social organization in India. However, this need not a priori be the one and only method to render the problem in question totally insoluble. This treatment might be compared to the way one sometimes sets about translating a difficult passage in a text. In this case the context will often provide an explanation. This context may in itself offer more difficulties than the passage at issue, though in spite of this, the meaning of the first passage has become clearer through this context. What will be put forward in this study about Indian culture may after all best be considered thus. It stands to reason that we do not intend to give an outline of Indian culture and in the mean time just steer epic research into better channels. At best we can suggest the direction and method in which, in our opinion, the right way may also be sought. Nor would this endeavour be too ambitious. For a description of the whole range of Indian culture in a small compass has repeatedly been tried. This is not the purpose of this treatise. It is here only a question of collecting data by which the connection between myth, rite and social organization of the Epic shall become clearer. Hence it is necessity that makes us enter the maze of the Indian social structure. The use of the word „maze" in this connection is no exaggeration. We shall try to determine a few fixed points in this labyrinth on which epic research may be based. That also in this treatise, after all that has been written about the social structure of India, one chapter is devoted to this many-sided subject, is not done in order to make this study more complete, We do so, because the literature available does not supply what we want. It is the large range of the literature relative to the matter that makes the study of this subject so difficult. It is also a deplorable fact that the theoretical speculations are more numerous than the collections of materials. Also for this interesting and much varied range of culture ethnographical fieldwork is urgently required. It is true, much has already been done by good ethnologists, but very much remained undone. Now that we venture into the labyrinth we shall by way of precaution take ethnology in hand as another thread of Ariadne. Even then the enterprise remains risky, for not everybody is a Theseus, but at any rate we are prepared to face great difficulties. Who says „the social organization of India" says „caste". This is the magical word that has moved so many pens. We shall not treat of all the theories. In short there are three factors which, either separately or in combination, are developed in the different theories: i. that the caste system is a question of race and the consequence of racial contrast (Risley); 2. that it is a hypertrophy of an organization of professional corporations (Ibbetson, Nesfield), 3. that it originates from an ancient clan organization (Senart). Of late there has been a tendency to attach greater importance to pre-Aryan culture. This was not so obvious at the time when the above mentioned scholars wrote (between 1883 and 1896). It stands to reason that also the pre-Aryan population knew some form of social organization, but there is not much known about it at present. Also in this case we feel little inclined, therefore, to ascertain the influence of the pre-Aryan substratum, not because we deny the influence of this substratum, but because it should first be studied what really may be found in Indian culture taken as a whole. To take a parallel case, when in a linguistic study one starts with eliminating what is owing to the influence of anolder substratumbefore one has considered the language in question as a whole, one begins at the wrong side. And the more so in a case like this, where the whole substratum is all but unknown. It is at any rate wrong to suppose that the pre-Aryan population should have taken over the social structure from the Aryans, or inversely, the Aryans from the pre-Aryans. For such an adoption there is much required, for adopting a different type of organization is not something like putting on another coat. address one another in terms of relationship in which this relation is expressed. These terms may be used to express that one knows a person as belonging to one's potential husbands or wives. Very often a term in which a man addresses his father's sister's — and his mother's brother's daughter is at the same time the term he uses to refer to the women of a clan with which his clan stands in connubial relation and whom he would be allowed to marry. In those cases both these cousins belong to the native clan of his wife and the terminology of relationship is not individual but classificatory. An equivalent for „father", for instance, not only refers to the generator, but to all the potential husbands of the mother, usually all the indicated men of the same generation from the clan, into which the mother would be allowed to marry in virtue of her clan membership. Nor is the tribe in the first place to be defined as an endogamous unit; the tribe is the representative of the totality. In common cases the tribe indicates the limits, within which clan relations exist. Within the tribe the clans stand in a definite connubial relation with each other. Consequently people marry within the limits of the tribe, because they do not stand in relation with clans outside the tribe. It is wrong to think that in a primitive organization nearly every marriage can be contracted, provided it is contracted outside the exogamous clan and within the endogamous tribe. Indeed, in a clan system marriage possibilities are often very limited, because there is a connubial relation with special clans. There are fewer possibilities than with kinship exogamy, where the first requirement is to marry outside definite degrees of consanguinity. It is indeed characteristic of a clan system that often persons who are closely related with respect to kinship are one another's potential husbands or wives. Now the Indian Sapinda exogamy is not conceivable in one of the clan systems that are known at present. If caste could simply be put on a level with tribe, the exogamy of the family nucleus would have to be put on a level with what Senart calls „familie" (also where a clan organization is found, no marriages are contracted that are incestuous in the limited sense of the word). Senart is not very clear on this point, for he writes about the sapinda exogamy: „II n'intéresse qu'indirectement la question qui nous préoccupe." 1) This cannot be right, for sapinda exogamy is kinship exogaray, which becomes prominent by the side of gotra exogamy, so that Sapinda exogamy should be put on a level with the exogamy of the bilateral family group. In an ordinary clan organization the children of two brothers belong to one and the same clan. The same is the case with the children of two sisters (because two brothers or two sisters should always marry within the same clan) it is immaterial whether this is their mother's clan (in the case of matrilineal descent) or their father's (in the case of patrilineal descent). The children of a brother and a sister, however, necessarily belong to two different clans, because they belong either respectively to the brother's clan and that of the sister's husband, or to the sister's clan and that of the brother's wife, accordingly as descent is patrilineal or matrilineal, Also the children of a mother's brother belong therefore to a clan different from that of the mother or the mother's sister; similarly the children of a father's sister belong to a clan different from that of the father or the father's brother. If a man is obliged, according to the connubial relations between clans, always to marry within the same clan in which the father married, i.e. the mother's clan, it will be clear that the potential wives are always the daughters of his mother's brothers (thistermagaininaclassificatory sense). Now the Sapinda exogamy prohibits a marriage to all relatives to the fifth degree, whether related through a man or a woman. This includes nearly all conceivable degrees of kinship. In this way a clan system cannot exist. At any rate we know of no clan system that would give room to a kinship exogamy carried to this length. Besides it should be kept in mind that, usually, a tribe does not count many thousands of relatives. This would decidedly be necessary in the case of Sapinda exogamy, for it has been calculated that exactly 2121 potential wives would consequently be excluded from marriage to a man and an equal number of potential husbands to a woman, if we assume that every married couple has one son and one daughter.2) This would in itself already be sufficiënt to prove 1) op* cit. p. 36. _ 2) V. N. Mandlik-Vyavahara Mayukha p. 352. that genuine clan exogamy cannot have existed side by side with Sapinda exogamy. Moreover, in the tribe the exogamy of the clan prevails over the endogamy of the tribe, whereas in the caste the endogamy of the caste prevails over the exogamy of the gotra. Besides this, the hierarchic organization of the castes has not at all been explained by Senart. It is true, for the improvement of this weak point scholars have tried to look upon the exclusivism of the priestly group as a rennet that curdled as it were the whole of Indian society into castes, but this conception is of too mystic simplicity. One may indeed assert that owing to the influence of the Brahman caste all other groups have arranged themselves „gradually" and „spontaneously" into a hierarchic organization, but the qualifications „gradually" and „spontaneously" do not deepen our insight. The tribe is the exponent of totality and the basis of the system of classification.1) Inside the tribe clan-membership usually means the possesion of a certain status. One instance by way of illustration. The Omaha Indians (North America) are divided into two phratries: i. that of the Sky people; 2. that of the Earth people. These phratries, the two exogamous tribe halves, are again subdivided into clans. The clans of the Sky-phratry had to look after the functioning of the primary cos mie powers. They were entrusted with the ritual. To this phratry for instance belonged the Sky-clan, which was in charge of the first initiation of a child some eight days after its birth. Another clan, being called „Head of the deer", was probably entrusted with the ritual that was connected with the night-sky, also bearing on the myth of creation. The clans of the Earthphratry had the more social charges of mankind on earth. There was for instance among them the clan „By whom to become angry" which was charged with the ritual of war. From this instance it may be seen that the clans are arranged into a definite hierarchic organization and that some welldefined functions are at the same time divided among them. In the tribal organization the status relation is at once expressed. The Heaven-phratry, entrusted with spiritual things was held in higher esteem than the other phratry that was in charge of the temporal concerns of human welfare. *) See p. 114. A person did not owe his status to the fact that he was a member of some tribe or other, but of a certain clan. With other tribes one had little to do. Seeing things from this standpoint one mightsaythatcasteis the representative of the clan, rather than of the tribe. The organization of castes in India resembles that of clans in a genuine clan system, but does not resemble a intertribal organization (or a confederacy of tribes), which is rauch less usual as a system complete in itself. It is exactly due to the tribe being the exponent of totality that it does not simply take the function that the clan has already within the tribe, when the organization is extended so as to include a number of tribes. Where the clan organization is deteriorating, one mostly sees that a new organization, whatever its character may be, anew takes the tribe for a pattern. The different tribes do not simply overlap but are merged into the new tribal whole. Hence it is not so easy to look upon caste organization simply as a grouping of tribal units. So the caste in it self cannot represent the tribe because the gotra-and sapinda exogamy cannot be exactly the same thing as the exogamy of the clan and the family nucleus in the tribe. Nor is anything known to us of a fixed connubial relation between the gotras within the caste. Also outwardly the classification of the various castes cannot have had a confederacy of tribes for prototype, because such an organization of tribal units is in the first place uncommon, and secondly a possible organization of tribes would in the long run not suffer the tribal units to exist in their old form, which, however, in case of the caste should be accepted. It is not only the connubial relations that are essential to the determination of a man's caste prerogatives. In daily life also certain food-taboos play a very important part. There are precise rules to prescribe the sorts of meat and drink one is allowed to accept and from which caste one may take them. „In Hindustan proper, castes can be divided into five groups; first, the twice-born castes; second, those castes at whose hands the twice-born can take „Pakka" food; third, those castes at whose hands the twice-born cannot accept any kind of food, but may take water; fourth, castes that are not intouchable, yet are such that water from them cannot be used by the twice-born; last come all those castes whose touch défilés not only the twice-born but any orthodox Hindu."1) This food community has rightly been associated with the food community that exists in a clan organization among persons who could partake of one and the same sacrificial repast. „In early times a sacrifice was the occasion for every important gathering or festivity, as is shown both in Indian history and legend. And the caste feasts.... seem to be the continuation and modern form of the ancient sacrifice".2) Now it stands to reason that the food community in a primitive organization is determined by c 1 a n membership. It is the members of a clan that give expression to their kinship in the collective communion. Consequently, if, in this respect, one wants to find in a clan system a prototype of the modern food community among the members of a caste one does not arrivé at the tribe again but at the clan. Does this dispose of Senart's theory ? In our opinion it does not. The relation with an old clan organization — no matter whether the caste must be associated with the clan or with the tribe — appeals too strongly to justify a denial of Senart's theory. Exogamy and endogamy undoubtedly point to a form of organization, determined by marriage possibilities. That he is not on the wrong track is proved by the meaning of the word „kula". „Die Verwendung von „Kula", das in der Regel „Familie" bedeutet, im Sinne von „Kaste", zeigt, wie sehr die beiden Begriffe für das Bewustsein der Inder ineinander übergingen, und wie eng sie in der That miteinander verbunden waren".3) Let us listen to Senart himself: „Par la notion de la parenté qui la pénètre, par la juridiction qui y règle assez tyranniquement la vie privée, marriage, nouriture, usages cérémoniels, par la pratique habituelle de certains cultes particuliers, par son organisation corporative, la caste rapelle en effect le groupe familial tel qu'on 1'entrevoit h ses degrés divers, dans la familie, la gens, la tribu,"4) We see that Senart is rather vague when he has to find out whether the tribe or the clan has actually been the prototype *) G. S. Ghurye-Caste and Race in India. London 1932 p. 7. 2) Russell (see p. 51 n. 2) — I p. 182. 3) Fick — op. cit. p. 22 n. 4. ') Senart-op. cit. p. 207. 4 of the caste. The eating together, according to many the only criterion of the caste, decidedly points to the clan and not to the tribe. The clan is the centre of the feeling of solidarity, still more than the tribe. A hierarchic grouping with a distribution of functions may find its pattern in an organi^ation of clans, but not in one of tribes. Everything points to the clan being the prototype of the caste and not the tribe. Still it is quite natural that Senart should make the tribe and not the clan the example of the caste, for the caste is endogamous, whereas the clan is exogamous. This is the root of the matter. Dahlmann has nearly approached this point in his criticism. We shall now try to deal more fully with this matter in order to elaborate Senart's theory on this point and to give it a firmer foundation. * * * We take Senart's example and begin with modern data, not in order to give a more complete survey of the social structure, but because older data are so scanty that they allow of all kinds of interpretations. Modern times may shed some light on the older periods, however dim it may be. First we shall try to state something about the clan grouping in India, in other words about the regulation of marriage. We have to lament the fact that the ethnography of India is not nearly what it should be.Hermann Niggemeyer rightly complains that the totemism of Australia is better known than that of India. It is also a great drawback that in the literature a discouraging amount of terms is used whose meaning is not always clear. Words like family, sept, section, clan, sib, caste, sub-caste, tribe, group etc. and a whole pattern-card of native terms make some publications into intricate cross-word puzzles. In his „Totemism and Exogamy" Frazer comes to the following conclusion for India: „Totemism and exogamy appear to be or to have been in former times universally diffused among the Dravidians.... The Dravidians combine as usual, totemism and exogamy with the classificatory system of relationship." x) In the North of India conditions are dif- x) P- 339- SOCIAL ORGANIZATION ferent. There exogamy is not lacking but totemism is less known.1) In the South of India the organization is more antique than in the North. Here the caste organization is carried further, totemism is also more obvious, there live more groups, here, that still know a genuine clan organization. So we shall first mention South Indian data. Moreover here we find Russell's fine and comprehensive work at our disposal, a we 11edited critical collection of materials that can inform us on many topics.2) An interesting ilJustration of the fact that genuine clan exogamy still occurs at least in the Central Provinces, may be found in the myth of creation of the Audhelias. According to this myth there were in the beginning three brothers. Out of the incestuous marriage of one of them with his brother's daughter rose the „gotras or exogamous divisions" as they are at present known for this caste. Russell adds: — and this is the gist of our remark — „A curious point about the story is that the incestuous nature of the connection is n o t taken to be the most pressing objection to the marriage of (the uncle) with his own niece, but the violation of the caste rule prohibiting marriage within the same gotra." 3) Taken as a whole the organization all over India is patrilineal. A person's social status largely depends on that of his father. There are two areas where matrilineal organization is found, viz. in Assam (among the Khasi, Garo, etc.) and in the second place in a quite different corner of India: the Kanarian district. Now it is a curious phenomenon that these areas are not organized according to unilateral principles. Matri- 1) We shall not go further into the matter of totemism in India. We refer to Niggemeyer's essay. We must confess that we are not always sure of totemism north of the Vindhya-mountains only being found among Dravidian groups. In order to establish the race the totemistic traits should not be used in the first place. We would indeed be more surprised if the available data could not be taken as indications of totemism than if the reverse were the case. 2) R. V. Russell assisted by Rai Bahadur Hira Lal. — The tribes and Castes of the Central Provinces of India-London 1916 (4 volumes). We once more remind our readers that our discussion of social organization in India was not meant, as an exhaustive study, but only to prevent our argument from being called strongly deductive. 3) Russell II p. 47. We do not believe this to be a reflect onofhistorical incestuous relations. This mythical incest (marriage between brother and sister, for instance) is often the sacral explanation of the relations between the phratries. and patrilineal systems are interwoven with one another. For the area in Assam Hodson concludes: „From these facts I deduce the conclusion that on the margin of the Khasi or matrilineal area we have a social organization which is or was patrilineal."1) And for the second area Niggemeyer states that it may be divided into several zones. There is in the first place a region with matrilineal clan membership and matrilineal inheritance. „Man könnte hier von vollmutterrechtlich-totemistischen Kasten sprechen." Secondly there is a region with patrilineal clan membership and patrilineal inheritance. „Es sind die reinen vaterrechtlich-totemistischen Gruppen, die sich als solche an das übrige Indien anschlieszen." Then, there is a region in between, where all sorts of blends occur. Here clan membership is often matrilineal and inheritance in the same group is patrilineal. 2) Matrilineal organization, as it seems, is not quite foreign to the Indian social structure, not only in these two areas but in general. Senart says: „On peut dire que partout il est interdit de se marier dans le gotra dont on por te le nom, dans le gotra paternel par conséquent. Mais cette interdiction n'épuise pas les empechemens légaux. La règle ordinaire est qu'un homme ne peut se marier davantage dans le gotra de sa mère; ni souvent dans celui de la mère de son père, ni quelquefois dans le clan de la mère de sa mère. L'exogamie du cóté maternel est d'une portée trés variable. On cite des castes ou tribus qui k cöté des gotras et au-dessous d'eux, connaissent des groupemens plus petits institués, semble-t-il, pour servir de cadre k l'exogamie du cöté maternel".3) Unfortunately we have no detailed information about the influence of this matrilineal organization. We shall mention a few data that drew our attention in our reading-matter. A Kamma man does not marry within his own, but marries by preference in his mother's gotra, and is obliged, when the case arises, to marry his mother's brother's daughter.4) The !) Notes on the marriage of cousins in India — Man in India V p. 163 seq. 2) See the instructive Schedule by H. Niggemeyer-Totemismus in Vorder-Indien. Anthropos XXVIII/1933 p. 441. 3) op. cit. p. 35. 4) Russell I p. 373. Chasas are divided into „got" or „septs" and into „varga" or „families". A man does not marry within his own „varga", but he usually does in that of his mother.1) A Bhil, however, will not marry within the family of his mother or maternal grandmother. In this case the marriage of first cousins would be prohibited. Russell draws the same conclusion, but at the same time he remarks that among the Mohammedan Bhils marriage with a mother's brother's daughter is known.2) This is not likely to spring from the requirements of the Mohammedan Fikh. So to a Bhil marriage within the maternal clan is at the same time prohibited and allowed. In so far he is in the same position as the Maratha 3) and the Kunbi4). To these marriage within the maternal clan is prohibited, whereas marriage with the mother's brother's daughter is recommendable. The mother's brother's daughter necessarily belongs to the same clan as the mother, because the latter and her brother bear her father's clan name. Russel goes on to say that the Mahar, who also knows m.b.d.5) marriage does marry within his mother's or paternal grandmother's sept.6) It will be seen that in the Central Provinces in one case marriage within the maternal clan is not allowed, whereas in the other case it is allowed. Indeed there are even cases in which marriage within the maternal clan is prohibited and at the same time prescribed. A closer examination would certainly add to the instances we have given. If we might suppose for a moment that there would exist a matrilineal system in the above mentioned cases, my m.b.d. would belong to a clan different from my mother's, for the m.b.d. would not belong to her father's clan (to which also her father's sister — my mother — belongs) but to her mother's clan. But there is a patrilineal system. Now, we propose the following hypothesis: There exists in India a patrilineal system of relationship, intersected by a latent matrilineal one. For in this case one may say that marriage 1) ib. II p. 426. 2) ib. II p. 287. 3) ib. IV p. 203. 4) ib. IV p. 22. 6) This abbreviation will further be given as an indication of the daughter of the mother's brother. 6) ib. IV p. 133. within the (matrilineal) clan of the mother is prohibited and allowed at the same time within the (patrilineal) clan of the mother. With this system of relationship each individual belongs to two clans, viz. to his father's patrilineal and to his mother's matrilineal clan.1) It is not impossible that also in other cases the division into a doublé set of clans, so inconceivable at first sight, should be explained in this way. The Parwar Bania are divided into twelve „gotra", which are again subdivided into twelve „mul". A Parwar Bania is not allowed to marry within his own „gotra", nor within his mother's, his (paternal or maternal?) grand- and greatgrandmother's „mul".2) If our conjecture is right the „gotra" was the patrilineal and the „mul" the matrilineal group there. In each „gotra" all „mul" are represented (as in the schema within the patrilineal clan i, the matrilineal clans 1) An illustrative plan. The = mark indicates marriage relations. To the left of the = mark, are the men, on the right side the women. The lines point to descent and brother-sister-relation. With the figures we mean patrilineal clans. A man from clan i (ego) marries, in obedience to the demands of exogamy, outside his clan, so, with a clan 2. Three generations are inaicatea. Generation of grandfather ..1=2 2 — 3 1 r 1 „ ,, father . . . T J 1 ' ego 1=2 2 — 3 It will be seen that it is impossible for ego not to marry into his mother's clan, and yet, with his m.b.d. Mother and m.b.d. of ego from clan 1, both belong to clan 2. We now give the same plan. The patrilineal grouping is now crossed by a matrilineal. So every man has the matrilineal clan of his mother (indicated by a character) and the patrilineal clan of his father (indicatedby a figure). 1 '' F~J A Generation of grandfather iC = 2D 2D -y 3A I I ' „ „ father . . iD = 2A 2A = H 1 1-1 „ „ ego . . . iA = 2B 2B - 3C Now it is sensible to forbid the marriage into the mother's clan, of ego (so, into clan A of ego) and yet to permit the marrying into the patrilineal clan of mother (so, clan 2) so, the marriage of m.b.d. The m.b.d. is 2B then. 2) ib. II p. 121, 158. A, B and C have been placed). It is, therefore, quite natural that such a latent matrilineal tendency should not be noticed because the matrilineal clans show themselves in this largely patrilineal grouping as subdivisions of the patrilineal clan. Another instance: The Gahoi Bania are divided into twelve „gotra" and seventy-two „al", which are subsections of the „gotras". A man does not marry within his own „gotra", or within the „al" of his mother or grandmother. Crooke mentions that the Gahoi does not marry within his own „gotra" or in the „al" of his mother's brother, father's mother's brother or mother's mother's brother.x) If we take the „al" to correspond with the matrilineal clan, and the „gotra" with the patrilineal clan, it is easy to see from the scheme that these seemingly foolish and complicated marriage prohibitions are really rather simple. Of course these remarks do not really prove our hypothesis. A special investigationbecomesnecessary. It isverylikelythata bilineal system, as we think to find here, is not due to occasional local circumstances. We expect bilineal descent is or was to be found all over India. In most places the matrilineal system is latent, in two places the patrilineal system is latent. These two areas are too distant from one another to allow of the hypothesis that there was an invasion of a group of population with a matrilineal orgamzation. In the conception we have defended there is room for these cases. They are not absolutely foreign elements, which, in an unexplainable way have invaded the country, quite independent of one another. They are only cases in which the generally latent matrilineal tendency has become obvious, whereas the patrilineal system is here latent, but still clearly perceptible. The tendency, somstimes still met with in ethnology, to presume a purely matrilineal phase of social organization, previous to the purely patrilineal system, is really a remnant of too strongly evolutional conceptions, which prevailed in the nineteenth century. For those cases, in which a matrilineal and a patrilineal system are mixed together, an explanation is sought in the direction of finding the origin of what is considered to be a blend. For instance one tries to eliminate the matrilineal characteristics *) ib. II p. 146. of a system. Matrilineal elements, however, if found somewhere, should not be eliminated a priori as foreign introductions. We, for ourselves, are of opinion that bilineal characteristics are normal for a primitive organization. At any rate, whatever may be the explanation of these bilineal tendencies, in India they have been noticed more or less obviously in many places1). For the rest, the question of the origin of the bilineal grouping need not interest us here. It will do for us just to have stated the phenomenon. It suits our conception of culture that we do not take this bilineal system in the same way as an interpolation in a text but as a constituent element of the culture where it is found. The second question we want to discuss is that of the marriage possibilities. Also with respect to this point India, as a whole, is not a homogeneous area of civilization. In the South the organization, as has been stated above, is more ancient than in the North. The most conservative groups are those of the more remote regions. Now in a clan-organization a very common type of marriage regulation is the cross-cousin marriage, where a man marries his mother's brother's daughter, or his father s sister's daughter. The two clans are with one another in reciprocal connubial relation, they exchange brothers and sisters.2) In 1907 Dr. W. H. R. Rivers wrote an essay on the relation between cousins, a very important element in a reciprocal marriage system.3) As it was known to him that a marriage with the m.b.d. is very frequent in India, Rivers examined *) The doublé clansystem has been studied for the Ashanti, the Herero and some Kongo-peoples in Africa. Consult works of R. S. Rattray; H. Vedder; E. Torday (Journ. Anthr. Inst. LVIII) and H. G. Luttig (The religious system and social organization of the Herero). *) r 1 , —1 Generation of grandfather ..1=2 2 = 1 I 1 , ■ I ~1 „ „ father ....1=2 2 = 1 ■ I U „ego i = 4 2 = i The mother's brother's daughter, who is identical with the father's sister's daughter (because the father's sister and the mother's brother, in obedience to the brother-sister-exchange, marry each other) is appointed wife. Mind that the terminology is classificatory. Consequently the marriage is possible with the appointed members of the father's sister s daughter's clan, not merely with the real father's sister's daughter. 3) J. R. A. S. 1907 p. 611 The marriage of cousins. whether also the f.s.d. marriage occurred, because in the reciprocal marriage system the f.s.d. who is the same as the m.b.d. (see the scheme) is the proper wife. In that case the common cross-cousin system would exist in India. And this would become evident in the terminology of relationship, e.g. in the fact that the term for father's sister's husband and that for mother's brother would be the same. This did not come right for by far the greater part of India, so that the frequency of a marriage, exclusively with one c.c. (the m.b.d.) remained for the moment incomprehensible. The important place that is assigned to the mother's brother in India, is, according to Rivers, a consequence of matriarchy. In the case of matriarchy a person does not get his clan name from his real father, but from his mother's brother, who also figures as father. The correlation between matriarchy and m.b.d. marriage as accepted by Rivers, is not exactly in accordance with the facts, because Rivers found that m.b.d. marriage was spread much wider than matriarchy. There are regions where the genuine c.c. marriage occurs e.g. among the Todas, who were thoroughly studied by Rivers, and among other groups that may be found in Frazer's „Totemism and Exogamy". But on the whole the following rule holds good: „In most of the castes of South India the most suitable bride for a boy is considered to be the maternal uncle's daughter."1) The normal type of marriage in South India is consequently the exclusive c.c. marriage, where the marriage with the m.b.d. is allowed and that with the paternal aunt's daughter is prohibited. Then, (owing to the demands of clan exogamy with its unilineal descent) a marriage with the daughter of the paternal uncle is prohibited and very strictly that with the daughter of 1) Fawcett-Man p. 37 1901 Russell-passim. Dubois (Manners and Customs p. 20) mentions „several strange and ridiculous customs", which are observed in marrying. „A brother's children may marry a sister's children, but the children of two brothers or of two sisters may not intermarry. A first cousin marries his first cousin." Dubois does not mention the exclusive c.c. marriage in so many words. He only writes: — „Among descendants from the same stock the male line always has the right of contracting marriage with the female line; but the children of the same line may never intermarry." Finally he also touches upon the sister's daughter's marriage: „An uncle may marry the daughter of his sister, but in no case may he marry the daughter of his brother. the maternal aunt1). One of the most striking features of the exclusive c.c. marriage has been thus formulated by Hodson: „An interesting feature about the single cousin marriage system is that it is capable of almost indefinite expansion according to this formula: (A + b) (B -(- c) (C+d)*>** [(R 2n) + (r — n)] [(R — n) + r] (R + a). By this scheme a man's son may not marry the daughter of his father's sister. It therefore requires as a minimum basis three exogamous divisions."2) In these words Hodson has put down what the system comes to. For the exclusive c.c. system is circulative. The number of participant clans is capable of almost infinite expansion, as Hodson rightly observes. Only with less than three clans is a circulative system impossible, because with two clans there must always be a reciprocal connubium between those two, so exchange of brother-sister (i.e. also the marriage with the daughter of the father's sister3). Hodson is quite right to connect the circulative marriage system with the myths of origin of the clans, in which it is told 1) F. T. Richards-Cross-Cousin Marriage in South India-Man XIV p. 1Q4. Richards continues: „His paternal aunt's daughter is next in favour," but adds (for the Komat Vaisyas): — „But if the sister should have a daughter, and a brother a son, (this is therefore, the father s sister's daughter's marriage), the sister is not obliged togive her daughter to the nephew, but may give her to whom she pleases. Richards is of opinion, that the c.c. marriage serves to enable the father to leave his property to his son in case of matrilineal inheritance. In case of matrilineal inheritance the property of the brother goes to his sister s children. If therefore the sister's child marries the brother s child, the brother s own child gets the property. This, however, does not explain, why exclusively a brother's daughter may marry a sister's son, which is, indeed, the case with the exclusively c.c. marriage. , 2) Notes on the marriage of cousins in India. Man in India V p. 103. 3) Hodson's meaning may be sketched as follows: Next another sketch from which it appears that only the marriage with the m.b.d. (and not that with the f.s.d.) is possible. Again three eenerations have been given. For clearness' sake we moreover mark tne man (to the left of the = mark) with a □ : the wife (to the right ot = with an O that the tribe has sprung from three brothers (or married couples etc.). In spite of this Hodson unfortunately has not fully appreciated the exclusive c.c. marriage. He apparently considered it as a sort of blend. It is not quite clear to us what he means by the following remark: „This scheme ignores all ties through female kin", for with the three-clan system kinship may as easily be derived patrilineally as matrilineally. Nor is the exclusive c.c. system an occasional Indian specialization, which outside India is only found among the Gilyaks in Siberia. On the contrary the system is rather common and will on closer examination on the spot probably turn out to have a greater extension than is now known. It has further been described for Melanesia, Further India and Australia, and it undoubtedly exists also in the eastern part of Indonesia. There are in the literature a few points that become clearer by the hypothesis of a circulative marriage system. The often repeated prohibition to exchange girls between two families is a denial of the reciprocity of the marriage relations, which naturally reminds of a one-sided marriage relation. The same aversion to the exchange of brother and sister is also found in the North of India. It is obvious that also there we look upon it as an indication for the existence of an exclusive c.c. system (now for the greater part archaic). In the North, as has been mentioned above, the organization is on the whole less antique than in the South. Clan exogamy has here been replaced by kinship exogamy to greater extent already.1) Still the m.b.d. marriage repeatedly occurs here.2) Clanship, the clan feeling of solidarity, is apparently not forgotten at all for: „Alliances ') How far this process has advanced, we do not know. According to K. P. Chattopadhyaya, the kinship terminology is still classificatory, which points, therefore, to clan-exogamy. (Man XXII p. 36). 2) See W. Crooke — The Tribes and Castes of theNorthWestern Provinces and Oudh. Calcutta 1896, IV vol. are made not on the ground of social position or wealth, but because the parties fulfil the conditions of descent, which make union between them possible. A rich land-holder naturally seeks a wealthy bride for his son; but from the scarcity of brides and complications in regard to prohibited degrees, he may have to marry him in a family of low position, and provide the marriage expenses himself. When the marriage day comes he finds the friends of the bride just as conscious of their own dignity arising from descent, and just as sensible and punctilious as he would be himself."1) In the North, where clan exogamy has been ousted by kinship exogamy, (so that the system there is more similar to the West-European system) marriage with the m.b.d. is very often strictly prohibited.2) This cannot be due to chance, but must be the expression of a tendency towards marriage in the prohibited degree. It is curious that in the marriage prohibitions sometimes all kinds of queer degrees of relationship are prohibited, which can only logically be accounted for in a threeclan-system. Otherwise what a strange arbitrariness must be considered a prohibition of the following degrees: daughter of mother's brother, father's mother's brother, mother's mother's brother, father's father's mother's brother, father's mother's mother's brother, mother's father's mother's brother, mother's mother's mother's brother?3) In North India the maternal uncle often comes into prominence as officiant of certain ceremonies e.g. those of marriage, a phenomenon that has been signaled as far as in the Panjab.4) Then there is another remarkable phenomenon, which, in our opinion, has to fit in with the marriage relation we have indicated, namely the so-called hypergamy, the preference for a bridegroom of a higher group than the bride. Seen from the man's point of view this means consequently the marriage with a woman of a lower rank than his own.5) This phenomenon is very ancient, so that some scholars (Risley, Bouglé) try to *) N. Crooke-Natives of Northern India p. 179. 2) Crooke-Tribes and Castes, Passim. 3) Crooke-ib. III p. 131, 2255 IV p. 127. 4) Crooke-ib. III p. 331, 408, 421 I p. ccvi, and H. M. Rose-The Mother's Brother-Man 1902. 5) The data are frequently given, e.g. by W* Crooke; oir ri. Kisiey; R. V. Russell; C. Bouglé. account for it by the circumstances of the time when Aryan immigrants invaded India, Scarcity of women prevailed among these Aryans, so that they were obliged to find their wives among the native population. When they had once got these women they, however, refused to give their own in exchange to the native groups but continued themselves to choose their wives from them.*) This hypothesis does not seem plausible to us. The fact that especially the higher groups (for there hypergamy is most common) were in this habit and kept it up to this day, makes this solution very unlikely. It seems Senart did not know what to do with it. Hypergamy, which connects not only sub-castes but also entirely separate castes, fits badly in with Senart's conception of the endogamous caste (which he thought to have had the endogamous tribe for a prototype). He simply calls this phenomenon „bizarrerie". 2) Crooke wants to account for the aversion to a marriage of a high-caste woman with a lower situated man by the anxiety of a degradation of status, which such a marriage means to a woman, as the wife follows her husband's status. 3) Russell searches deeper to find a reason in the mentality of the archaic marriage by capture. A group has to part with a woman and feels this as a defeat. This feeling they try to compensate, if possible, by marrying the girl into a higher clan. 4) However, the phenomenon is more than mere „bizarrarie". When the periodical „Man in India" was set up, Rivers mentioned hypergamy as the first point on which further data have to be collected.5) So when we spend so many words on this subject in the present short survey, it is not an endeavour to magnify an unimportant phenomenon with which to pad the gaps in our hypothesis. Hypergamy, in our opinion, is a phenomenon that accompanies the circulative system. As has been stated, patrilineal grouping prevails in India. The male sex is of paramount importance in the organization. Also everywhere in religion all that is classified as male predominates. In so far Sarat Chandra Roy has found the right explana- 1) Bouglé-op. cit. p. 123. 2) Senart-op. cit. p. 31. 3) Crooke-op. cit. I p. 305. 4) Russell I p. 151. 6) Rivers-Kinship and Marriage in India. Man in India I p. 6. tion for hypergamy in our opinion.1) He rightly points out that it is partly in religion that an explanation is to be found of the caste phenomena, though, perhaps, he does not mean exactly the same thing as we do. The conception of the superiority of the male (purusha) certainly plays a part. Only, we do not agree with him that people married Dravidian women on account of this superiority. Then they had better stick to Aryan women, for through the marriage of an Aryan man with an Aryan woman this superiority would have been shown more clearly than by a marriage of an Aryan man with a non-Aryan woman. And in general one might say, as a matter of fact, that the marriage with a lower-placed woman could hardly contribute to show the man's superiority to woman. Russell rightly advanced the fact that especially the higher classes, Brahmins 2 ) and Rajpüts put hypergamy to practise as an argument against the explanation of caste phenomena as resulting from race antithesis.3) In the Indian classification the female side is inferior to the male. This inferiority is expressed in the terminology of kinship by using the terms for „wife's brother and „father-in-law as invectives.4) From the diagram it may be seen that every clan provided women for one clan and received them from another. This is different from a reciprocal relation, where this kind of superiority could not be thought of, because the women are exchanged and consequently a clan after taking, as the „male" clan, women from one clan figures as the „female" clan in its turn with respect to the „male" clan to which women are given. In a circulative system, however, there arises a gradual scale of status relations among the clans. The clan to which women are given is higher than the one that gives them. The clan from which one gets women is lower than one's own clan. Clan i is higher than 1) Roy-Caste, Race and Religion in India. Man in India XIV p. 39 2) ^The phenomenon is also calles Kulinism, because the Kulin Brahmins exercised marrying as a profitable trade, the drawbacks 01 which were neutralized by the dowries. , 3) „If caste was based on distinctions of race, then apparently tne practice of hypergamy would be objectionable, because it would destroy the different racial classes." Russell — op. cit. II p. 363. *) Russell I p. 29. SOCIAL ORGANIZATION clan 2, clan 2 than 3 and clan 3 again higher than clan i.1) The phenomena described here and taken together: aversion from the exchange of brother against sister, from marriage with the m.b.d. and some m.b.d. of other generations and hypergamy, justify the conclusion that behind North Indian kinship exogamy there must be thought of a circulative clan system, of which the clan idea has not by any means disappeared entirely. Roughly speaking India shows a circulative, probably bilineal system of kinship in the South, which in the North has been replaced by a marriage system with kinship exogamy, with here and there a c.c. type with exchange of brother-sister. There is often another important element in a clan organization, which has only been incidentally mentioned, viz. the division of the whole tribe into two exogamous phratries. If there are no more than 3 clans there can naturally be no room for a real division into two. But there is one if there are four or a greater number of clans (provided itisanevennumber)2). *) If there is no gap in the system, there is of course, in the last case, no question of hypergamy, but on the contrary, of hypogamy. See lower down p 85. 2) For 4 clans the plan for three generations is: r= '3 $ = *——ftl 7 s rJ ^ snL s Hl 'i=s rt~[ Now the clans 1 and 3 form the one, and the clans 2 and 4 the other phratrie. This may be seen in the following sketch: □ man O wife .... brother-sister marriage. It is clear that clan 1 and 3, taken as a whole, form an exogamous group. By this division into four, the division into three does not become ïllusory again, for each clan regularly forms a three-group with two others. Father's sister's chiid Ego Mother's brother's child 4 1 2 By the side of the division into three there is in India also a division into two, though perhaps not so evident there, as in a complete c.c. system. Real phratries are apparently found among the Janappan (Madras), Korava (Arcot) and Bili Maggi (Mysore)1). Less known exogamous divisions into two have also been observed among the Sansia and the Gond.2) With the latter the dual organization went perhaps together with an m.b.d. type of marriage. According to Russell this type originally occurred most frequently with the Gonds whereas the f.s.d. marriage did not become in vogue until later times.3) Alongside with these cases of real dualism is a pronounced tendency to a division of castes etc. into two groups, which are at least suggestive of a dual phratry-organization. The numerous cases in which castes are divided into a Northern and a Southern group, or an Inner and an Outer, a Right and a Left, a Great and a Small, a Genuine and a False, a Siva and a Parvatï group immediately remind one, who has the slightest acquaintance with clan organizations, of the wellknown phratry contrasts, which are also taken as two complements (of husband and wife, right and left etc.) forming together a unity (the tribe). Hence Niggemeyer comes to the following conclusion: „Die Einteilung einer Kaste oder eines Stammes in zwei Hauptgruppen findet sich in Indien allerdings nicht selten". The question, however, whether these dual organizations justify the assumption of a phratry system is answered in the negative by Niggemeyer; firstly because these dual divisions are not attended by exogamy, which is decidedly required for a real division into phratries; secondly because this dualism is not correlated with the „Mutterrecht". We must dispose of the latter argument because this correlation is chiefly founded on the idea of a „Mutterrechtskulturkreise", which is for us just as problematic as the „Kulturschichtung", of which it is a part. Niggemeyer thinks that the dual division in question sprung up, because the children of endogamous marriages within the caste put themselves over against children of mixed marriages. 1) Niggemeyer-loc. cit. p. 421. 2) Russell IV p. 489; III p. 64. ») Russell. III p. 71. He derives this from the names „Genuine" over against „False", or „Pure" over against „Mixed". Presently we shall have an opportunity to point out that from the breaking up of the clan organization not only two but many more blends come into existence. What Niggemeyer further brings in to explain the connection between real dualism in India and patrilineal organization is merely a wild guess, hopelessly at war with the facts. The local-historical explanations he gives are insufficiënt. If it were not that the dual divisions in question are not exogamous, the connection with the real phratry-division would be luce clarius. Below we shall try to prove the connection with the phratry-division also in this respect.1) On another point, again, there is direct resemblance with a phratry-division: viz. the characteristic rivalry between the two groups, which in some of the divisions under consideration too are by no means lacking. „Est-ce k la persistance des souvenirs, des inimitiés qu'elles (sc. les variétés d'origine et de race) éveillent que se doivent ramener les dissensions qui en maints endroits se perpétuent entre castes diverses? Elles frappent d'autant plus que cette population est naturellement plus pacifique. L'hostilité la plus durable, la plus fameuse, est celle qui, dans le sud de 1'Inde, partage ce qu'on appelle la „main droite" et la „main gauche". Les deux catégories correspondent, semble-t-il, au moins en gros, h une répartition en castes d'artisans et castes agricoles. L'origine et 1'histoire n'en ont jamais pu être éclaircies. Ce qui est sur, c'est que leur rivalité a été et est encore la source des conflits violens qui divisent la population en camps ennemis. Certains privilèges que revendique 1'une ou 1'autre „main", au moindre empiétement, allument la lutte. Elle a souvent provoqué des soulèvemens qui, „se communiquant de proche en proche, semaient le trouble sur une grande é tendue de pays, donnaient occasion h des excès de tout genre et se terminaient fréquemment par des batailles sanglantes". Des faits pareils, quoique plus circonsrits, sont signalée en bien des régions. Souvent ce sont des prétentions rivales k des avantages honorifiques, qui, cause ou prétexte, donnent naissance k ces conflits".2) One need not *) See p. 86. 2) Senart-op. cit. p. 30. 5 think of race antagonism in these cases simply because for the greater part there is no difference in race between the groups. One might rather think of phratry-antagonism.1) Dual divisions of this kind are not only found in the South butalso in the North. Among the Rajpüts, for instance, there is a division into two really rather insignificant groups that have the characteristic natnes of Sun and Moon.2) Occasionally it is not social but religious groups that have been divided into two. How persistent and comprehensive such a division may be, is seen from the remarkable dualism of the Brahmin caste. This is divided into the group of the Panch Gaur, the group of the North, consisting (as the name indicates) of five subdivisions and that of the Panch Dravid (again divided into five parts), living in the South of India. The dividing-line between the two groups is said to be the river Nerbudda, whichalso serves as boundary between Hindustan andtheDeccan. This boundary must not be considered geographically, for the Gujarati Brahmins belong to the Southern group, although Gujarat lies North of the Nerbudda. A mythical division into North and South, must rather be thought of, which has also been applied geographically here. And it is wrong to suppose that a dividing-line drawn over such an enormous stretch of land cannot mean anything, as we also learn from the following quotation of Père Coeurdoux in the year 1767: „Et encore aujourd'hui ceux qui sont nés dans des pays plus septentrionaux se regardent supérieurs au méridionaux".3) This feeling of superiority of the Northern group may be more than a reminiscence of the importance of the North as a cultural center of influence, for what is classified in the North is held to be superior to what is classified in the South. * * * Our research must now be continued more historically. The great difficulty is here that the light of data becomes gradually !) Hocartwrites aboutthe c.c.marriage withphratry-division: „Another way of approaching the problem is by looking for divisions that fight one another", and then mentions „the hostility between the right-hand and the left-hand fractions of Southern India". Further he mentions some cases of „joking relationship" between the c.c. Ind. Ant. 1923 p. 270. !) Russell — IV p. 418. 8) Indo-européens et Indo-iraniens, p. 4. scarcer. To find out the way we have to take we shall always use the phenomena we found as beacons, by which we shall go. As, if we do so there is a great chance of wrong orientation, the following must be considered to be suggested in the Modus Potentialis. Modern data lead us to the presumption that in the North (where the MBh. takes us after all) we shall have to do with a community that knows a circulative clan organization and doublé unilineal descent. For we notice something like that in the South in so far as conditions have not changed too much owing to the formation of castes. As the North is of old the cultural centre of diffusion one may reasonably presume that on the periphery of this cultural area, so in the South, the old conditions have maintained themselves longest. Also with regard to the social structure the North is more modern: kinship exogamy has ousted clan exogamy to a greater extent than in the South. With a view to the close cultural relations between the North and the South we assume, however, that the data we found for the North may be taken as indications of a clan exogamy, much older than the kinship exogamy now extant. For these reasons we are on the look-out for the social structure we mentioned. We begin with a quotation from Russell, which shows the modern Brahmin's conception of the regulation of marriage based on the old tradition. „The prohibition of marriage within the gotra or exogamous division bars the union of persons related solely through males. In addition to this, according to Hindu law a Brahman must not marry a girl of his mother's or maternal grandfather's gotra, or one who is a sapinda of his father or maternal grandfather. Mr. Joshi states that sa pin das are persons related through being particles of the same body. It is also understood that two persons are said to be sapindas when they can offer pindas or funeral cakes to the same ancestor, and the relationship is derived through the father of either party. If either is more remote than fourth in descent they apparently could marry. If the relationship of the couple is through their mothers in each case, then they cannot marry if they are third in descent from the same ancestor, but may do so in the fourth or subsequent generations. It is of no importance whe.ther the intervening links between the common ancestor and the proposed couple are male or female; descent is considered to be male if through the father, and female if through the mother. In practice, marriages are held to be valid between persons fourth in descent from a common ancestor in the case of male relationship, that is, persons having a common greatgrandparent in the male line or a common grandparent in the female line can marry. Other rules are that girls must not be exchanged in marriage between two families, and a man may not marry two sisters, though he can marry his deceased wife's sister." x) It needs no prolonged consideration to see that this regulation is somehow based on a clan system. But it is also evident that one has after all to do with kinship exogamy. Nor is this regulation clear to us at all. How can e.g. people that have a common greatgrandparent in the male line intermarry, without infringing upon the exogamy of the patrilineal gotra? Who is the grandparent in the female line? Is it the girl's paternal grandfather and the young man's maternal grandfather, so that the girl is his m.b.d.? If this be so, how can the marriage with a girl of the gotra of a man s mother or mother s father be prohibited? Is there also a latent matrilineal tendency present here, so that, at one time the exogamy of the patrilineal clan, at another time that of the matrilineal clan, is thought of without the investigator being aware of it? The regulation of marriage is especially intricate, because the clauses derived from the old texts by Hindu lawyers are not always distinguished from the actually obtaining regulation. Take, for instance sapinda exogamy. If all its demands would be strictly observed, there would be according to V. H. Mandlik's calculation 2121 relatives prohibited to either party, if it is assumed that there is only one brother and one sister in each household. It would be a very dark prospect for young people who are desirous of marrying, if the law did not help them by expressly stating that local custom may prevail on the legal regulation.2) As a matter of fact all kinds of curious exceptions are made in determining sapinda kinship. If the relationship indicates for instance that the common ancestor ») Russell II p. 362. 2) Vyavahara Mayökha p. 352, 355- is third or fourth to the man, and fourth or fifth to the girl, in the ascending line, the marriage is allowed. If it is first, second, fifth or a higher degree to the man, and third degree of the ascending line to the wife where they have their common ancestor, the marriage is prohibited. Strictly prohibited it is especially when the ancestor is fourth in the ascending line. Curious enough the marriage is allowed if the common ancestor is for both second in the ascending line. „(We now) give rules in the case of the third female who becomes eligible for marriage according to the text „the third female or the fourth female on both the sides (may be married)". „The third female on the mother's side may be either mother's brother's daughter or mother's sister's daughter, while on the father's side the third female (may be) either father's brother's daughter of father's sister's daughter. Of these the father's brother's daughter should be left out, since she is of the same gotra (family). Daughters of the father's sister and of mother's sister should also be left out in accordance with the text of Manu, viz. „a wise man may not marry three females, viz. the father's sister's daughter, the mother's sister (and) mother's sister's daughter.. The mother's brother's daughter (is therefore) the only third female who may be married.... (and that too) by the third male alone (and) not by the fourth or subsequent male. Some allow a marriage with the daughter of the father's sister under difficulties."1) It seems to us that this strange and seemingly arbitrarious set of regulations of which there are only a few summed up here, becomes reasonable if we are to think of a bilineal descent. We shall return to this below. Sapinda exogamy is a sort of kinship exogamy. It extends so far, that no conceivable clan organization can function by its side. By sapinda exogamy so many persons of all possible degrees of relationship are excluded from marriage that there is no room for a clan system with classificatory kinship terminology by the side of sapinda exogamy. In a clan system it is exactly cousins who rather often intermarry. „The nearer the relationship, the more easily are marriages contracted," says Dubois.2) 1) Mandlik — op. cit. p. 354. 2) Manners and Customs p. 21. So the relation between gotra-and sapinda kinship cannot possibly be the same as that between clan and family group. We have already advanced this as a serious objection to the form of Senart's theory. Though this would suffice, we shall include a parenthetical digression on sapinda relationship into our argument. For the determination of sapinda relationship different criteria come into account, depending on the question whether one starts from marriage possibility, or communion with the pinda offering to the ancestors, or descent, or the right of succession. It can not be mere chance that these very points determine one by one the sapinda relationship. At first sight it is incomprehensible that on the strength of so entirely different criteria groups of relatives are indicated, which are always put under the generic name: sapinda. On further consideration one discovers a central principle in this, now so complicated notion. This central principle is the function of the clan. For in a clan organization clan membership not only determines from whom one can inherit, who one may marry, andtowhom one is akin, but also who is allowed to partake in the clan rites. The clans are always in a relation of mutual service with each other. Now one of the most striking aspects of sapinda relationship is the participation of the mortuary rites. It may very well be that there is no difference between the clan of which one says: „This clan should be present at the pinda offering to the ancestors", and that of which one says: „In that clan we may marry". Only when the clan organization breaks up, the connection between these notions is no longer feit. The pinda sacrifice is in the first place for father, grand-, and greatgrandfather, or also for the three maternal ascendants, according to one text for mother, grandmother, and greatgrandmother.1) Manu III. 147 lays down that to the offering to the dead virtuous and learned Brahmins should to be invited as guests. A. M. Hocart writes about this: „The following subsidiary rule is recognized, which is always practised by the good: a man may feed his mother's father, and his mother's brother, his sister's son, father-in-law, son-in-law, maternal 1) w. Caland-Proc. Roy. Ac. Amsterdam. Dep. Lett. XVII p. 10. relation (bandhu) *), both the sacrificial priest and the institutor ofthesacrifice. Further on Manu selects the daughter's son for special mention: „One should at a funeral feast strenuously feed the daughter's son, even though he be under a vow (that is, still a youth)". The paternal relations nowhere appear, save in one exception that proves the rule. For v. 220 lays down that if „His father be alive he should offer only to those before him, or feed his own father at the funeral feast as one of the Brahmans." In other words the father can partake not as a relation but as a learned and virtuous Brahman; he comes in under the primary rule. In conclusion, then, we may say that in India too, the maternal relations consume the sacrifice as representatives of the gods or the fathers.... The inclusion of the father-in-law and the son-in-law among the partakers of the sacrifice thus offers no difficulty, since in the c.c. system they are identical respectively with the mother's brother and the sister's son."2) When we consider the different relatives summed up by Manu it will be seen that in a clan system they would all belong to the clan of the mother's brother and the father's sister. We do not think, with Hocart, of the existence of a complete but of an exclusive c.c. marriage. With this marriage regulation it is indeed natural that men of the clan, of which the ego's clan obtains wives (the clan of the m.b.d.) and of which the ego's clan obtains husbands (the clan of the f. s. son) should assist the ego's clan at the ritual for the dead. The own clan is of course also admitted to the ritual, This could hardly remain out of consideration as „an exception that proves the rule". So originally the sapinda relationship possibly expressed the relation between the clans that are in connubium with each other and which consequently were also to assist at the ritual with mutual service. We imagine that at the discontinuation of clan exogamy in the first place those persons were excluded form marriage who belonged to the clan with which one's own clan was in sapinda 1) „Bandhu means kinsman in general, but more particularly a maternal kinsman. It is obviously meant here in the restricted sense, or the enumeration is pointless. In v. 264 it is used in opposition to jati, paternal relations." (Note by Hocart). 2) Maternal Relations in Indian Ritual-Man 1924. relation. This kinship exogamy (sapinda exogamy) is still very well noticeable in the terminology of the clan system. In course of time kinship exogamy spread wider and wider. Why sapinda relationship is extended roughly speaking to the fifth generation we shall try to point out presently. In short: sapinda exogamy is indeed kinship exogamy, though it is still connected with and directed against clan exogamy. It is certainly to be regretted, that the meaning of this sapinda relationship should not be absolutely clear, for a wellknown precept demands: Daran kurvita' sagotran matur asapindan (Gobh. III 4.1.) a similar regulation as Manu III 5.5. and MBh. XIII. 44. This may teil us that the proper wife should be asapinda matuh and asagotra pituh.1) It has been mentioned above that the gotra is an exogamous patrilineal group. On the marriage regulations of the Brahmins, to whom the gotra is especially familiar, Abbé Dubois remarks: „They (the Brahmans) are all supposed to know the gotram or stock from which they spring: that is to say, they know who was the ancient Müni or devotee from whom they descend, and they always take care in order to avoid intermarriage with a female descendant of this remote priestly ancestor, to marry into a gotram other than their own."2) So in itself the gotra is a traditional unilineal group of relatives, in other words a clan.3) Yet there is a difference between gotra and clan, for, if we have understood the data right, the different gotras are not in a fixed connubial relation with each other. So the gotra is in a way a clan, but then a clan disconnected with a clan system. So, here too, Senart's theory does not hold good, for the gotra has not the same function in the caste as the clan has in the tribe. What exactly the gotra was in the primitive clan organization is only a matter of conjecture. Is perhaps gotra a name for a formerly matrilineal clan ? It has been mentioned above that in India there seems to exist a latent matrilineal organization. This latent matrilineal system, which we also expect to find in old India, perhaps manifests *) A. Weber-Ind. Stud. X p. 75. 2) Manners and Customs p. 21. 3) It is not to be recommended, as is often done, to indicate all smiler exogamous groups by the name of „gotra". itself in the gotra.1) Some data point to this. We admit that positive evidence is extremely scarce. At any rate it is certain that terms like kula, jati, gotra, vamsa, varna, etc. refer to a patrilineal organization. It is only a question whether this has always been so. There are traces of a general matrilineal system. Professor Weber observes e.g. that the Dasapeya ceremony might only be partaken of, if a person could mention ten ancestors, both through the father and through the mother; according to Laty. IX. 2.5. even ten female ancestors through the mother.2) Yajnav I. 53, 54 demands the same for a marriage, besides the requirements concerning sapinda- and gotra relationship.3) Here the prestige of a matrilineal clan apparently comes into account. Inheritance in the maternal line was also known. The Stridhana (a wife's - peculiar - property) did not fall to the son but to her daughters. It is exactly the gotra which the only positive evidence of matrilineal grouping refers to. J. Przyluski says: „Tout est suspect dans la légende de Maya, mais nous savons, et ceci est digne de remarque, que Gautama était du même clan que sa tante maternelle Gautamï. Dans 1'Inde ancienne le nom et le clan se transmettent parfois en ligne ut ér ine. 4) On the Stüpa of Bharhut the donators are sometimes mentioned by the name of the mother's gotra. It is doubtful whether Dr. Fick should be right to account for this as an endeavour to distinguish the children of a polygyne royal family. Finally K. P. Chattopadhaya has wanted to show that matrilineal clans must have existed among the Satakarnis and Sungas.5) Then there is a curious utterance in Panini IV. 1.162 according to which the gotra begins with the grandson (Gotram pautraprabhrityapatyam) 6) This gives the impression that to *) D® Josselin de Jong opines that virtually in every unilateral tracing of relationship, a more or less apert, doublé unilateral tracing is to be found. These types must not at will be reduced to blends. 2) Weber-Ind. Stud. X p. 86. ' ) ib. p. 88 n. 1. 4) Journ. As. 1927 p. 177. ') Social Organisation of the Satakarnis and Sun gas — T. A. S. Bengal N. S. XXXIII/1927 p. 503. ' 6) C. V. Vaidya remarks: „Here comes in the Sütra of Panini.... explaimng the generally used word gotra or gota as it is used in modern Indian languages. Gotra in this sense means the descendants from the grandson onwards of any well-known man". Proc. First or Conf. Poona I p. XV; II p. 34. Panini the gotra was not a common patrilineal group, for in that case there would be no sense in beginning with the grandson. It is also curious in this connection that the father s mother should be entitled to inherit in virtue of Sagotratva, or gotra partnership.1) In this regulation, which rather worries the modern Indian lawyer, who only knows of the patrilineal gotra, a reminiscence may be seen from a time when the father still had the same gotra as his mother. The etymology of the word gotra would allow of the translation „matrilineal clan" if necessary, for gotra means „cow-house". The cow, representative of Mother Earth, is sometimes found as the pendant of Prajapati, the creator, thought as the buil. Tradition usually mentions eight original gotras. According to MBh. XII. 298. 17 there were at first four root gotras (mülagotra). The gotras differ from each other in the outward appearance of their members (especially in their head-dress) and in different ritual concerns.2) Whatever one might think about the gotra as a matrilineal group, we are rather certain that there really are traces of matrilineal grouping. In the Jatakas the ideal of immaculate descent was thus: „Beiderseits von guter Familie, sowohl mütterlicher- wie vaterlicherseits, von reiner Empfangnis bis zum siebenten Gliede, unbescholten und vorwurfsfrei hinsichtlich der Ge- burt."3) .... We have seen by now that sapinda exogamy is indeed kinship exogamy, but that there is still a clan exogamy behind it; besides, that this clan organization may as well be bilineal as the modern organization, whether this matrilineal clan has or has not been the (presently patrilineal)4) gotra. Next comes the question: Has this clan organization also known a circulative marriage system? The answer to this question too could not be anything more than a theory of probabilities, though less so than the idea of bilateral descent. Although, as has been observed, the sapinda relationship extends, broadly speaking, to the fourth of the ascending line, „so erhellt doch der Umstand, dasz diese >) Mandlik — op. cit. p. 360. 2) Weber-Ind. stud. p. 88. ») Fick-Soziale Glied. p. 128. _ ') About the possibility of this transition, rel. to p. 80. eigenthümliche und strenge Satzung in der alteren Zeit noch keine allgemein anerkannte war, mit Sicherheit aus folgenden in Qat. I, 8, 36 vorliegenden Angaben: „von einem und demselben Mann stammen Geniesser (Gatte) und zu Geniessender (Gattin): denn die Geschlechtsgenossen (jatyah) pflegen sich jetzt weidlich am (Liebes-) Spiel zu ergötzen, indem sie sagen: Beim vierten Manne (Gliede, oder) beim dritten vereinigen wir uns (fleischlich, coimus)". Sayana bemerkt hiezu, dass die Kanva bloss „beim dritten", die Saurashtra dagegen „beim vierten" lesen; die Dakshinatya schliessen sich nach ihm den Kanva an, da sie die Heirath mit den Mutterbruder-Töchtern und den Vaterschwester-Söhnen verstatten, die je vom Grossvater ab gezahlt das dritte Glied bilden. Dasselbe berichtet der schol, der Vajrasüci .... von den Andhra und Dakshinatya, giebt indes zugleich an, dass die Schule der Vajasaneyin die Heimführung der MutterbruderTochter verbiete." *) It may be seen that sapinda exogamy was preceded by an organization with cross-cousin marriage. There is every reason to assume that the latter was an exclusive (unilateral) c.c. marriage, as Sayana states for the Dakshinatyas and the Kanvas. When the passus at issue was recorded in the Sat. Br. the c.c. marriage was consequently not yet superseded in the North. To this day the passage is the locus classicus pro m.b.d. marriage for the Hindus. The quotation from Weber also shows the aversion to the m.b.d. marriage, which, as has been said, still exists in the North of India. The rising endogamy naturally turns against this marriage relation, which was practically involved in the Indian clan system, consequently aigainst the m.b.d. marriage. Baudhayana mentions the m.b.d. and f.s.d. marriage among the characteristic customs of the South.2) It is possible that in his days the c.c. marriage did not actually occur in the North; it is also possible that this type of marriage, which really was not the type of the endogamous caste, as will be pointed out below, was attributed by him to the less orthodox South, while it was still also in existence around him. To a circulative marriage system also points the strong aver- *) Weber-Ind. Stud. X p. 75. See also S. B. E. XII p. 238 n. 1. ») S. B. E. XIII p. 146. sion from a marriage in which the woman is superior in rank to the man. Marriages outside the caste are indeed always sinful, but there are gradations. Marriages „against the hair" (pratiloma) are much more detestable than those „with the hair" (anuloma), consequently the alliance of the lowest man (Südra) and the highest woman (Brahmin)isheldtobemeanest. „In the case of the first three castes Baudhayana declares that the offspring of the male of the higher caste and a female of the next lower caste are „savarnas" or of equal caste. Consistently he gives the name of the issue of the marriage of a Brahmin with a Kshatriya female as a Brahmin, and that of the union of a Kshatriya male and a Vaisya female as a Kshatriya. This seems to represent the old practice, when marriages between males of higher castes and females of castes next in order were regarded as perfectly regular and entailed no disqualification on the issue." *) This points, in our opinion, to a unilateral, and in no case to a reciprocal marriage system. With such a hypergamy the possession of daughters is not a source of unmixed pleasure, because it is not always easy at all to find a bridegroom of a higher class than the bride. Matali, the charioteer of Indra gives utterance to the following lamentation: „Ah, it is no easy matter if in a family of blameless, high, and excellent men of good character a girl is born. Her father's family and her mother's and besides that into which the girl is married, three families such a girl causes to run risks".2) Dhik khalvalaghusïlanam uchchhritanam yasasvinam Naranam mridusattvanam kule kanya prarohanam Matuh kulam pitrikulam yatra chaiva kanya pradïyate Kulatrayam samsayitam kurute kanyaka satam It may be seen from this fatherly complaint that there were three kulas (patrilineal clans?) wanted for a marriage. This is the case in a circulative or three-clan-system. The m.b.d. marriage is also known in history. The royal families of Magadha, Kosala, Avanti and that of the Vatsas „were united by matrimonial alliances, and were also, not seldom in consequence of those very alliances, from time to time at ') Ghurye-Caste and Race in India p. 77. 2) MBh. V. 97. war". King Ajatasattu of Magadhafought with his step-mother's brother Pasendai of Kosala and married his daughter, in other words his m.b.d. x) An instance in the more sacred history is the marriage of Arjuna with Krishna's sister, Just because the family relation of Krishna and Arjuna is so important, (which will be further explained below) this datum is of great importance. If, to prove the sum, one looks at the terminology of kinship one finds that the term for mother's brother is not the same as that for father's sister's husband, so that on this head the existence of a common c.c. marriage system is improbable.2) The terminology is generally very supple. One and the same term can easily be used to express actual kinship and the clan relation, as is the case, e.g. with terms like pitrisvasrïya (father' sister's son) and matuleya (mother's brother's son). For that matter, the same holds good for kinship terminology as for the whole system3) of relationship: at present it mainly fits in with a kinship exogamy, but it still reminds one vividly of a clan system. One should always pay attention to the fact that terms like mahapitri, putra, snusha, etc. can also be used for groups of persons, which also the Indian scholars themselves have pointed out.4) In epic literature, for instance, the women of the Pandavas are to Dhritarashtra in the relation of snusha to svasura. Another tracé of the classificatory system is found in the name bhratri (brother), by which the Pandavas and the Kauravas address each other. In the system of genuine kinship exogamy they would be cousins, as they are brother's sons. Rama always calls all his father's wives „mother".5) As may be expected in a clan organization, where division into phratries has not been absent (see p. 86) the mother's *) Rhys Davids-Buddhist India p. 3. 2) See p. 56 n. 2. 3) Not of all terms the meaning is incontestable. Hocart assumes that the meaning of bhratrivya is „mother's brother's son", and not „father's sister's son", (as in the Vedic Index) This sounds more probable indeed, because brother's sons (at least in the Epic), call each other bratri-brother, which points to a clansystem. Hocart — The cousin in Vedic Ritual Ind. Ant. 1925. Wackernagel (Indog. Jahibuch V.p. 70 embraces an old meaning, „brother's son", later changed into „father's brother's son". 4) Mandlik — op. cit. p. 33. 6) Winternitz-Gesch. Ind. Lit. I p. 410 n. 2. brother very often comes into prominenceintheoldestliterature as well as his social complement: the sister's son. These two represent the two phratries, and besides, two generations. In a three-clan-system the m.b. is the father-in-law of his sister's son. Moreover he is, as the representative of the other phratry, the proper man for the rites of initiation. It cannot be mere chance, therefore, that Sakuni, who plays such an important part in the dicing scene, is just the mother's brother of Kauravas. Nor can this be the case in the relation of Krishna and his father's sister's son Arjuna. If we put two and two together, we think there is every reason to assume (also for the old period, with which we are concerned), the existance of a circulative patrilineal clan system with latent matrilineal tendencies. The well-known data point to this, and there are no other data that make our conception impossible. There is only one datum known to us that in itself would rather suggest a genuine type of c.c. marriage, namely Rig Veda Khila VII 55.8a, where it is said of Indra that he has no less a right to the offering than to the f.s.d. and the m.b.d. For that matter, a genuine c.c. type often exists by the side of the circulative type, as is still the case in the South. Of much greater importance is an article about Buddha and Devadatta, written by A. M. Hocart in 1923» in which he proves that the genuine c.c. type was known.1) Buddha s genealogy (schematized for convenience's sake after the method we have always applied) is printed below.2) Hocart writes there: „Anyone who has the slightest acquaintance with kinship systems will immediately diagnose the case. It is the crosscousin system, under which a man's children are expected to *) Ind. Ant. Lil p. 267. l) Jayasena = Dêvadaha = (Sakya) I IK.011) marry his sister's children, but not his brother's children. In technical language a man marries his cross-cousin, a term invented to express the fact that they are cousins through parents of opposite sexes. Such a form of marriage results in a system of reckoning kin, in which the maternal uncle is the same as the father-in-law, the paternal aunt as the mother-inlaw, and so forth, as any one can work out for himself on the above pedegree." This is right, yet we do not draw the conclusion that in Buddha's time the c.c. marriage was prevalent. First, because the other data point to the impossibility that this should have been the normal type of marriage. And especially because one can only draw the conclusion that the text, in which this genealogy was mentioned, belongs to a milieu where the c.c. type was known. This text is the Mahavamsa (II. 15 ff.), one of the Pali chronicles of Ceylon, probably by the poet Mahanama of the last quarter of the fifth century A.D.1) The Mahavamsa does not closely follow its sources, so that the work is not authorirative for Buddha's time. From Hocart's article one may consequently conclude that the c.c. marriage in India is very old and probably known in Ceylon a few centuries after Christ. * * * Ëmile Senart is the man who has best understood the caste system, in the opinion of many experts. Though we have sympathised with this, the foregoing may have given the impression, as if we wanted nothing else but to prove the untenability of Senart's theory. Yet this is not the case. It was only our purpose to show that the similarity he found between caste und tribe does not hold good. We want to propose a few things, salva modestia, in support of Senart's theory. In the preceding, where we tried to determine something about the regulation of marriage, we have hardly occupied ourselves with the caste as such. For instance a certain dualism was stated, which, however, unlike the genuine dualism of a phratry, knew of no exogamy, but of endogamy. We stated already that the caste system with its distribution of functions, and its hierarchic grouping resembled a 3) Winternitz-Gesch. Ind. Lit. II p. 167 f.f. clan —, rather than a tribe organization. If it was not the same here: the clan being exogamous, whereas the caste is endogamous, we (and probably Senart long before us) had directly associated caste, not with tribe, but with clan. So, the following question naturally suggests itself: Is there in certain clan systems a tendency towards endogamy noticeable, which would eventually change the exogamous clan into an endogamous caste? This question may be answered in the affirmative. With the Banaro (New Guinea) the largest social unity is divided into two endogamous groups, called Ironwood and Bamboo-flute. This endogamous grouping intersects an exogamous patrilineal clan grouping, so that e.g. a man of the Ironwood half of clan i. marries a girl of the Ironwood half of clan 2. This phenomenon is even more striking among the Natchez Indians. There existed an exogamous division into two of Nobility and Commoners. The nobility was divided into ranks: a. the „suns", of whom the kingwas the highest „sun"; b. the,, Nobles"; c. the „honoured people". The commoners bore the rather disagreeable name of „Stinkards". The status was heritable through the female line, so that the children of a noblewoman were nobles themselves. The children of a nobleman, however, did not follow the rank of their mother if she was a commoner, nor did they follow the rank of their „noble" father, but descended in status by one degree. It is not likely that the exogamous division into two of the Natchez may be detached from other exogamous divisions into two. So it is reasonable to assume that the Natchez formerly knew a common clan system and that the function of these old clans lived on in the ranks of the nobility. This class as a whole had been substituted for the ancient phratry, the nobility was probably the originally superior, the commoners being the originally inferior phratry. It is now characteristic that among these Natchez the nobility showed a tendency towards endogamy and that the commoners were no longer strictly exogamous either. Here we consequently see the caste in statu nascendi. 1) *) Most of his views de Josselin de Jong has expounded in lectures at Leyden University. For this problem we refer however to: Proc. Twenty-third Int. Congr. of Americanists. Sept. 1928 p. 553 — The Natchez Social System. A similar development is probably perceptible among the Ngada in Flores (East-Indies). Now there is every reason to assume that the social structure of India shows parallel phenomena. In a recent study on Totemism in India Hermann Niggemeyer also speaks about the division into endogamous groups.L) From his startingpoint Niggemeyer comes to the conclusion that endogamy is a younger development; a conclusion that is all the more remarkable as Niggemeyer can in no way be suspected to bring the facts more into harmony with the caste theory that we defend. He says: „Nach diesen Erwagungen ist wohl wahrscheinlich gemacht, dasz einerseits das Prinzip der endogamen Untergruppen ein verhaltnismaszig junges Element in der gesellschaftlichen Entwicklung Indiens ist, das erst spat in die totemistische Gliederung der Stamme und Kasten eingriff und diese modifizierte, und dasz anderseits durch die Stamme und Kasten, die keine Einteilung zeigen und die ungeteilt 2) in exogam-totemistische Klans zerfallen, ein alterer Zustand reprasentiert ist. Daraus laszt sich wohl entnehmen, dasz einst die einfache Gliederung der Stamme ohne endogame Einteilung viel weiter verbreitet war als heute." 3) The process of the development of clan into caste does not belong to prehistorie phenomena, whose existence must only be believed on the strength of a general parallelism. The phenomenon was (and quite probably is) observable and describable. The Rajpüts e.g. are divided into clans. Many of the large clans were divided into „branches" (sachae), which are partly local groups. These „sachae" are in a certain status relation with each other. Now, Russell writes: „Besides this (i.e. clans and sachae) the Rajpüts have go tras, named after eponymous saints exactly like the Brahman gotras, and probably adopted in imitation of the Brahmans. Since, theoretically, marriage is prohibited in the whole clan, the gotra divisions would appear to be useless, but Sir H. Risley states that persons of the same clan but with different gotras have begun to intermarry. Similary it 1) Anthropos XXVIII/1933 p. 407 and p. 579. 2) „Zerfallt eine Kaste oder ein Stamm ohne Einteilung in (endogame) Untergruppen sofort in exogame Totemklans, so wird die Gruppe atomisch genannt (Note 12 by Niggemeyer). 3) ib. p. 418. 6 would appear that the different branches of the great clans mentioned above must intermarry in some cases; while in the Central Provinces, as already stated several clans have become regular castes and form endogamous and not exogamous groups. (Spaced out by the present writer). In Nothern India, however, Mr. Crooke's accounts of the different clans indicate that marriage within the clan is as a rule not permitted. The clans themselves and iheir branches have different degrees of rank for purpose of marriage, according to the purity of their descent, while in each clan or subclan there is an inferior section formed of the descendants of remarried widows, or even the offspring of women of another caste, who have probably in the course of generations not infrequently got back into their father s clan. Thus many groups of varying status arise".*) This description leaves little doubt as to the character of the process. One could only wish to learn a few more facts. In short the course of events is thus: Clan exogamy declines owing to the tendency towards endogamy. Now, not exactly as many castes suddenly spring up as there were clans before, but after a period of wavering a definite hierarchic order of status-indicating groups arises in one form of kinship exogamy or another. These status-indicating groups show a specially strong tendency towards endogamy. What has been noticed in this one case among the Rajpüts has occurred on a large scale in the whole of India and probably still occurs. Remarkable in this connection is also an utterance of Niggemeyer in the essay we have just mentioned: „Auffalligerweise haben alle exogamen Gruppen, die sich teils neben die Totemklans, teils an ihre Stelle gesetzt haben, etwas mit Blutverwantschaftsexogamie zu tun, sei es, dasz die Heirat innerhalb gewissen Generationen, vom Vater oder von der Mutter gerechnet, verboten war, sei es, dasz die Heirat innerhalb einer Sippe oder Familiengruppe als unerlaubt angesehen wurde." 2) It is obvious that within certain limits there are all kinds of possibilities and phases when a clan is becoming endogamous. *) Russell IV p. 418. Crooke signalizes the formationof new endogamous groups, (sub-castes) for instance among the Chamars (11 p. I74J and among the Sunars (IV p. 336). 2) loc. cit. p. 437. Owing to this the social structure of India shows such a peculiar, confusing picture of numerous groups intersecting each other, so that it gives now this now that impression according to the aspect turned towards the observer. This should also be kept in mind when estimating what we have said about clan organization. Closer investigation on the spot is necessary to form a detailed conception. If we suppose for a moment that a clan organization with circulative marriage system tends to the formation of castes, it would be expected that in India (where the male takes the highest place in the classification) there would arise the greatest aversion to a marriage of a woman of the highest with a man of the lowest clan. Or, if we take the scheme on p. 58 before us, that there would be a stagnation in the relation of clan 1 with clan 3, because clan 1 would object to having its women married to the men of clan 3. The consequence would be that in clan 1 a surplus of women is formed, and in clan 3 a surplus of men. If our hypothesis holds good in general, one must see that there arises in the highest classes a surplus of women and, to a less degree, a surplus of men in the lower classes, that are connected with these higher classes. The surplus of men will not be so striking, as the lower classes cannot be so fastidious in the choice of their mates as the higher ones. This is also seen in reality. The Rajpüts did not know sometimes how to escape from the impasse but by killing their daughters. Hypergamy may generally lead to objectionable situations: polygamy in the higher and scarcety of women in the lower classes.1) If we now turn back to ancient times, which most interest us with regard to the MBh., it must first be stated that in the older literature an increasing tendency towards endogamy is perceptible. So this shows that clan exogamy was only beginning to disappear at that time. As the Vedas throw little light on the social organization, it cannot be stated with certainty when the phenomenon at issue begins to become evident. It is well-known that the caste was not always strictly endogamous in epic literature. In the period of the Jatakas endogamy *) See Hastings E. R. E. s.v. Caste; Russell I p. 40; II p. 364; 435; IV p. 419. was „rather a universal custom than a rigid rule of caste".x) During the period of the Brahmanas and the Sütras a right to have three women in consecutive castes is conferred on the Brahmin, two women on the Kshatriya and one on the Vaisya. According to some all were permitted to take a Südra woman besides. The right to have four women is expressly conferred on the king, which proves that the king was at that time already considered a central figure.2) The prohibition to have intercourse with a Südra woman is often given as a special regulation, from which Albrecht Weber concludes that this was not generally prohibited.3) According to one source the exogamy of the caste was even regular. 4) When, on certain occasions one had to mention ten matrilineal ancestors, one apparently often came across mothers of a caste different from that of the mother, for which case special measures have been taken. „Hieraus geht zur Genüge hervor, dasz haufige Heirathen zwischen br. (Brahmanen) und Frauen aus anderen Kasten stattfanden." 5) So, on the whole there are reasons to assume that the castes have sprung from clans. The castes are classes that have the function of clans. In so far data are available, one sees endogamy spring from exogamy. At what time this transition must be dated is only a matter of conjecture. We have the impression that one has to think of the time of the Brahmanas, for in the Sat. Br., a source that would certainly not fight for clan exogamy, a cross-cousin marriage is still known. Caste endogamy has not been carried through so strictly at that time: kingship is only clearly noticeable at the time of Buddhism. Everything points to the caste having originated at this time. How this transition was effected is easier to understand. The reason of this tendency towards exogamy is the growing prominence of the status indication, which was expressed already in the clan system. For the clans are not all equally high in rank. To begin with there is the division into phratries that ') Ghurye — op. cit. p. 80. s) Weber-Ind. Stud. X p. 21. 3) ib. p. 74') ib. p. 87. 5) ib. p. 87. is not lacking in India. One phratry is always higher than the other. In the classificatory system one phratry is male, the other female. That is also the background of the typical rivalry between these two groups. It is the same with the clans. The reader may be reminded of what has been mentioned about the Omaha Indians. One clan has a function different from the other, and consequently has a status different from the other. When the tribe assembles, each clan has a place of its own in the encampment. The exclusivism of the clans is in continual polarity with the necessity of cooperation within the tribal organization. Necessity — for they are interdependent e.g. in the ritual and in marriage. When this exclusivism is intensified, the cooperation, based as it is on the mutual rendering of services, slackens and assumes different shapes. This is an important dynamic element in the formation of castes, for the caste is after all an exclusive group for status indication. The idea of cooperation has never quite vanished here. Otherwise one could not even speak of caste organization. If we suppose that there was in olden times a circulative marriage system with a division into patrilineal clans (which we may be allowed to call „kula") and matrilineal clans (suppose these were called „gotra"), the result is the following hypothetical construction: With the risingtendency towards endogamy the most distinguished kula refuses to give women any longer to the lowest kula. On the scheme: clan 1 refuses to give women any longer to clan 3. But for the time being the highest clans go on taking women of the lower clans, especially those of the next in order, besides that, owing to endogamy, they marry girls of the own clan (hypergamy). In practice: the Brahmins marry Kshatriya girls, but the Kshatriyas do not marry Brahmin girls. The strongest aversion rises where the circle of the exclusive system would close viz. marriage between Brahmin women and Südra men. According to the system of Manu the issue of this alliance is the despised Chandala. Marriages between Südra men and Brahmin girls are not on record. So our conception cannot be proved. Only it is extremely strange that the Brahmins should consider women of their own caste the mothers of the most despicable people, if there were no reason whatever. The endogamy focussed especially on the patrilineal clan; people began to marry girls of the own kula. But there remained an aversion from a marriage with somebody of the own (the mother's) matrilineal clan, which we have called gotra. When, owing to the endogamy, the kula has once turned into a caste, such a caste assumes the character of the totality of the tribe. This is evident from the frequent divisions into two of the caste. For the caste has endogamy just like the tribe. One has only to do with the members of the caste. So, the caste, just like the tribe is the whole of society in nuce. As the patrilineal grouping predominates, the matrilineal gotras, whose endogamous character is as much retained as possible, become patrilineal clans. This occurs e.g. when the grandson bears the same gotra-name as his father, while the latter has inherited the name from his mother. As the clan organization proper, and consequently the connubial relation between the clans is discontinued, the only characteristic feature of the now patrilineal gotra left is its exogamy. For the matrilineal line one sticks to a rigid kinship exogamy. We do not mean to say that we are entirely certain of the correctness of our hypothesis. Though we are convinced of the transition of theexogamous clan into the endogamous caste, the change of the matrilineal gotra into a patrilineal grouping remains for the present problematic. *) So we are pretty certain that the divisions into two that are found in India go back on common phratry divisions, in spite of the fact that these divisions are endogamous, whereas those of the phratry are exogamous. There is now the advantage of concord with the Indian literature itself. The Brahmin conception of an originally small number of castes, increased by intermarriage, may be accepted. One need not make on this point a difference in literary type between the Vedas and the Brahmanas. There is now room for a developmental history of the caste system. This disposes of the chief objections to Senart's theory. ■ ) Among other things, also the relation between the ordinaty c.c. marriage system and the exclusive system is not yet quite clear. Remarkable is what Rivers says about hypergamy: „There is reason to believe that this practice forms an intermediate process between the strict endogamy of the orthodox caste-system and the state found in other parts of the world where difference of rank presents no insuperable bar to intermarriage." Man in India I p. 7. Besides, the hierarchic grouping of the castes and the distribution of functions among them is clearing up, — problems with which Senart is at a loss what to do. For we assumed above that endogamy should exclusively focus on the clan. This must not be expected to be found in practice. There also, the unity sprung from local grouping and professional community will have played a part. The caste will not only have arisen from the clan proper, but also from the functional group. The further the caste system developed, the more groups will have become endogamous. We do not stealthily shove in a new element, for the clan grouping is, as we have seen, at the same time a hierarchic organization into functional groups, to which of course a certain local organization must also have been assimilated. In many cultures a division of labour is found among clans and family groups which goes together with a certain exclusivïsm and hierarchic organization. Different groups have special hereditary functions of different kinds, technical as well as shamanistic. Sometimes the whole group is in a special function, sometimes the group has only to provide a single functionary. This phenomenon has been studied e.g. for the Patwin Indians. It is plausible that the distribution of functions over the castes in India must be seen in the light of such a distribution over the different groups within the tribe. Here we also touch on Nesfield's theory. This theory is not satisfactory, because it starts from a phenomenon, which, in its turn, depends on other phenomena, in this case on the social structure. At bottom the theories of E. Senart and M. Nesfield are not incompatible, because a „guild"-like organization of groups is already present in the tribal organization. Dr. Fick mentions kulas of blacksmiths, potters, joiners. Separate local groups of joiners, potters and smiths are on record. To the contemptible professional castes belonged the bamboo-workers, cartwrights, carpenters, blacksmiths, basketmakers, flute-makers, weavers and barbers.1) Also here racial contrast is called in. When the immigrants invaded India they are supposed to have found a native population, which could only work with bamboo and was especially accomplished in ') op. cit. p. 173 f.f. the art of making carts. This is very improbable, because one cannot earnestly maintain that the immigrants could not e.g. forge or do carpenter's work. Apart from the impossible narrowness of such a local-historie explanation,we have a convincing proof of the untenability of this conception in the fact that Rudra is already called in the Y.V. Lord of the carpenters, cartwrights, bow-and-arrow makers, potters, smiths, hunters and, in other sources, Lord of the thieves and all kinds of rabble. All these crafts can hardly be attributed to a pre-Aryan population, unless one also wants to give pre-Aryan descent to Rudra. Besides, it is difficult to see why exactly these crafts would have been familiar to the pre-Aryan population. It is much likelier that an explanation of the taboo of these crafts must be found in religion,notably, in the ritual of the tribe. We cannot enter into this question here, but suffice it to say that several crafts went together with shamanistic practices. The gay note be not absent. About the occupation of a barber, according to Fick, one need not assume that the taboo of this craft should find its origon „in irgendwelchen ethnischen Verhaltnissen". According to him we are to think of the incleanliness inherent to this honest occupation. The modern barber has apparently accomodated himself to the demands of an antiseptic attendance, because he serves the distinguished classes as a matrimonial agent and attendant of the Brahmin at marriage ceremonies; for the lower classes that cannot afford a Brahmin he even holds the profession of a priest. Is it not more plausible to account for the taboo of the barber by the fact that he had to assist at the ritual for other reasons (think of hair-cutting as a certain form of initiation)? In this modified form Senart's theory also explains the hierarchic organization and distribution of functions, which are noticeable in the caste system. But — one will ask — did racial contrast exert no influence at all? No doubt it did, but as a minor factor. The cultures (for it is better to keep race and culture apart) that came into touch with one another here, certainly influenced each other. But this cannot explain the caste system as such, though it is possible that the tendency towards endogamy should have been stimulated by a (possible) difference in cultural level. How far this influence reached we cannot determine, neither is the question of primary importance to us. ♦ * * There is one point in this modified conception of Senart's theory we want to insist upon for the sake of our MBh. research namely the fact that there was a period in the Indian civilization when clan exogamy made room for caste endogamy. At this period, probably to be dated within that of the Brahmanas, culture goes through a stage of transition and wavering. Of this we think we can find the reaction in the MBh., on which the societies (one of the typical phenomena of this situation) and the ritual connected with them, have undeniably put their stamp. It will be the purpose of the next pages to elaborate this more in detail. APPENDIX It is our aim to load our construction of the MBh. with as little dead weight of accidental theories as possible. Not every hypothesis we give is a conditio sine qua non for our conception of the Epic. Most of what we said so far is, in our opinion, still more or less connected with our conception of the Epic. We cannot say this, however, of what we want to propose in this appendix. Moreover, for all we know, we find ourselves quite alone here in the face of a communis opinio established for tens of years. Sufficiënt reason to use the appendix as a suitable strategical point to retreat to in case of emergency without endangering the main body of our argument. We doubt, namely, whether the Südras constituted the native population, which as such was opposed to „Aryas". In the Rig Veda the well-known division into four: Brahmins, Kshatriyas, Vaisyas and Südras only occurs in the Purushasükta (R.V. X. 90). The primeval man Purusha is represented there as the origin of the four groups. From his mouth proceeded the Brahmins, from his arms the Kshatriyas, from the lower parts of his body the Vaisyas and from his feet the Südras. The same representation is found in Manu I. 91 and in MBh. XII. 296. 6. Already on the ground of this passage there might rise justifiable doubt if one might look upon the Südras as aborigines. When we remember that the Brahmins were the fiercest fighters for the caste organization and that Brahmin literature of more recent date was anything but kindly disposed towards the Südras, we may understand that with the ruling conceptions the investigators objected to acknowledging that already in an old text a place in the order of creation was reserved for the Südra. In that case it would for instance become difficult to see why just in the oldest times the aboriginal population got a good place in an unsuspected Aryan text, if at the same time one has to believe that from the clash of Aryans and natives resulted almost all of the typical character of Indian culture. Besides, in the Purushasükta the Südra is in generic, and besides sacral, relation with the other three varnas. It is said that the Purushasükta is young. Why? The reason can not be that the division into four occurs there. In that case the more recent date was only a theory formed pour le besoin de la cause. So this cannot be brought forward as an argument, as has been done by Professor Zimmer when he calls the Purushasükta „ein Lied, das erst entstanden sein kann, als das indische Volk streng nach Brahmanischer Ordnung gegliedert war." *) The data that have led to the objection against the assumption of a division into four in ancient India and to the identification of the Südras with the aborigenes, are in the first place several expressions in the Rig Veda. Over against „our colour" (varna), the Aryan colour, the Aryan name (naman), there is the colour of the Dasa, the black (krishna), or dark (asikni) skin (tvach). The Arya is occasionally described as „godless" (adeva) and for that reason connected with the Dasa. The Arya utters inimical words (or practises magic-mridavach) just like the Dasa or Südra. But the Dasyu-Dasa is reproached for knowing other rites (anyavrata) or no rites at all (avrata). He is not human (amanusha) and does not honour the gods (adevayu). The Aryas take their women, Dasïs, who have black wombs (krishnayonï). Once the Dasyu is called anas and once vrishasipra. The former word is rather obscure and means „without a nose" or „without a mouth". The latter word (with buli's l) A.i. Leben p. 1x7. lips) means „with protruding lips". In this personal description the Dravid was recognized. x) Now P. Giles has already protested against the more romantic than scientific conception that the immigrants should have looked like fair-haired Teutons.2) It is therefore wiser for the present to wait for a moment before stating the superiority of one race to the other. But then, one cannot possibly maintain that there is such a sharp contrast between Dasas and Aryas as one wants to have it. It is plausible that the term Dasa should also refer to the Dravidians, but by no means that the Dasas should be Dravidians. As a matter of fact the word Dasa itself indicates that there is no question about a sharp dividing line covering that of the races. „The aborigines, however," says Professor Macdonell, „were not only known as enemies, but were already in part attracted to the Aryan community as a servile class. For the word Dasa not only means „aboriginal foe" but is also clearly used in the sense of „slave" or „servant" in the Rigveda." 3) Similarly Oldenberg says: „Schon das rgvedische Altertum hat die dunkelfarbigen nicht allein als Feinde, sondern als der arischen Gemeinschaft attachierte Unterworfene gekannt." 4) So in the old literature there is not much positive evidence of racial contrast to be found. Consequently one must come to the conclusion that the mixture of races has exerted its influence in prehistorie times, because the historie data do not make much of this influence. Also with the further spreading of Aryan civilization there is little noticeable of the influence of racial contrast. In the Jatakas the Brahmins and Kshatriyas are put over against the Vaisyas and Südras, irrespective of birth. So Fick says: „Vergebens sehen wir uns in dem hellen Lichte, das die Jataka über das wirkliche Leben des alten Indiens verbreiten, nach einer die gesammte arische von der gesammten nicht-arischen Bevölkerung trennenden Scheidungsgrenze um.... Wohl haben wir für die ersten auf die arische Einwanderung folgenden Zeiten eine Absonderung der dunkelfarbigen Eingeborenen von ihren hellfarbigen Besiegern x) de la Vallée-Poussin — op. cit. p. 142. ") Cambr. Hist. I p. 66. 3) American Hist. Rev. XIX p. 240. 4) Z.D.M.G. LI p. 274. anzunehmen; sie wird bezeugt durch die Gegenüberstellung von arya varna und dasa varna im Veda, Aber schon frühzeitig wird — und zwar je weiter nach den Grenzgebieten der arischen Kultursphare zu, um so mehr — eine Vermischung mit den einheimischen Völkerschaften eingetreten sein." *) It must be granted that the racial contrast in India was not very sharp, judging from these data. Now it is said, — and here we approach the gist of our argument, — that the antithesis of Arya to Dasa Varna, which is found in the Vedas, is the same as that of Arya to Saudra Varna. Hence it follows that the Südras and the Dasas are identical and that via the identification of the Dasa with the pre-Aryan, they must be reckoned a mong the aboriginal population. The chief argument for this construction is the use of the word varna. But it is obvious that the word varna is not meant in a psychological, but in a mythicalclassificatory sense. For it is also sometimes said that the varna of the Südras is black, that of the Vaisyas yellow, that of the Rajanyas red, and that of the Brahmins white, or sometimes that the varna of the Vaisyas is white and that of the Rajanyas dark. And most remarkable of all is that the texts deliberately maintain sometimes that this colour actually refers to the skin, which „daher nicht wenig befremdet" as the learned scholar, Weber, observes.2) Some scholars remain consistent also here and discovered white, red, yellow and black people. On what biological phenomena the cameleon-like variations in colour of the different varnas must be based, is unknown to us. We will not spoil our case, however, by eliminating a priori all real racial contrast. But we still doubt whether an identification of Südras with pre-Aryans is exactly right. Neither is it seen, for that matter, that the Südras gradually come to be received in the ritual life of the Aryans, but it is just the opposite, they are more and more excluded. The prohibitions of marriage with the Südras gradually become stricter. In the ritual the Südra only officiates in the solstice ceremony, the Mahavrata festival, as the partner of the Brahmin, with whom he fights a ritual combat over a piece of white leather, which is thought to represent the sun. As if the fact that here a Brahmin *) op. cit. p. 163 n. 1 and p. 201. ') Ind. Stud. X p. 10. and a Südra function together in a sacral event were not sufficiënt, Sankh. XVII, 6. 2, informs us that this custom is not new, but „old, obsolete, and negligible".1) Also in the ceremonial game of dice to whose importance we shall return later on, the Südra plays as the King's opponent. It is said of the king that he may have four wives, consequently also a Südra. The fire necessary for the Agnyadheya can be obtained by rubbing, but also by taking it from the fire place of a rich Brahmin, Rajanya, Vaisya, or Südra. According to others the Südra is excluded from this.2) From these instances, to which many might probably be added, it is seen that the Südra could partake in the ritual. Max Müller observes that the Sütras never expressly exclude him from the initiation, which was later on strictly prohibited to him.3) We cannot but state a growing and not a declining aversion from the Südras. Now it should not be said that the Südra only appears where he is obviously represented as the inferior of the other three varnas and that consequently he is yet a Dravidian. To appear in the ritual is always the recognition of a right, the attributions of a certain ordination. Otherwise one simply cannot officiate in a ritual. There seems to be no reason whatever for us to doubt that the division into four, of which the old texts speak, is right, while also the Südras fit in the same whole as the three other varnas. It is unacceptable that a group of population, with which one had nothing to do, should have been given a place in the sacral order of creation that embraced the own tribe. The division into four consequently refers to a homogeneous cultural conception and cannot be dissolved into three classes of Aryans with a group of aborigines, called Südras. Apart from the Purushasükta its existence appears from the wellknown expression panchajanah, by which the whole people is meant. The number five has here a sacral meaning and is accordingly taken by the Indian commentators as the indication of totality, represented by one in the series: gandharvah pitaro deva asura rakshamsi; by the other in the series: Br., Ksh., V., S. and Nishadas. 4) It has now been definitely established that ') Weber — Ind. Stud. X p. 4 n. 4. 2) Hillebrandt — Vedische Myth. II p. 76. 3) Ghurye — op. cit. p. 56. 4) Zimmer — A.i. Leben p. 119. the five-group is the image of totality, meaning a division into four groups, together with a fifth group as the group of totality. Five different tribes of importance in Vedic India need not be thought of. The interpretations of the commentators are not so wrong as Zimmer wants to have them. That this division into four is not a Brahmin falsification of reality becomes again evident from the fact that children of mixed marriages might be received again in the same varna after a number of generations. Oldenberg, therefore rightly doubts of Senart's conception in question.x) An antithesis between the varnas such as has been generalized into a racial antithesis, is very well conceivable within the limits of the tribe. Think only of the well-known phratry rivalry. Besides there is nothing commoner than an association of the different clans with contrasting colours. So we arrivé at the conclusion that racial contrast probably did not play such an important part as one might think, and besides that the immigrants have approached a possible racial contrast from their own standpoint. They would in our opinion rather consider the Dravidians as members of the lower phratry or as Dasas, potentially as members of the lowest clan they had themselves, if we suppose that they were at all implied in their considerations. In this light the difference between the Südra nirvasita (out-caste) and the real Südra (caste-fellow) must perhaps be seen. The four castes are consequently the four ranks that have sprung from the division of the tribe into four. We do not mean to say that with all the tribes the four patrilineal clans are called by the names of the four castes. This is possible, but not provable. We only say that the division into four was prevalent and that at the change of exogamy into endogamy the four castes or classes, whatever one wants to call them, have taken the functions of the four clans. From all this itfollows that the name Arya is not directly used for all the people that now go by that name. The name Aryans means about the same as if we called the English by the name of Lords. What is in a name? If from now on one wants to call the English: Lords, there is no objection, only one should remember that there are still other Englishmen than the genuine *) Z. D. M. G. LI p. 277- Lords. Also from this point of view there rises some objection against the use of the word Aryan to indicate races or languages. In the mean time history would have it that this loose term enjoys a reputation, so that its perpetuation will be granted. But one should bring into account that the despicable Südras were also „Aryans", though they were not „Aryas" in the Indian sense of the word. So it seems probable to us that there has existed in India a circulative marriage system with a four-clan arrangement. Besides we must take into account the possibility of a doublé unilateral reckoning of kinship. Let it be allowed toustoassume for a moment that the social structure really was thus, and that the tradition of the four root gotras is right.*) There are a l) The sketch of a patrilineal fourclassystem with exclusive c.c. marriage-system looks like this: r= '3 * i a ftl '* 7 * $ = $ ^~T ^ 4 7 ' 'i = 5 ^ ^ i s rt] and the same sketch, but now doublé unilateral: It is plain that every man has the patrilineal clan of his father (indicated byafigure)andthe mattilineal clan of his mother (indicated by a letter). One does not marry into the mother's matrilineal clan, but does marry into the mother's patrilineal clan, viz. with the m.b.d., net with f.s.d. For clearness' sake we shall give another scheme, to show that phratry-division is very well possible: few data that would possibly fit in with such a system, so that we want to adduce them as indications of the conception we propose. The first datum is the still unexplained fact that the Sapinda degrees ascend to the fifth generation. The ancestors, for whom the pinda sacrifice was meant, are the grandfather and the greatgrandfather. Sankh. observes with respect to sraddhakalpa that after a year the fourth parent must be left out. Similarly Parask. III. 10. 50, 51. „Mit „dem Vierten" ist gemeint derjenige der drei Pitaras, welcher nach der Zulassung des Jüngstverstorbenen der Vierte würde geworden sein für den Qraddhaverrichter, also der Urgrossvater des Verstorbenen. So sagt auch Manu (IX. 186): „Dreien gibt man die Spende, dreien die Pinda, der Vierte ist der, welcher sie gibt; der Fünfte hat nichts damit zu schaffen." Nach Mark. Pur. XXXI. 1 und Brhatpara?ara (D.S. II. 178. 5) wird nach dem Sapindikarana der Urgrossvater des Verstorbenen: „lepabhug"."1) The passage Mark. Pur., to which Professor Caland refers here, runs as follows (translation Pargiter): „After performing of the ?raddha to the deceased sapindas, he who is the father's greatgrandfather passes to (the class of) those who feed on the lepa,2) having lost (his share) in the pinda offered to the pitris. He who is the fourth there above among those who feed on the lepa bestowed to the (deceased's) son, ceases to eat (thereof) and obtains the satisfaction that he is freed from relationship." If, in the scheme, one counts proceeding from a man iA it is seen that with the fourth generation (that of the great-greatgrandfather) one again arrivés at the own clan of ego. The group of a person's pitaras consequently reaches as far as that. Then one arrivés at the clans of the sons, and in the sons, according to the well-known Indian notion, the father lives again. Secondly: regarding the regulation of marriage, we have already pointed out that sapinda relationship generally reaches to the fifth generation, if we include the ego's generation. We have assumed that sapinda relationship referred in the beginning to the relation founded on mutual services rendered by *) Caland — Proc. Roy. Ac. Amsterdam Dep. Lett. XVII p. 33. s) I.e. the wiping of the hands after offering the funeral oblations to the three sapindas (note by Pargiter). the clans of father's sister and mother's brother. The sapindas of clan i would then be e.g. the members of the clans 4 and 2. The scheme shows that after the fourth generation there come no new clans. All clans have had a turn by then. This seems to us an explanation of the fact that sapinda relationship reaches as far as the fifth generation. Now, we have pointed out above that, according to sapinda relationship a man may e.g. marry a woman, if they have a common parent to the fourth degree of the man's — and to the third degree of the woman's ascending line. But there is no doubt that a marriage of a man with a girl, who have a common ancestor to the fourth degree of the ascending line, is prohibited. The fifth man must never marry the fifth woman.*) It is probable that also this regulation, though strange in itself, should give expression to the aversion from a marriage with one of the same matrilineal and patrilineal clan, who might be found in the fourth degree. Finally: If at a sacrifice a person sums up his Pravaras (a groupofancestors) hemay mention one, two, three or five, but never the fourth of them. It may be seen from the scheme that with the fourth parent one arrivés at the ego's clan.2) From these indications appears a peculiar interest in the fourth degree of the ascending line, of which the marriage system we have pointed out, might give an explanation. Conclusive evidence of the existence of such a system has of course not been given here. But it is in itself not void of importance that the necessarily scanty data of India fit in with this marriage system, which is also known elsewhere. As the poorness of data is sure to remain a great drawback for all study of old Indian culture, we think we are justified to advance the possibility that we have pointed out above. Mandlik — op. cit. p. 354. ') Vaidya-Proc. First Oriënt. Conf. Poona; Ramogachari-Proc. Third Onent. Congr. Madras; Bhargava-Proc. Sixth All-India Conf. Patna. 7 II MYTH, RITUAL AND COSMIC ORDER. If it were not for the one thing that stands in the way of drawing up a chronological outline of the mythological contents of the Epic, the most obvious course of procedure would be — says Professor Hopkins—in the first place to treat the mythology of the Epic chronologically in its reaching back, on the one hand, to the Puranas, and in its being rooted, on the other hand, in the Vedas.1) The obstacle to which he here refers is that „chaos of mythology", the legacy of an older age, which has become more rather than less chaotic as a result of the tendency to generalize so characteristic of the Epic's compilers. In the oldest gods the primary naturalistic substratum is undeniably there, however thoroughly even gods like sun and moon came to be anthropomorphized. Animism and naturalism become merged in each other; and then another element crops up, the star-cult, the worship, in general, of the heavenly bodies, which must have had a certain influence upon the Epic, though it is not possible to decide as to what degree. How, in fact, would it be possible to do more than give a mere summary of the main elements of a system of mythology, which boasts of a pantheon in which „hope, heil and hunger, cows and corn, wisdom and the West etc. are called gods, constantly new images invoking new personifications". The lists contained in the Epic itself of the gods reveal striking differences. But few of the epic gods can lay claim to possessing any real significance. They associate freely with Rishis and also with spirits, good and evil; they are not even averse to cultivating the society of their counterparts, the Asuras. And in spite of the fact that three and thirty is stated conventionally to be the exact number of the divine occupants of this pantheon, the Epic speaks just as convincingly of deities in their thousands. l) Epic Mythology. Gr. I. A. Phil. III i B p. 3, 52. ■ Visualizing things in this way Hopkins did the only thing he could do under the circumstances: he devoted his energies to a detailed description of the mythical material of the MBh. Now, going according to one's own impressions of epic mythology, even with the systematic recapitulation furnished by Hopkins at hand, there is no making out how on earth the poets of the Epic themselves ever managed to find their way in this mythological „confusion worse confounded" without the aid of so felicitous a guide. The student, nevertheless, employs this knowledge of epic mythology as an aid to reconstructing the developmental history of the Epic, as is done, for instance, by starting from the figure of Krishna. The method pursued for the purpose of acquiring an insight into the mythology is practically the same as that employed for the purpose of making a reconstruction of the Epic itself in its oldest shape. One tries to discover the primitive myths by eliminating all those elements which are considered interpolations of a later time. It is naturally the myths around and about the more complicated figures which in particular lend themselves to such treatment. Generally one goes to work in such a way, that, the accidental features having been eliminated, a natural substratum is left, just as a natural substratum is assigned to the gods of the Vedic pantheon, as existing behind all their trappings, though no longer immediately evident to the eye, but revealed, all the same, in the derivations of their names. Consequently, in its main features, the history of Indian religion is supposed to be as follows: from out a more or less simple belief in the existence of gods still closely connected with natural phenomena, as represented in the oldest Vedas, there gradually evolves the more philosophic system of belief in divine revelation as manifested in the Brahmanas. In the course of time all religion became such an artificial thing, that in several parts of India repulsion and spiritual need resulted in religious reactions, such as Buddhism and Jainism. This „insuffisance de la vieille théologie brahmanique, dont les divinités s'étaient effacées peu & peu derrière des abstractions trop subtiles pour la conscience des masses" a) ultimately became l) Barth-Oeuvres I p. 141. the ,,rnison d'être "of Vishnuism and ï^ivaism, s£cts; which in contradistmction to Buddhism and Jainism did not break with the Vedic past; just the opposite, in fact: they maintained that they were continuing that past, nay more, they were that very past itself. Now, considering this idea of the evolution of the gods in Indian religion and studying the figures of the Vedic gods more closely, it becomes immediately evident, that even as far back as the Vedic age these gods were not merely worshipped as personifications of natural phenomena in an intellectual, naively scientific sense of the word. Even in the case of a figure like Dyaus it is not in the first place the sky that is worshipped but principally the All-Father, the begetter, together with Prithivi, the Earth-Mother, of the visible universe, and comprehending all created things within himself.*) The same thing is even more evident in the case of other figures, the names of which etymologists have so far failed, either wholly or in part, to explain. The etymological significance attached to the words Rudra or Varuna, for instance, leaves us practically in the dark concerning the naturalistic substratum concealed beneath these names. The preference given to the interpretation of myth by means of natural phenomena, the so-called nature-myth method of investigation, so notable in the study of the mythology of ancient India, also in the mythology of the Epics, is a legacy of the philological school of mythology with Max Müller at lts head. This school of so-called „comparative mythology" saw in the myths simply tales of the gods tacked on to the names of gods erroneously understood, once the relation between the name and the natural phenomenon it stood for had ceased to be clearly feit by people in general. Professor A. A. Macdonell, writing in 18^7, says that myths and tales of the gods and demons „have their source in the attempt of the human mind, in a primitive and unscientific age, to explain the various forces and phenomena of nature with which man is confronted . ) According to this scholar, then, a myth springs up when the imagination seeks to explain a phenomenon of nature as a thing 1) Macdonell-Vedic Mythology. Gr. I. A. Phil. III i A p. 22. 2) Vedic Mythology p. 1. done by, the action of, a personified being. Then, in course of time, the myth becomes detached from its naturalistic substratum, with the result, that, under the influence of „poetical embellishment", the natural phenomenon that gave it birth fades so completely into the background, that, often enough, a great deal of philological manipulation is required to detect it again. Now it goes without saying, that Macdonell's attitude towards the philological system of nature-myth interpretation is not in every respect an uncritical one. He can find no satisfaction in all this frequently quasi-etymological enigmatization, all this monotonous rooting in selfmade nature-myth puzzles. But in principle he does not disagree with the method of investigation he feels called upon to criticize. However, this old method is, in fact, hopelessly done for; it cannot be saved either by closer and more accurate examination or by severer critical judgement for the simple reason, that it is supported by such a flimsy foundation that it cannot possibly carry the weight of the demands that Macdonell puts upon it. We shall study this question somewhat more closely, the method of investigation into the nature of epic mythology being so inextricably bound up with it, since the naturalistic substratum looked for in Vedic literature, though not, in the proper sense of the word, epic, is, nevertheless the given foundation, as it were, upon which the entire edifice of epic mythology has been raised, thus determining in a not inconsiderable measure its architectural character. A number of objections arise from the outset. And the principal objection is not that which Andrew Lang, years and years ago, drew attention to, viz. the utter impossibility to reduce the wild confusion of so-called etymological „finds" to a safe and ordered communis opinio. We might indulge the hope that continued investigation will one day succeed in establishing a certain order inthischaosofnaturedeities,whereeveryimaginable phenomenon of nature is to be found personified, gods of wind and storm, of thunder and lightning, of light and fire, of sun and moon and stars, of sky, earth and water, and the like. The greatest objection we are up against is the peculiar attitude adopted by scholars who pursue this method of investigation towards the form of religion they have taken upon themselves to study. If we take as our point of departure Max Müller's conception of mythology as properly a form of thought so essentially determined by language that it may be described as „a disease of language", we can see in religion at most a system of irrational aberrations of the human mind, but never religion as such. That wonderful thing we call religion is not to be explained as owing its existence to a lack of rational insight into the operation of cosmic laws as determinative of physical phenomena. Religion is something apart, obviously a serious something that requires to be studied as such. Religion is not an inveterate error of the human mind resulting form a momentary sense of impotence over against the forces of nature and the consequent attempt to arrivé at a satisfactory explanation of such uncomprehended and, therefore, fear-arousing phenomena, a sort of short-sighted natural philosophy of a pitiful nature long discarded by the modern mind. It is very much more than that. The existence of religion in all its varying forms has been psychologically characterized by Dr. R. R. Marett as that sense of fear and awe that comes over the human mind in the „x-regions" of the supernatural. It is not: only a step from the province of philosophy to that of religion. Nowhere, and not in India either, has a philosophical phase of thought ever been found to exist at a time when uncomprehended physical phenomena occupied men's minds and religion was as yet unknown. What is continually being found is religion itself. It is inadmissable, therefore, for the purpose of explaining religion to go back to a pre-religious and, at the same time, prehistorical, philosophical stage of spiritual culture. And it is not merely one single element in the method of investigation that is touched — incidentally by our objection, we are here concerned with a difference of mental attitude and outlook upon which a great deal depends. The significance of the great change that the mental attitude towards the subject of religion has undergone since the last century can hardly be overestimated. We are no longer justified in approaching the study of this subject in the spirit of such as see in primitive religion merely a lack of knowledge decked out with fantastic imaginings. In discussing the subject of nature mythology Professor Dürkheim rightly remarks, that religion is simply being made a meaningless metaphor of by reducing it to a system of anthropomorphized musings upon the activities and manifestations of nature.1) And wherefore should all such errors have continued to exist even when it had become evident to all thinking individuals that they were scientifically untenable ? Religion cannot possibly adopt a now defeated attitude, otherwise it could never have resisted the encroachments of science. We are not here concerned with a question of appreciation, proceeding from the peculiar religious ideas of the individual investigator; it is with differences of a more concrete character with which we are dealing. And this is of the greatest consequence to the practical study of the MBh. To see in myth a small nucleus formed by the purely religious contemplation of natural forces and phenomena around which a proportionately much larger quantity of poetic embellishment has in the course of time collected, the original mythic tale being thus more or less covered up beneath a promiscuous weedy overgrowth, fantastical clingings to the names of gods that had gradually assumed the character of veritable enigmas, to see myth in this light, shorn of its predominantly religious values, is, in fact, simply the same as saying, that myths are „développements parasitaires, qui sous 1'influence du langage seraient venus se greffer sur les représentations fondamentales et les dénaturer".2) In accordance with this idea of myth it is obvious that the next thing to be done will be to endeavour to get down to the original narrative by removing all the fanciful ornamentations that were later added to it. Then, naturally enough, for lack of objective criteria, recourse will be had to the employment of aesthetic norms, and attempts will, in the very first place, be made to do away with the unaesthetic and offensive elements in the myth in the hope of thus being able to detect its primary form. It is in this primary form that one expects to see the figure of the god in its original shape; and this „disinterred" figure of the god will serve in turn to assist the philologist in making hiscriticalanalysisofthetextas the goldsmith's testing of allo ys isfacilitatedbythe use ofhis touchstone. J) Les Formes Élémentaires de la Vie Religieuse p. 100—123. *) op. cit. p. 115. This method of mythical study is unacceptable. For, as Dürkheim rightly observes, if a distinction is to be made between a religious and a non-religious element in myth, the same course must be fcllowed in respect of the rite belonging to the myth, Robertson Smith having shown that myth and rite are closely connected with one another. The nature of the rite, of the cult, is dependant upon the peculiar character of the god worshipped in the cult. Frequently the rite is simply a dramatized form of the myth. So if the myth is to be looked upon merely as a fanciful growth round about a simple central religious idea, the same thing must be done in the case of the ritualistic practice so inseparably connected with the myth in its original form; virtually, then, under the circumstances, the practice of ritual in general must be seen merely as a sort of „pastime", resulting from „a mood of playful and erratic fancy", otherwise we lay ourselves open to the charge of being arbitrary. Ritual is worship organized, the detail and circumstance of adoration, and its practice concerns not the phenomenalistic substratum but the person of the god himself as he is known to exist in the myth. If we see in the Kamsavadha the conflict between Krishna (the obscured sun) and darkness, an explanation must likewise be forthcoming as to why that darkness should have come to be considered the uncle of the sun. In this respect nature mythology together with the conception of myth attaching to it is likely to conduct us into perilous ways. And why ? Simply because religion cannot possibly have emerged from a primitive philosophy of nature. No religion was ever born out of the scientific observation of physical phenomena. To aim at achieving a fine discrimination between a religious and a non-religious factor in the „make-up" of myth is entirely unmotived, for myth is in its entirety an object of faith. That means to say, that in studying myths one will do well to begin by not attaching much value to one's own aesthetic appreciation. If we want to know the religion of a people then, all we have to do is to look for the religious tales and use them as steppingstones to the acquirement of that knowledge. Alongside the religious tale or myth there are sometimes tales of a profane character, and not very reliable as such for the study of the religion, but they are to be recognized by the very fact, that they are non-sacred, often enough in practice by their being not related to the ritual.1) We must not first look for a religious element in myth in order, thus, to acquire a knowledge of the religion it enshrines; we must be primarily concerned with what myths really are. In its study of myth as so closely associated with ritual, nature mythology can boast of but a very unsure foundation. The presence of a naturalistic substratum in myth is inferred by almost every investigator, according to his own particular preference, frequently from a rather questionable, etymological explanation of some name. The factor of religious belief has been entirely neglected. What was it that people believed? It is not the naturalistic substratum of a myth that people believe in, but in the myth itself as a whole; they do not split up a myth and then proceed to analyse it,* they simply take it as it is in the form in which it has been handed down. As such it is an object of belief. No, it is evident, that Marett was right in laying particular stress upon the sense of awe as one of the leading elements in the religious consciousness of primitive man, not that religion sprang from fear, but that fear was an enduring element in religion. And not all the „awe-inspiring" phenomena in themselves have been assimilated into the religious consciousness. We find exactly the same sort of differentiation made in primitive states of society between the religious and the commonplace, between the sacred and profane as we ourselves make. Even though we agree with Marett, therefore, that religion is not to be explained as resulting from an imperfect understanding of nature's manifold manifestations, we still have to ask ourselves why it was that one „awe-inspiring" phenomenon was looked upon as being quite ordinary, whereas another was given a religious significance. Why should one story be a myth, a sacred tale, in one place, and, in another, a fairy-tale, a profane story? Religion without doubt belongs to the domain of the supernatural, but supernaturalism is not always religion. Moreover we must take care to preserve a sense of proportion in this connection; the sense of awe, the feeling of impotence over against the „awe-inspiring", was not so predominating a thing 0 Dürkheim-op. cit. p. 117 n. 2. in primitive religion as we might be inclined to imagine, for it is frequently a certain sense of possessing power over nature rather than a feeling of simple piety which is at the bottom of ritualistic practices. In this position of affairs the safest thing we can do is to adopt the distinction made by the French sociologist, Dürkheim, between religious and non-religious as being the difference between the sacred and the profane, and so to take religion as it manifests itself, and to apply, in principle, to religion neither psychological nor nature-myth standards of judgement x). In the domain of ethnology the point, in so far as the nature myth method of interpretation in particular is concerned, has been carried. The fact of a certain partiality being manifested in Indological studies for the use, in a moderate degree, of the nature myth system of interpretation, even though one is not concerned any longer with the problem of a general naturalistic substratum is to be accounted for not only by a somewhat exclusive philological interest in the subject, but also, in part, by the peculiar character of the Vedic texts themselves, which are frequently so deeply religious and at the same time so profoundly human, that the association of myth with ritual, the mainstay, therefore, of mythical study, is not immediately manifest. So, too, is to be explained, for example, the lack of approval that waited upon the attemptsof Professorvon Schröder in 1908 and 1915 at explaining a number of Vedic hymns by means of ethnological data. Concerning von Schröder's labours it must be said, that they were undoubtedly distinguished by a certain acumen, but also that they were of a rather speculative nature, for the very cogent reason that there is nowhere any evidence as to the hymns thus treated really fitting the ritualistic practices he describes. But with the single exception of a man like von Schröder, whose keen and versatile interest also went out to the study of Animism and Primitive Monotheism, the science i) We do not mean to say that our discussion is entirely independent of oreconceived ideas and that we are the first to „let the facts speak for themselves". But in the present work, in which theoretical speculation would be out of place, we have adhered to the distinction, in our opiruon, essentially right, drawn by the great French sociologist, Emile Durk heim; there being no need then, at any rate, to hark back to the obscurity of prehistory or to tread the maze of bewildenng psychological subtleties. of ethnology failed to win even the minimum of attention of the large number of Sanskrit scholars; even names held in such high repute as those of E. B. Tylor and Sir James Frazer being only now and then incidentally mentioned. And yet it is quite evident that the religion of ancient India must become as incomprehensible as any other great religion once it is disconnected from the more primitive phases of religious life which are studied by ethnology. It is one of the great merits of the late Dr. Nathan Söderblom, archbishop of Upsala, to have conclusively demonstrated this in his work, „Das Werden des Gottesglaubens". This scholar distinguishes three main aspects in religion: the belief in souls; in a non personal super-natural power; and in a creator. The relation between these three elements determines the character of a religion. In India it is the belief in a non-personal power that constitutes the main element, the belief in Mana, discovered by R. H. Codrington, and which in India is expressed by the word Brahman, the ritual formula that constitutes the power of those beings who are able to employ it as, for example, the Brahmin onearth.1) Later on this Brahman developed into a philosophical idea, the basic principle underlying an idealistic monism which could no longer satisfy the cravings of the heart in its eternal seeking after spiritual satisfaction. The process of religious development as worked out by Söderblom is all too simple and straightforward. At all events it cannot be accepted in the form he gave to it. His conception of the Supreme Being is too much the result of intellectual construction. This is particularly apparent from the work of K. Th. Preusz. The ritualistic formula among the American Indians, whom Preusz studied, endowed those who were able to ') „Das Brahman ist das Fluidum oder die Potenz geistlich-zauberhafter Macht, samt ihrer Verkörperung einerseits in heiligen Sprüchen und dergleichen Riten, andererseits in dem Stande der Brahmanen, welche jene Macht besitzen" (Oldenberg). In the works of Oldenberg Geldner, Osthoff and Hillebrandt, Söderblom finds the selfsame definition given of Indian Brahman as of Melanesian Mana, so there can be hardly any doubt as to the identity of these twin ideas. Oldenberg, however, who kept abreast of the ethnological thought of his times (R. d. V. p. 35, ssq.), supposes Mana to be the supernatural dynamic power and Brahman essentially the ritual formula, the two ideas, therefore, not overlapping. They are undoubtedly indentical, though. (Lehre der Upanishaden, p. 49). employ it with the power andimposeduponthemtheobligation to regulate, to a certain extent, the forces and phenomena of nature. This ritual formula originates from the Supreme Being, which means that Man, by meticulous compliance with the prescribed rules of ritual and, hence, by his maintenance of cosmic order, serves the Supreme Spirit itself as one and coeval with that order. Ritual, therefore, is not a means of approaching, in the first place, the Supreme Deity from whom it derives, but the other gods, which does not, however, mean to say, that the Supreme Being must be understood as being entirely a deus otiosus. So we can say of the Primeval Being that it retired, after having created, established and fixed everything, into itself. But the Primeval Being is none the less a god, which means that he must not be considered the prima causa, deliberately devised by a conscious and systematic intelligence, but the source of life, experienced and acknowledged by the believing heart. The gods are not free to control the forces and phenomena of nature to their own end, but have been given a place in the cosmos by the „Urheber". Just as human beings they must uphold the order of the world by means of rites. Considering the results of the investigations of Preusz and Söderblom together, we are struck by the nicety with which they fit as a whole into the idea of Brahma as the All-Father, one and coincident with the Cosmos, from whom proceeds Brahman, the ritual formula, the ritual itself, being a nonpersonal power, by means of which the priest and, in general, holy persons are able to exert their influence upon the forces and phenomena of nature and even upon the gods themselves, likewise engendered by Brahma in the Cosmic order. This identification of ritual, power and deity is especially striking in expressions such as: „the sacrifice is Brahman" and: „the sacrifice is Purusha". „The sacrifice rejoiceth when one that knoweth, draweth nigh". The various parts of the sacrifice correspond to the various parts of the human being, to the various gods and the powers of nature 1). In the Bhagavadgïta we read that everything that lives comes from food and food i) de la Vallée-Poussin-Indo-européens et Indo-iraniens p. 261. from rain; rain from the sacrifice, and the sacrifice from Karman (the ritual act); Karman comes from Brahman; Brahman from akshara; hence the omnipresent Brahman is continually to be found in the sacrifice 1). So we see that ritual is closely associated with the Supreme Being, as appears from the word Brahman attaching so closely to that of Brahma, and that the Brahmin controls the cosmos by means of ritual, thus becoming the wielder of that impersonal power, expressed in the words, mana and Brahman. The two meanings of Brahman, „word" and „the powerful" meet in the ritual.2) Now to ask oneself which is the primary meaning, „word" or „powerful" is altogether wrong from the ethnological point of view. The ideas of power, supreme spirit and rite are so intimately bound up with one another that there can hardly be any talk of priority. The ideas of power and supreme spirit, Brahman and Brahma are kept too far apart by Söderblom. Naturally the three elements, power, supreme being and ritual do not always and everywhere appear in the selfsame relation to one another. In India impersonal Brahman seems to be more in evidence than personal Brahma, especially during the period of the Brahmanas, when ritual and, as a result, the position of the Brahmins was the predominating influence in religion. Hence the rise of Vishnuism and Sivaism may be seen as an explicit confirmation of the „insuffisance de la vieille théologie brahmanique, dont les divinités s'étaient effacées peu è peu derrière des abstractions trop subtiles pour la conscience des masses" 3). We are therefore — it is true — justified in speaking of a leaning towards theism, manifested in Hinduism *) Bh. G. III. 14. Also vide Söderblom-op. cit. p. 259. 2) Oldenberg's conception of Brahman as being the magie ritual word which later on developed into Brahman as the idea of supernatural dynamic power, from which Brahma eventually proceeds, was influenced by the so-called pre-animistic theories, for instance, those developed by Marett. In the time immediately following upon its discovery the idea of Mana was given a decidedly too vivid theological colouring, which explains how Oldenberg was so easily led to associate it with the theological ideas underlying the speculative philosophy of the Brahmanas. In our opinion the Brahman-atman speculations constitute, indeed, the beginnings of a system of theological philosophy, but there is nothing which suggests a pre-animistic substratum underlying that philosophy. s) Barth-Oeuvres I p. 141. proper, althought we must bear in mind, when employing such a term so frequently suggestive of the theology of the West, that power and supreme spirit belong together, and that, in spite of the difficulty of determining whether there is such a thing at all as a supreme spirit in the Vedas, and if there is, who it is. We must not forget that the Vedic hymns can do but little to increase our knowledge of Indian mythology as such. So there may have been a Supreme Being that the Vedas did not mention 1). In the MBh. Brahma, the supreme deity, in any case appears as the „Urheber", without otherwise occupying an exactly predominant position.2) Brahma is atmabhü, svayambhü, selfexistent. He is denoted in a number of terms as the creator and lord of the world, even though he generally leaves the positive act of creation to another divinity. He produces spiritually „mental sons" and they, in turn, become the begetters of all living creatures. The highest appellations are bestowed upon this lokapitamaha. And yet there is a certain dualism in his character resulting from his impartiality. He is not the father of the gods alone, but likewise of the demons, and, as such, he bestows the boon of invulnerability upon gods and demons alike, both parties, thus, being urged to do all that lay in their power to bask in the sunshine of his favour. He is the same for all, thus being equally well-disposed towards gods and demons alike. The demon Ravana derives his power from him. He gives to Ravana's son the brahmastra by means of which he catches Rama, and goes a step farther by giving him the wherewithal to conquer Jambavat, Brahma's very son.3) i) As to Dyaus as a Supreme Being vide L. von Schröder-Arische Rehgion expiain Brahma by means of the nature-myth method of interpretation must from the outset be doomed to failure; he will not thus be explained. „Brahman lacks the hold of tbe naturegods upon the popular imagination . The highest gods of a pantheon are seldom entirely abstract, Blahma, at any rate, being associated with heaven and the gods. The supreme deity of theKora Indiansis associated with the moon, the earth and the underworld. According to Hopkins the Brahma of the mythological pantheon evolved form the philosophical idea of Brahman, „the universal world-power , invested by „the personifying power "with,, figurative anthropomorphism . (bpic Mytn. p. 189). 3) Hopkins-Ep. Myth. p. 189, 195. We shall later on dweil at greater length upon the relation between ritual and world-order, but this much at any rate is evident enough already, that the myths which contain a detailed account of the manner in which the power of Mana operates in ritualistic practice, may be deemed to furnish us with an insight into the nature of cosmic phenomena, and that from such myths, therefore, we can become acquainted both with the ritual and with the happenings in the domain of the sacred. Myths do not, therefore, transport us into a purely imaginary heaven, but conduct us directly or indirectly to the spot where the rites are celebrated. Andrew Lang long ago drew attention to the fact that various myths lend themselves admirably to being interpreted as explanations on the sacred plane of some social arrangement or other. The heaven spoken of in myth cannot be better known than by a knowledge of ritual and what is meant by it. „Heaven" then, according to the conception of it in primitive thought, is not situated somewhere in the air, in primitive thought „heaven" is the world itself, merely that world in another manifestation,a non-profane, a sacred world. The men and women in myth are real, living human beings, but they are dwelling upon another plane, that of the sacred. It is far better and far safer to make with Dürkheim a distinction between a sacred and a profane world, a sacred world in which the power of Brahman is employed and the substance of the myth enacted, and a profane world in which life is lived in the ordinary human manner and its more commonplace sensations experienced. The difference, therefore, between a sacred and a profane world is not a mere difference of place but essentially a difference of mental attitude adopted by those that profess the form of religion manifested in the myth towards the twofold nature of the world in which they live. We might say that the sacred world i s the profane world, but then entirely different in the mind of him that conceives it. Mythical personages, therefore, by reason of their acting in the region of myth, become metamorphosed into beings of a more or less sacred nature. „Le sacrifice transforme le sacrifiant en dieu" 1). That does not mean to say that everybody else who is not *) de la Vallée-Poussin op. cit. p. 263. immediately connected with myth is to be considered entirely profane, for every human being has something holy within him; but holiness can become so dangerous a thing, that it can only be operated after long and scrupulous preparation on the part of the „saint", as, indeed, is done in ritual. Indian ritual is so detailed and elaborate, that one can only expose oneself to the dangers attendant upon the celebration of a rite after a long and careful preparation; for it is by means of Brahman, the formula and the rite that the operator controls the cosmos. Everywhere the most scrupulous care was taken to prevent mistakes being made in the performance of sacred rites, the harmony of the universe being more or less dependent upon their faultless celebration in detail, with the result that the celebrant in Indian ritual ultimately found himself exalted to the dignity of lord of the world. No holier, no more dangerous beings than the celebrants of the religious rites, the Brahmins, who were held in fear by gods and men alike. Also the persons figuring in myth are endowed with the same sacred character. They are entirely or in part either gods or demons. As a rule the action of a myth is also sacralized as to time, i.e. it is made to take place in a mythical age when other conditions of life prevailed. Such, too, is the condition of affairs in the Epic. Although the persons figuring in it are not exactly gods themselves, the Pandavas, for example, are nevertheless represented as the incarnations of Dharma, Vayu, Indra and the Asvins; Duryodhana as the incarnation of Kali. From as early a source as Panini we know that in his own time already there existed a sort of Arjuna-Krishna cult. There is not the slightest reason to endeavour to explain this deification in a euhemeristic way. The MBh. wasvenerated from of old as a sacred text. It is, moreover, selfevident that the Epic plays in a mythic world, no distinction being made between heaven and earth, at least one is not sensible of any distinction being made. We see the human dramatis personae in this socalled heroic narrative so freely associating with gods and demons, yea, even dweiling in the divine mansions, that we are bound to conclude that the scene of the Epic must represent the sacred world. For, whatever we may do, whether we consider the Epic as the product of a humanization of divine tales or of a deification of human, we always knock up against the same, apparently insurmountable, difficulty with which we are confronted when taking the nature-myth method of interpretation as the basis of our investigations into the origin and development of this labyrinthine poem, and that difficulty is, that the transition from „playful and erratic fancy" to sacred story and vice versa cannot take place so easily as people seem to think. And if, moreover, — which we dare say nobody doubts — the Epic of the Bharatas may be said to attach to laudatory poems sung on the occasion of the Imperial Sacrifices (as supposed by Professor Weber), then we must needs admit that the more intimate connection between Epic and ritual is thereby already in fact established and the mythical character of the MBh. assured from the very oldest times. Further, seen chronologically, the Epic in the eyes of the Indian is not merely a literary product of some centuries B.C. but a sacred narrative of events which took place in the age of myth. The action takes place between the Dvapara and the Kali age, the latter being the age in which the world now lives. Now there are those who will contend, that the Epic in its war-narratives is certainly a reflection of actual events, though perhaps somewhat fantastically conceived. Our answer to this objection is, that such is indeed the case in the myth also, attaching as it does to the ritual. The relation of the myth to reality is vouched for by the ritual connected with it. In this sense the myth attaching to ritual is a narrative of actual events. We mentioned in our preface, that the study of the MBh. has led scholars to distinguish a narrative and a didactic element. „Personne ne doute qu'il y ait au fond du Mahabharata un admirable fable épique" 1). If our own conception of things is right, then, we must look upon the MBh. as being not „an epic narrative" but „a myth". This is of primary importance in the critical study of the poem, an analysis of a purely literary nature having no grip on the Epic. There is no question of the religious element having in some unaccountable manner found its way into the poem at a later date; it constitutes an essential part of the poem from the very beginning. And even if we are *) Barth-Oeuvres IV p. 358. 8 not willing to admit it, it must at any rate be considered an entirely unjustifiable proceeding from a methodologie point of view to remove the religious kernei from its epic husk or vice versa to separate the religious husk from the epic kernei. In this manner the idea of religion simplygoestothe wall. Hence, it is not a literary but a religious analysis that is here required. * * * We venture to presume that it is now clear, that the present method of interpretation, still more or less akin to that employed by the school of Max Müller, must be deemed unreliable and even dangerous, and that, right away from the start, by reason of the faulty mental attitude towards the textual material. We have shown in short, that the results of ethnological study, the discovery of the mana conception, and of the belief in a supreme being and the like, must needs be of far-reaching importance in the study of Indian religion and myth. It has become manifest that Brahman and Brahma, that mana, the „Urheber" and ritual are closely interrelated, these relations being explained and accounted for in myth. It is mythic material that we have before us in the Epic, hence we must try to reconstruct from it the conception of the universe peculiar to the epic period. There being, as we have already said, no illimitable gulf fixed between the sacred world of myth and the profane world, we are able by study of the mythic material at our disposal to acquire at the same time a certain insight into the organization of human society as it really existed in the period of the Epic. It is not the fantastic imaginings of a super-religious mind with which three quarters of the MBh. is filled. When we speak of Brahma as the creator of the universe and the laws that govern it, the establisher of the cosmos or even more or less its substitute, the word cosmos is not to be understood as implying a Western systemof philosophy, but an ordering of thingsaccordingto a system of classification peculiar to India, in the manner of the classificatory systems dealt with by E. Dürkheim and M. Mauss in their well-known essay1). The things with which primitive man comes into contact are classified by him in a manner different to that which we are !) De quelques formes primitives de classification-Année Sociol. VI. accustomed to.1) Atfirst sight itisextremelydifficulttounderstand, for example, how on earth an American Indian tribe studied by Preusz could ever have come to identify a certain cactus, the pejote, with the deer. The classification becomes quite plausible, however, once we get to know something about the ritualistic practices of these tribes. The various things different clans are constantly in touch with, being generally shown by their social grouping, in most cases determined by the position occupied by them in the ritual of the tribe, the social organization often constitutes the basis of the system of classification. But the classificatory system does not always seem to be regulated in the first place by the peculiar organization of the clan, the place in tribal gatherings allotted to the various clans with respect to the points of the compass also frequently serving as a basis of classification. But it is only ostensibly that another basis is here adopted. For when the tribe gathers the clans are so encamped, for example, as to give the celestial clan a place in the direction where, according to the system of classification, the celestial regions are situated. The Omahas, for instance, are divided into a Northern phratry, with which thunder and lightning, sun and sky and day and the masculine sex are associated. This is the phratry of the „sky people", and a Southern phratry, representing the feminine half of the cosmos 1) An example of a form of classification based upon the social structure of an Australian tribe in Dürkheim's Les Formes Ëlémentaires de la Vie Religieuse, p. 202: PHRATRIES CLANS CHOSES CLASSÉES DANS CHAQUE CLAN L'arbre a bois noir, les chiens, le feu, la glacé, etc. Le kangourou, 1'été, le soleil, le vent, 1'automne, etc. and associated with earth, moon, night and the female sex. This is the phratry of the „earth people". In the tribal camp, constituting in itself the entire community, it really does not very much matter whether a classification according to the points of the compass is adopted as a startingpoint or one according to the clan, the various clans being allotted their places in the camp-circle according to the points of the compass.1) In India it is the classification according to the points of the compass which is of especial importance. In numbers of places various things and divinities are classified according to the points of the compass, a classificatory system so familiar, that it may be deemed redundant to further establish proof of it by the arraying of detailed evidence. As far back as the Rig-veda mention is made of the four points of the compass. By the „four quarters" together is meant the earth as a whole, five points being occasionally mentioned in the same sense, i.e. when the centre is added.2) The idea of the dikpalas as guardian gods of the quarters of the firmament is familiar enough.3) Thus we also find in the Epic a classification of the four nereditary classes or castes, the Brahmins being associated with white, the Kshatriyas with red, the Vaisyas with yellow and the Südras with black 4). Now Professor Oldenberg has shown that underlying the conception of world in the Brahmanas there must be some sort of classificatory system 5). But he does not attach to his discovery 1) Among the Hereros of Africa the village is built in the shape of a circle, the southern half of which is inhabited by the men and the northern half by the women. Situated in the east is the hut of thechiefand his wife. In the centre is the sacred tree, the shed for the sacred cattle and the place of the sacred fire. This classification is, of course, not primarily the result of hygienic or rational considerations but of mythological. H. G. Luttig — The Religious System and Social Organization of the Herero (Leyden 1933) p. 33- 2) Macdonell-Ved. Myth. p. 9. a) Prof. Vogel gives several examples of this manner ot classitication in his „Indian Serpent-Lore", p. 198; 9, 118, 212, 226. In Gobh. G. S. IV. 7. 27—41 the following: Indra (East), Vayu (South-East), Yama (South), Pitaras (South-West), Varuna (West), Maharaja (NorthWest), Soma (North), Mahendra (North-East), Vasuki (Down), and Brahma (Up into the sky). «) MBh. XII. 188. _ „ uf.ru 5) Die Weltanschauung der Brahmana-Texte. Vorwissenschaftliche Wissenschaft 1919. the importance which it, in our opinion, possesses. In the essay by Dürkheim and Mauss, which we have already mentioned, likewise mentioned by Oldenberg, it was there for him to read that there is here no question of „ursprünglich formlosen Superstitionen", destined later on in the Brahmanas to take on a „halb-wissenschaftliche Farbung".1) And yet Oldenberg emphatically says that the identifications of the Brahmanas are far from always being the result of playful inventiveness „sondern tiefer in mythologischen und rituellen Vorstellungskreisen begründet"2). But Oldenberg does not take those „mythologische und religiöse Vorstellungskreise" seriously enough, it being the very same peculiar mental attitude towards myth and ritual, which also played this great scholar false. The sub-title he gave to his work, „Vorwissenschaftliche Wissenschaft", speaks volumes. It is not incorrect, but it makes us think rather of something that is not yet what it was to be than of something that has become what it now is out of what it once was. „Prescientific". Sure enough. But „post-religious", too, then. By these identifications is meant: that the things identified with one another are born out of one another; or belong to one another; or govern one another; or constitute one another's potency. The relationship existing within the classificatory system between one thing and another is expressed by the word rüpa („form, outward appearance, colour, shape, the right form i.e. beauty, character, peculiarity, play").3) Rüpa is not simply the form of a thing, for „eine Text sagt, dasz ein Rüpa da ist, so lange Atem ist; wenn aber der Atem entweicht, wird man zu ein Stück Holz; und unnütz bleibt man zurück als kein Rüpa. (J. U. B. III. 32.)" 4). The animals are said to have put their *) op. cit. p. 56. 2) ib. p. 112. 3) It is not quite clear, how Oldenberg explains this use of the word, rüpa, to which he himself draws the attention. He draws a distinction between the rüpa, „das Wesen von Seiten seiner Erscheinung oder kurzweg seine Erscheinung" and tanü, „das von Seiten seiner Körperlichkeit betrachtete Wesen". Bezeichnend ist der beide Woite verbindende Satz des Rigveda über Indra, dasz er rüpa auf rüpa wird, Zaubergebilde schaffend um seine tanü (III. 53, 8. vgl. auch X. 112, 3; 169, 3.); da ist tanü das Wesenhafte, rüpa das Blendwerk der Erscheinungen." op. cit. p. 109. *) ib. p. 105. rüpa away. (S. Br. IV. 3.4,14.) *)• That does not mean to say, of course, that the animals were, therefore, considered to be an „ungeformte Masse". That would be mere abracadabra. Perhaps it was the purpose to show, that the Brahminical theologian who is holding forth is excluding animals from the sacredworld of the classificatory system. The sacerdotal speaker has entirely outgrown the totemism of the classificatory system, but the system as such has him firmly in its grip still. By the creation of the cosmic order ancient Indian literature means the giving of rüpa to the elements. „Von altester Zeit an erscheint rüpa haufig zusammen mit naman in der festen Verbindung, „Namen und Gestalt", (auch in umgekehrter Rïihenfolge). Dies sind „die beiden groszen Riesenwesen des Brahman", mit denen jenes, wie ein Brahmana sagt (£. B. XI. 2.3.), in die Welt eingegangen ist"2). Here, therefore, we see clearly enough that the primeval being is identified with the classificatory system through his identification with the ordered universe; the cosmic principle, the fundamentais of order in heaven and earth and all that is in them, and by means of which things are grouped together in their mutual relation to one another, being Brahman itself. The Indian conception of the act of creation, therefore, is the classificatory arranging of things according to certain norms. When there is talk of creation we must beware of considering that idea in the light of western philosophy, thereby laying undue stress upon the act of creation as generation out of nothing instead of on the arranging, the ordering of primordial matter. The Indian speculative mind was not primarily concerned with the origin of primordial matter as such. The primeval being is already there, being selfexistent, or he is born out of an egg. However that may be, there was never a time when nothing existed. Something ever was, and Brahma merely introduced into that something the various rüpas. The rüpa is not a philosophical but a mythological idea. x) ib. p. 104 n. 4. .. . . . .. 2) Noch ist viel von der alten Vorstellungsweise lebendig, fur die der Name, weit entfernt davon zu einem blosz konventionellen Zeichen abgeblaszt zu sein, auf das Engste mit dem Wesen der benannten Sache zusammengehört, geradezu ein Teil dieses Wesens ist.... ich weisz dein Namen, sagt man zu der Wesenheit, über die man beim Zauber seine Macht betatigen will". op. cit. p. 103. „(Ch. Up. VI. 2.f.f. erzahlt), wie das Urwesen, das „Seiende", in die drie ersten Geschöpfe, Glut, Wasser, Nahrung, eingehend, in ihnen „Namen und rüpa" aus einander legt. So entstehen Feuer, Sonne, Mond, Blitz, in denen jedesmal ein rotes, ein weiszes ,ein schwarzes rüpa enthalten ist: das sind die ïüpa von Glut, von Wasser, von Nahrung" x). Things were not seen as a rule in an isolated condition: „(Rüpa) pflegt, ebenso wie tanü, haufiger von der Beziehung zweier Wesenheiten als vom Dasein einer einzigen gebraucht, die eine als Erscheinungsform der anderen zu bezeichnen" 2). The knowledge of the classification of things, i.e. of the rüpa and the naman, of the mutual relation of things one to another (virtually, therefore, the knowledge of Brahma himself, who is made up of these two vast entities) is not in its entirety to be acquired without more ado by anybody and everybody. One is known, the other is not generally known. „Die Brennholzer in denen Agnis Wesen gegenwartig ist, sind sein verborgenes rüpa, Agni selbst ist das offenbare: neben der alltaglichen Ansicht der Dinge tut sich da der Bliek in ihre geheime (X'dnung auf" 3). Frequently they go the length of considering, a knowledge of the mythical significance of the ritual identical with the meticulous performance of the ritual itself.„Hewhoknoweththis (ya evam veda), hath immortality". One wonders whether the ritual act (karman) is not dependent, for instance, upon „la connaissance (jnana), et non plus la connaissance des identifications relatives au sacrifice, mais la connaissance d'un être suprème dont 1'homme est une émanation, un fragment ou un substitut amoindri et aveugle: Quisque Deum intelligit Deus fit. On réservera 1'immortalité au „vrai brahmane", au savant, k 1'ascète: on condamnera k la transmigration tous les êtres, dieux compris, „qui ne savent pas".4) The reading of a text comes to be considered identical with the celebration of the sacrifice of which it speaks. Though one be lying upon a soft couch, annointed, clothed in full pontificals and having had one's fill of food and drink, if only one studies one's text, the sacred fire shall continue to burn within one, even to one's very *) op. cit. p. 105. *) ib. p. 105. 3) ib. p. 107. 4) de la Vallée Poussin-Indo-européens etc. p. 263. finger-tips.x) Sacred knowledge, the knowing of the cosmos, of the ordered universe, is identical with the performance of the ritualistic ceremonies. By the rüpa, therefore, we must understand the entire mythical complex of numbers of things grouped around the thing started from. By the creation of rüpas the Heavenly Father establishes order in the chaos of orderless things. They are classified according to their naman, their place and function in the system of which they form a part. Hence naman and rüpa may be deemed to coincide with Brahman, Brahman being coincident and coeval with the cosmos, with the world as an ordered whole. This association between ritual formula (Brahman), cosmos, classification and Brahma is evident with uncommon clearness from S.B. VI. i, 2, 6. The Creator creates by „copulating with Vach by means of his manas"2). Vach is the Word, and can hardly be anything else but the ritual formula. The same thing is just as clearly evident from another passage in which the sacrifice is said to have thought: „That word is woman; I shall call unto her, and she will certainly call me to her". He, therefore, that knows the classification of things and hence the universal order, such a one has power over the universe. „Ohne weiteres ist es klar, dasz es sich um mehr als blosze Symbole handelt. Für das Bewustsein der Alten zeigen sich hier reellste Wirklichkeiten und Wirkungen. Ueberall ist es ja eben darauf abgesehen, durch geschicktes Ausnutzen der Identifikationen Wirkungen zu erzielen" 3). Among the characteristic things that immediately strike one in the classificatory system, we shall mention, in the first place, the peculiar preference enjoyed by the number five. In the Vedas we find the word „Panchajanah" denoting all the peoples of the world understood as a whole. The five (four) points of the compass represent in the Vedic scriptures the entire world. In epical philosophy man has five „breaths". Creation is made up of five mortal and five immortal elements. The five immortal elements are termed the fivefold breath. A remarkable thing is, that the names of these five „breaths" are constantly M Weber-Ind. Stud. X p. 112. 2) Oldenberg op. cit. p. 80; p. 84 n. 4. See: Weber-Ind. Stud. IX p. 473. Vac und lióyo; 3) Sparing bij me. H. used in a different sense. From this we do not conclude in the first place, that at that time the conception of the spiritual elements must have been established though their fixed appellations were still wanting, but that in the eyes of the theologian of the Brahmanas the classification itself was of greater importance than the „realities" to be classified. His only concern was to arrange all sorts of things in all sorts of „fives" after the example of the four clans or castes together with the whole four taken together and composing a total fifth, with the purpose of showing that this ideaofthe„fivefold" was to be understood as constituting a perfect whole in itself. Another remarkable thing is, that the manner of classification was more or less variable, the various things not being once for all grouped together in an unalterable association of ideas. In one instance a certain thing is allotted the highest place, in another instance the same place is given to a different thing, the first then being put lower than the second. It continually depends upon what other things a certain thing is classed with. This relation, then, of one thing to another is the very same as that which exists among the clans in a circulating connubial system, each clan in its turn being „male" or „female". The cosmos created by Brahma, comprehending, as we have already said, a form of human society based upon a clan-organization and a circulating connubial system, we can understand that the classification of things should have undergone the selfsame peculiar shiftings of relative values which determined the place occupied by the various clans in a circulatory connubial system in respect of one another, each clan in turn being at a given moment higher in the classificatory series than one clan and lower in the same series than another. It is, as we have already said, a peculiar thing in clan-organization, that each separate clan considers itself as constituting the unity of all the clans together, everything that is especially associated with each clan in particular thus more or less comprehending everything else that is associated with all the other clans together. „In dieser Weise wird gesagt: „Regenzeit fürwahr sind alle Jahreszeiten... Die Regenzeit ist ein rüpa aller Jahreszeiten (Q.B. II. 2, 3, 7)"1). We may go on drawing the same parallel with the organization ') Oldenberg-op. cit. p. 112. of the clans in other things, too. Frequently enough the totality in the classificatory system is expressed by means of a dual organizition, such as we can see in the tribal sub-division into two phratries with contrasting characteristics. The relation between the two phratries is generally represented as that between heaven and earth, light and dark, man and woman and the like. We are immediately reminded of a relation of a similar nature, when reading in S.B. I. i, i, 21, that it is forbidden to walk in between water and fire, water being a woman and fire a man, cohabitating with each other *). When we consider that „water" may represent the earth or the underworld and „fire" the sky, we see that this prohibition is a link in the universal relation existent between the two component elements of the universe. Thus we come to understand the meaning of this taboo, not as the expression of an unbridled fancy but as a precaution resultant upon the fear of interrupting the ordered course of things in the world of nature. In another place (S.B. III. 3, 2, 2.) we are told, that „the universe consists of two parts; and there is no third". „The gods are truth and men falsehood"; or „The gods are truth and the Asuras falsehood". Thus tribal organization in its most striking feature, that of sub-division into two phratries with contrasting characteristics proves to be regulative here of cosmic ordination. The system of classification we are here discussing is also to be found in the Epic, employed in full measure. Practically speaking, the whole of the so-called philisophy of the Epic is deeply rooted in the rank soil of such an over-philosophized system of classification. This very idea is expressed by the term samkhya (numeration).2) Now without further venturing to date or to explain this philosophical speculation we shall proceed to quote a few passages from the Bhagavadgïta, in which the motives of the old system of classification are sufficiently illuninating in themselves. It is said of the Triguna: sattva, rajas and tamas: Sattva is connected with happiness; rajas with the act; and tamas, which clouds the mind, obscuring know- !) Here, then, Agni is classified with the masculine side, the sky. In other instances again, he is given an abode in the water. Within the confines of the system the positions of the things classified are subject to change. 2) Oldenberg-Lehre d. Upan. p. 351. n. 129. ledge, is connected with carelessnes. And there is sattva, exceeding rajas and tamas. And, further, tamas, exceeding rajas and sattva. Likewise rajas exceeding tamas and sattva.1) Sattvam sukhe sanjayati rajah karmani Bharata Jnanam avritya tu tamah pramade sanjayatyuta Rajas tamaschabhibhüya sattvam bhavati Bharata Rajah sattvam tamaschaiva tamah sattvam rajas tatha So there exists a perfectly circulating relation between these three gunas.Thethreegunastogether constituteprakriti (nature). Now prakriti is said to be to purusha (spirit) as man to woman, or as perishable (kshara) to imperishable (akshara); so here, too, as one phratry to the other.2) Aksharaksharayor esha dvayoh sambandha ishyate Strïpumsor vapi Bhagavan sambandhas tadvad uchyate But it is also said of the same purusha, that he is understood as being the intransitory in the transitory. And there are, indeed, two purushas: a perishable (kshara) and an imperishable (akshara). And above the purusha there is also a paramatman. We have constantly to do with tribal division and connection, with the relation of one phratry to the other in the same tribe: the one is above the other, but exists only through the other, with which it is in connubium, proceeding in its turn from the other. And yet each separate phratry with its male and female aspect is a complete whole in itself, considering itself, at least, as constituting the totality of the two 3). It appears to us, that in the epical Samkhya system with its vyakta over against avyakta, its mahat and ahamkara, its five coarse over against its five fine elements, its five organs of cognition over against its five organs of action together with the the five and twentieth element, we must imagine the influence of an old system of classification.4) The number of the whole is obtained by continually adding one to the whole already obtained, just as the number five is obtained by adding as a fifth to the four castes or clans the tribe or community as a whole. ») Bh. G. XIV. 9—10. 2) MBh. XII. 305. 1. 3) L'Inde Antique p. 238. *) We shall enter more fully later on into the subject of the curious rivalry that exists between one phratry and the other. In this manner the philosophical system may be made to grow ad infinitum, and a four and twentieth element be taken as a key-stone just as well as a five and twentieth. We may be permitted to discuss a few points in the classification system somewhat more fully, not with a view to finding a definitive solution of the difficulties they involve but, more especially, for the purpose of drawing the attention of scholars to the possibilities which now present themselves. In the passage we have just quoted we read that prakriti is to purusha what woman is to man, and that the purusha infuses the various rüpas into the female principle by uniting the female and the male gunas. The attributes ascribed to the man are: bones, sinews and marrow; and to the woman: skin, flesh and blood 1). 1. Aksharaksharayor esha dvayoh sambandha ishyate Strïpumsor vapi Bhagavan sambandhas tadvad uchyate 2. Rite tu purusham neha strï garbham dharayatyuta Rite striyam na purusho rüpam nivartayet tatha 3. Anyonyasyabhisambandhad anyonyagunasamsrayat Rüpam nivartayatyetad evam sarvasu yonishu 4. Ratyartham abhisambandhad anyonyagunasamsrayat Ritau nivartate rüpam tad vakshyami nidarsanam 5. Ye gunah purushasyeha yecha matrigunas tatha Asthisnayuscha majjacha janxmah pitrito dvija 6. Tvanmamsasonitamcheti matrijanyapi susrumah Evam etad dvijasreshtha vede sastrecha pasyate 7. Pramanam yat svavedoktam sastroktam yachcha pasyate Vedasastrapramanamcha pramanam tat sanatanam 8. Anyonyagunasamrodhad anyonyagunasamsrayat Evam evabhisambandhau nityam prakritipurushau. Briefly set forth, these component elements are semen and blood, so male and female respectively.2) From the union of these two elements results conception- If the male element preponderates, a son is born, if the female element is the *) MBh. XII. 305. That in the same adhyaya skin, flesch, blood, fat, bile, marrow, bones and nerves are classified with Prakriti is not so peculiar as Hopkins thinks (Great Epic, p. 178). Here again Prakriti stands for the totality of things or perhaps the poet is now thinking of the material as opposed to Prakriti. a) Cf. MBh. XII. 320, 116. XIII. iii. 27; XIV. 18. 4; 24. 6. stronger of the two, a daughter. Concerning the human physiology the Epic teaches, that man consists of a mesh-work of veins with ten main arteries, which convey the rajoguna „like rivers in a sea", through the body, warming it. In the centre of the body, in the heart, is the trunk of the arterial system, called (in the man) manovaha, which conveys and distributes the seminal fluid to all parts of the body. The veins fed by the great artery terminate in the eyes, and from them the eyes derive their lustre. In this complex arterial system the seminal fluid is „churned"1). So we see that manas and sukra belong to one class, since the veins that distribute the seminal fluid (sukra)through the body are called manovaha (manas-duct). In another place manas and semen are even completely identified. It is related that the rishi, Vyasa, while engaged in the drilling of fire, was so transported at the sight of the Apsaras Ghritachï, that he had an effusio, the sperm dropping on to his fire-sticks. Then we read: pravisritam manah. Taking no notice of what had happened the ascetic went on rubbing his fire-sticks. Thus did Suka come into being, proceeding from the seminal fluid (sukra), rubbed by the aranis or fire-sticks.2) We presume that we may conclude from this, that manas (soul) was a male principle. If we may be permitted to delve somewhat deeper into this doctrine of man's spiritual elements we would propound the question, whether it would not be plausible to see in the differentiation between manas and asu a differentiation between male and female, in the same sense of man's being physiologically composed of a female element (blood) and a male element (semen). These two elements together ,says Oldenberg, considered in their relation to tanü (body), must be seen not as corporeal over against spiritual, but as living over against lifeless, to which we would add, that such a contradistinction in the classificatory system is not established as an incontrovertible truth for all time, running parallel, as it does, to other distinctions, such as that, for instance, between Purusha and Prakriti.3) Thinking, feeling and wishing are not as a rule associated with ») MBh. XII. 314. 12. 2) MBh. XII. 324. 9. 3) Oldenberg-Weltansch. d. Brahmavat. p. 63. asu in the religion of the Vedas. Manas, on the contrary, is the mind as the seat of consciousness, thought, wishing and volition. It is frequently said of the manas, that it is situated in the heart, where, as we have already seen, the great artery or trunk of the arterial system, Manovaha, is imagined to be. Whether the body shall live or die, is determined by the tarrying in it or the departure from it, each alone or both together, of asu (Lebenshauch) or of manas (Geist)x). In respect of tanü (body) they are naturally both together considered to be male over against female, Purusha over against Prakriti. We should expect, that, after death, the asu would go to the mataras and the manas to the pitaras; but such is not the case, for we read that the asu also sojourns with the pitaras. We could only hazard a guess as to the reason of this indeterminateness. At any rate, it does not put us entirely in the wrong, the relation of the Pitaras to the Devas being very frequently that of one phratry to the other. This is to be seen in the calendar, and even more clearly, if such were possible, in the hostility prevailing between gods and manes, although the pitaras are gods themselves. In the Rigveda the Pitaras are called the enemies of the gods (devasatravah) just as the Asuras 2). For the rest, in the ancient Indian culture, as we have seen, the matrilineal organization has always remained more or less in the background. Now if we may be permitted, instead of adducing evidence of a distinction between asu and manas in the sense of the former being considered the female component of the latter, asu being the female soul over against manas, the male soul, to assume an other hypothesis, we would venture to ask, whether, perhaps, the word Asura might serve the purpose. The Asuras stand in the same relation to the Devas as the two phratries of a tribe to one another. Now Asura and Asu are often associated with each other, an etymological connection which does not yet seem to have been established with certainty 3). But if we may assume that Asu is associated with Asura we should then have 1) Oldenberg-Religion des Veda p. 525 seq. 2) Hopkins-Ep. Myih. p. 176. s) Oldenberg R. d. V. p. 160 n. 2. an indication of the kind of grouping suggested by us1). * * * We have shown in the foregoing how the cosmic world of myth and ritual is constructed upon the basis of the social organization of India. We went for our subject-matter not only to the Epic but also to the Brahmanas, on the one hand because we had Professor Oldenberg to fall back upon, on the other hand, because, in our opinion, the same classificatory system underlies the epic conception of the world as the world represented in the Brahmanas, although in the latter the social basis of the classificatory system has become more theoreticalized. Moreover the „epic period probabi y covers a long number of years, so there is no need to keep religiously to the epic text. In what has so far been said concerning the classificatoi y system, a sort of static analysis has been given of the early conception of the world. What we now propose to do is to give what one might call a dynamic analysis of this early conception of things. What were the ideas of cosmic evolution expressed in this classificatory system? How did the world comeintoexistence and how would it pass away? That is similar to asking: In what manner did the clans associate with one another ? What were the laws regulating marriage ? We shall take the myth, which treats at greater length of Vishnu's sleep2). All beings consist of the five elements which drew together for Vishnu's sleep. At the time of the Praia ya the earth contracts into water, water into light, light into air, air into space and space into manas, manas into the manifest (vyakta), vyakta into avyakta, avyakta into Purusha, Purusha into the Universal. Then there is darkness (tamas, one of the triguna, sattva-rajas-tamas), which consists of Brahman in the form of Purusha. Purusha is Hari (Vishnu), here to be understood as avyakta and the triguna. From Vishnu's navel proceeds Brahma, representing sattva. Further there come into being two demons, representing rajas and tamas, who hinder -.The division of the spiritual elements between a patrilineal and a matrilline group is to be found, fot example, among the Ashantees of S- M» ») MBh. XII. 347. Vide also III. 203; VI. 67. Brahma in his work of creating the Veda. Now Brahma awakens Vishnu, who assumes the shape of Vishnu, the Horse-headed, and from whose body springs the world. In a loud voice he chants the Vedic mantras, which were sungin theritual, whereupon the two demons come to Vishnu and are slain by him. Thus was the Pravrittidharma established. It is evident, that the consummation of things denoted by Pralaya must not be understood here as implying final destruction succeeded by re-creation out of nothingness, but as an evolutionary and involutionary process for which the words srishti and pralaya are employed. Oldenberg indentifies the word „srij-" with the verb „ to bear, to give birth to". Prajapati, feeling the desire to create (srij-), and there being nothing with which he could unite, he assumed an androgynous form and proceeded to cohabitate with himself *). It seems to us that „to bear" is identified with „srij-" exactly the wrong way round. „Srij-" means „to cause to pass out of oneself, to let loose, pour out, send forth", and „pra-lï" is „to slip into, to disappear". The underlying thought is that of a movement such as the drawing in and putting out of the limbs, retraction and protrusion, e.g. as is done by the tortoise2). In the myth we have just related the re-generation of the world through Vishnu, the srishti following upon the pralaya, is termed the establishment of the Pravritti-dharma. The translation of srishti by creation and of pralaya, samhara, by destruction are really misleading, the emphasis in these translations being laid upon creation out of nothingness. According to the myth there was once a time when the alternate activity of pra n Oldenberg-Weltansch. d. Brahmanat. p. 272. . 2) MBh. XII. 247. 4, 14- The turtle probably occupies an important nlace in the mythological conception of the cosmos. The shape of lts föy is remeSscent 0gf the cosmic egg from which Brahma is supposed to have originated. lts two shells or shields represent the upper and the lower world, the underworld or the earth, 1^hich we conceived of as two bowls turned concavely towards each, other Vit. 5- see below p 142). In the Amritamanthana the turtle, representing the totalify? of things upholds, Atlas-like, the rotating mountain a h avatar of Vishnu. The creator, Prajapati, is also pleased to assume the shape of a turtle. Further the ancient race of Kasyapa (turtle) was proud to be able to tracé back its ancestry to the same reptilian progemtor, which also has something to do with the Indian cosmogony. It is eer Sy not wfthout reason that Oldenberg seed vestiges of totermsm in this idea (R.d.V. p. 82). vritti and nivritti (the counterpart of pravritti as pralaya or samhara is the counterpart of srishti) did not exist, there being no death. In the beginning the process of evolution began by Brahma's creating living creatures (prasrij). And there being no such thing as death, the earth became over-full and made complaint to the creator. Brahma thought and thought for a long time to find a way to return to a state of involution (samhara) again, to draw his whole being back into himself. Then was his wrath kindled and his wrath became a consuming fire. And beholding this, Siva-Sthanu turned to the supreme Brahma, and besought him to make an end of the work of destruction, to set a limit to the process of samhara, to cause the fire to draw back into himself and to allow living creatures continually to return again to the world he had made. And in answer to Siva's prayer Brahma made pravritti and nivritti. Death was born of the fire of his wrath, a woman, black-eyed and clad in red, and her mission was to walk the earth and to sow death among all the creatures thereof, among the wise and the unwise alike. In the adhyayas that follow is the striking description of the woman, death, protesting against the commands of the creator. She beseeches him with tears to relieve her of the dire necessity of executing so terrible a task; and all sickness and disease that prey upon human beings, and all living things singling them out for destruction, were born of her tears. Then follows an exhortation never to forget, when that the heart is laden with sorrow for the death of one dearly beloved, that ever upon death shall follow the new birth. Ever upon pralaya shall follow srishti. In this connection we would once again draw attention to the problem of reincarnation. For what is reincarnation other than a like cycle of successive deaths and births, death, followed by ever new life? How it really came about that this particular doctrine, which, in a broader sense, as a belief in transmigration constitutes the very key-stone of Hinduism and also of Buddhism, which presupposes this belief, sprang into existence and won such universal recognition, remains as yet an unsolved problem. The idea contained in the oldest Vedas regarding the afterlife is not quite clear. According to the Rig-veda the dead 9 belong to the Pitaras, though we possess no evidence by which we are justified in according the Pitaras immortality. But however that may be, according to the Brahmanas, even the very gods themselves were only gifted with immortality because they performed rites and had defeated the Asuras. According to the Sat. Br. XII. 9. 3, 12. the person, who is unable to participate in the boon of immortality, because he has no son who can perform the funeral rites enjoined upon the male children after the passing away of their parents, is doomed to go on dying, to suffer „repeated death" (punarmrityu). Many scholars are of opinion that this belief in transmigration attachés to this idea of punarmrityu. Professor de laVallée- Poussin, however, remark, as against this, that punarmrityu is not to be understood as being a repeated dying in this world but in the next.1) According to the Brahmanas the father is, indeed, born again in his son, the son being the atman, the „self" of the father. So one would be inclined to say, that those who have not been blessed with children have escaped death for good, there being no possibility of their being reborn again. In the religious belief of the Upanishads, however, the doctrine of metempsychosis plays an important part. Man passes from one life to another, the nature of each separate life being determined by his moral conduct in the preceding form of existence. Various forms of existence are graded in such a manner, that each succeeding form is always a hundred times better than the one immediately preceding it in the sequence of existences denoted by the appellations, Pitri, Gandharva, Deva, Prajapati and Brahma. The religious philosophy of the Upanishads teaches that the dead can follow the way of the gods and so, attain at length to the heaven of Brahma. Others who have vanquished the witchery of the world, by means of offerings to the gods, the giving of alms to the needy among men or the practice of severe self-discipline and selfdenial, such journey by the road of the Pitaras to the moon, whence by way of the moon, ether, air, rain, food, semen and embryo, they return to earth again. Others again, who, knowing neither way, walk neither, become worms, locusts, or gnats. In the opinion of some scholars the belief in the transmigration of souls constitutes part of the teaching of Yajnavalkya, *•) de la Vallée-Poussin-Indo-européens etc. p. 277 seq. who himself explains it as resulting from the influence of karman. That the foundation of such a fundamental conception of things, such a determinative vision of life and the after-life, is to be sought in the mere preaching of a single teacher, must be deemed almost incredible; hence de la Vallée-Poussin's judgement that the foundation must indeed be considered all too slight to support so grand a structure. If we well vinderstand the Belgian scholar's purpose, he means to draw a distinction between the universal Brahminical belief in the doctrine of reincarnation on the one hand and the belief in the transmigration of souls on the other which could easily become associated with the vague ideas, held by the Aryan race, regarding life and death. Reincarnation he calls the belief in a rebirth of the father in his son, while transmigration, a much wider conception, is the belief in a metempsychosis whereby one comes after death to live again in another existence, be it one of man or animalor other according to the value of one's actions in a previous life. The doctrine of the transmigration of souls, however, as conceived of by the Buddhists, has, in the same scholar's opinion, nothing to do with the belief in the Pitaras, the reincarnation. The belief in the transmigration is founded upon a theory of conception, which, as de la Vallée Poussin rightly remarks, strongly resembles that of a number of primitive peoples, such as the Aruntas of Australia, viz. that conception is brought about by a Gandharva entering into the womb of the mother: „le Gandharva est un être désincarné qui cherche une matrice". This significance attached to the word Gandharva is, indeed, considered by most scholars to be pretty firmly established. Professor Hillebrandt, however, disputes the meaning of „Wesenskeim" (germinal being) attached to Gandharva, pointing out, in the first place, that, in the earliest products of Indian literature, the Gandharvas are not conceived of as insignificant beings yet to be born, but as beings of a higher order in between gods and men1). In the mural sculptures of the temples they are are simply represented as ordinary human beings, and in Maya's supernatural conception of Buddha the embryo penetrates into her womb in the form of a white elephant. In the Epic and the Puranas the Gandharvas are represented as companies of ') 84ster Jahresber. Schles. Ges. f. Vaterl. Cultur. dancing and singing beings of a higher order, heavenly minstrels, plying their subtle art on the sacred mountains and elsewhere and in Vedic times they constituted the guard of Soma. Hillebrandt concedes, that Buddhiscic literature is not devoid of some places, in which the Gandharva really is spoken of as a prefoetal being, but attributes it to an erroneous conception of the meaning of the word Gandharva, which, in his opinion, means giant or giant spirit of the air. It is also true, that commentators of the kosas explain the expression antarabhavasattva (the being in an intermediate state) by declaring that the Gandharvas stand between death and birth, (maranajanmanor antarale sthitah). But he calls this comment an „afterthought". Oldenberg, who likewise considers this meaning of Gandharva (viz.foetus)a„sekundare und nebensachliche Seitenentwicklung", thinks Hillebrandt unconvincing, though, on the other hand, he cannot deny, that Hillebrandt certainly showed that there is no reason to see in the Gandharva a feeble, prefoetal being and that it never was seen as such.1) De la Vallée Poussin sees more in the word. We shall now give our attention to the idea of conception, mentioned by him, obtaining among the Aruntas of Australia, thereby following the exposition of the idea given by the French scholar, Dürkheim in his highly important study of „Les Formes Ëlémentaires de la Vie Religieuse2). The researches of B. Spencer and F. J. Gillen have shown that the Aruntas of Australia do not imagine the souls of the new-born to be the product of separate, ever new creations. They think there is a limited number of souls composed of the souls of their ancestors of early mythical times, and tarrying, in expectation of re-birth, in the places where these ancestors died. They know where such places are and are able to point them out exactly. They are the holy places where the clans keep their bullroarers 3). It is when one of these ancestral souls penetrates into the body of a woman that conception takes place. So every indiv- !) Oldenberg-R. d. V. p. 253 a. i. ) P- 352 seq. 3) A flat, oval piece of wood which, swung round and round on a piece of string, produces a peculiar buzzing sound, to which a supernatural significance is attached. These bull-roarers are of great ïmportance in the non-profane or „sacred" life of the tribe. idual is the avatar of a certain ancestor. After death the soul returns again to the ancestors. These mythical ancestors were so closely associated with their totem, that they became inextricably bound up with it, as it were, being in the course of time entirely identified with it. An ancestor of the Kangaroo totem for instance is imagined to be half kangaroo, half man. Alongside the information furnished upon this important point by Spencer and Gillen, there are, however, the conclusions arrived at by Strehlow, which are different. Perhaps de la Vallée-Poussin was referring to Strehlow's publications. According to this ethnographer the human soul goes after death to the land of the dead, where after various vicissitudes it comes to dweil for good; hence Strehlow denies the possibility of a reincarnation. Strehlow is of opinion that conception is thought to take place as follows: Upon the spot where a mythical ancestor terminated his journeyings upon earth, disappearing into it, there is a rock or a tree, in the neighbourhood of which the ratapa, the „Kinderkeime", „spirit-children", or however they may be called, are thought to be waiting an opportunity to be embodied in human form. Now when a woman of the suitable marriageable class passes by, the ratapa penetrates into the womb of the woman through her thigh, after which she is acquainted with her state of pregnancy by the usual prognostic symptoms. Another mode of conception is this: the ancestor, to whom the ratapa in question belongs, himself takes an active part in the work of fructification by throwing the woman a small bullroarer of a special shape that penetrates into her body. Both modes of conception are equally frequent. A third and less usual manner is, that the ancestor himself assumes corporeal form in the woman's womb. The contradiction that exists between the accounts of Spencer and Gillen and those of Strehlow is, however, in Dürkheim's opinion, more apparent than real, it being evident from Strehlow's description of the ratapa, that it is really nothing else but the mythical ancestor in miniature. And on closer inspection the bull-roarer, too proves to be simply a mythical equivalent of the ancestor himself; in India one would say: it is a rüpa of that ancestor. It is not any ancestral embryo that a man may have within him; far from it; it is a special embryo belonging to him and to him alone. It is obvious, then, that, according to both Strehlow and Spencer and Gillen, there is a supernatural something in every individual emanating from the mythical ancestor. The difference between the two versions is simply, that this something by means of which conception is brought about, returns ultimately in the shape of a soul to the group of ancestors according to Spencer and Gillen and that, according to Strehlow, it disappears altogether after death. In the first instance, therefore, the number of souls is continually being added to by the souls of the dead, in the second instance the number of souls is by nature inexhaustible. In any case, then, it is quite evident that in Australia the ratapa or spirit-child is related to the ancestors; an Indian would say: the Gandharva is related to the Pitaras. The assertion that the father is reborn in the son, as the doctrine of incarnation is concisely formulated in the Brahmanas, must probably not be taken word for word in a Western sense but in the sense of a classificatory relationship system or kinship grouping. The meaning would then be: the Pitaras are reborn in the younger generation, it being usually a group of ancestors in general that is alluded to by the word, Pitaras. It is plausible, that the deceasedwas considered after the fifth generation to belong definitively to the group of mythical ancestors, the deceased from that generation onwards no longer being commemorated in the ritual of the dead 1). Then it is very well possible that the group of the Pitaras is really identic with that of the Gandharvas; hence Gandharva might also mean: a mythical ancestor who desires to be embodied in human form again 2). It is a remarkable thing in this connection, that in the Vedic ritual accompanying the offerings to the Pitaras the man gives one of the cakes of meal (pinda) prepared for them to his wife, when she wants a son 3). Even in our own times a dead person is called a „pret" immediately after death (and before the sapindïkarana?), likewise the spirit of a deformed person, also a child that has died early as a result of the ritual for the oee p yu. 2) The meaning, i.a. attached to Gandharva in the P. W. is: die Seele nach dem Tode, bevor sie einen neuen Körper erwahlt hat. After death one belongs to the Pitaras. 3) Oldenberg-R. d. V. p. 568. formation of the embryo (garbhadhana) having been neglected, or the ritual for obtaining a male child (pumsavana). It is still believed that the soul takes up its abode within the foetus or fully developed embryo in the fifth month of pregnancy1). Uninitiated children are not cremated, but buried, in the earth, the spirit, in this case, not ascending to the ether, and there being a likelihood of its being reincarnated in another female member of the family 2). In India then as in Australia among the Aruntas it is apparent that there was a similar possibility of a person either coming to belong after death to the Pitaras or ancestors and then being embodied again in human form or of his going to heaven and staying there for good 3). The Gandharva, then, may be the same as the Pitri among the Indians just as the ratapa is the ancestor among the Australian Aruntas, and in the second place there is the conception of the Gandharva being not itself the Pitri but anemanation from the group of the Pitaras just as the ratapa in Australia is an emanation from the mythical ancestor. If the analogy with the Australian ratapa belief as demonstrated by de la Vallée-Poussin holds good, then it is conceivable that the Gandharvas are not human beings in statu nascendi, but beings intermediate between gods and men, of a higher order even than the Pitaras. In the classificatory system the Gandharva is classified with life and the Pitri with death. The translation, Gandharva = foetus, plausible enough at first sight, is on second thoughts nevertheless misleading, the significance attached to this word being more physiological than mythological, the physiological colouring being too strong, the mythological too weak. What is further known concerning the Gandharvas coincides, in our opinion, with our conception. In myth the Gandharvas are pictured as being especially enamoured of women. There existed various exorcizing incantations with which to expel the Gandharva out of the woman. In the first nights of marriage, which had to be spent in sexual continence by the newly-married couple, a staff was placed in between them, resembling the annointed and draped sacrificial pole, wrapped *) The first cry of a new-born child is the complaint of the soul, conscious of its being caught in the meshes of Maya, the toils of „illusion". 2) Crooke-Rel. and Folkl. of North.-India p. 186. 3) Cf. BhG. VIII. 24 seq. round with a piece of cloth or wound round about with thread, the symbol of the Gandharva Visvavasu 1). Von Schröder describes the Gandharvas as hairy little manukins with large testicles, eaters of raw flesh, that try to get to the female genital organs. They have to be hunted away from her. They 1 ie with women in the shape of brothers and fathers. Gandharvas and Apsaras govern female fertility, and are worshipped by those who desire children 2). The shape the Gandharvas are thought of as assuming is also noteworthy. Scholars are apparently uncertain as to whether they are to be given a human or an animal shape, hesitating between the two. They are sometimes represented as half horse and half bird 3). Prof. Vogel of Leyden is concerned with the question as to whether the term, antarabhavasattva employed by Amarakosa, explained by commentators, as we have seen, as an intermediate stage between life and death, might not possibly mean a form of existence intermediate between that of man and beast. One of the well-known figures with wings, claws or a tail, familiar to us from the plastic art of India has been designated byDr.Burgess as a „gandharva or kinnari"4). If the Gandharva can be assumedasbeinganalogous tothe ratapa or spirit-child, representing the mythic ancestor in miniature, the blending of man and beast in the shape of the Gandharva would be still more strongly suggestive of a totemistic classificatory system. The mythical ancestor of the Australian Aruntas was likewise imagined as a being intermediate between the totemanimal and human-being 5). Likewise noteworthy is the association of the Gandnarva with the fire-stick in the myth of Purüravas and Urvasï and with various trees, in which he is supposed to have his abode. But there is still very much that is far from clear, even though the solution proposed by us be abided by. Are the Gandharvas so given to dancing and singing, because the ancestors are thought of as being constantly occupied with the rites accompanying religious festivals? Were the *) Oldenberg-R. d. V. p. 88 n. 2; 252. !) Myst. und Mimus p. 61. Macdonell-Ved. Myth. p. 137. 3) Some scholars imagine a relationship between Gandharva and Centaur. *) Hillebrandt-loc. cit. p. 8 n. 2. 5) Vide supra p 133. Brahmins ignorant as to what caste the Gandharvas belonged to as a result of the ancient clan-organization having been thrown out of gear by the rïse of the caste-system, and the Gandharvas therefore no longer being able to enter into the right woman? 1). Whatever one may think of this solution as such, we ourselves,in any case,are convinced that further investigatorystudy as to the Gandharvas will have to be conducted in this direction. An „afterthought" — which will have to be explained for the matter of that — a mere „afterthought" is certainly not what we are inclined to see in the Gandharva understood as foetus. The explanations which have so far been given of these mysterious beings, giants of the air and spirits of the mountains or brownies or gnomes or imps or hobgoblins, the rising sun, cloud-spirits, rainbow and the like, are all quite in the air, literally and figuratively. The connection made by de la ValléePoussin between the idea of Gandharva as „germ of life" and metempsychosis or the transmigration of souls appeals to us as being most satisfactory. A precise description of their outward appearance is lacking; and, indeed no definite idea about this seems to have existed. They are intermediate between different forms of existence, just as the totem-ancestor, the key-stone of the classificatory system2). In our opinion the belief in reincarnation and that in the transmigration of souls are not to be thought of as unconnected with one another; on the contrary they have so much to do with one another, that the belief in transmigration is merely a more comprehensive form of the belief in reincarnation with an ethical basis given toitby the doctrine of karman. Community of soul between father and son if understood in the classificatory system does not include father and son only, since the parental and filial relationship also determines the cosmic continuity of the clan, and, hence, of all those things ') Perhaps the Gandharva is even „la négation de la caste", as de la Vallée-Poussin suggests (op. cit. p. 288). That this is the root-idea underlying the mythological fancy of the Gandharvas seems to us extremely improbable. 2) Their female counterparts are the Apsaras, who are associated with water as the Gandharvas are with the air. That does not mean to say, however, that it is our intention to represent the Apsaras as the fe male „spirit-children". associated with the clan. In this sense the doctrine of reincarnation is not only a way of expressing the relationship between father and son, but, at the same time, in a much broader sense, the relationship existing between a person and all those things with which his clan is associated. The doctrine of the transmigration of souls is simply the same relationship expressed in a more elaborate form, a form of belief which will arise all the more easily where the system of classification finds itself compelled to adopt other forms, e.g. as a result of the very nature of the clan-organization in itself becoming different under the influence of the development of cast es and the like. The fact of belief in the transmigration of souls not being met with in the earliests products of Indian literature need not, therefore, be deemed very surprising under the circumstances. The remarkable thing about the doctrine of metempsychosis in the gospel of Buddhism is probably its dependence upon karman. In an earlier phase a person's place in the cosmos, i.e. in the clan, therefore, was dependent upon his father's clanmembership, and, likewise, his place, on that ground, in the ritual of the tribe; but the Buddhist moralists looked at it in a different way altogether; a person's place in the cosmic order was, in the new teaching of the Buddha, to be determined by his own individual merits or demerits; the nature of a man's relationship to the universal cosmic order was to be determined by what he did or did not do. This simply meant that the Buddhistic conception of the transmigration of souls was in harmony with the supercilious attitude adopted by the Buddhist towards the whole idea of caste superiority and exclusiveness. It was not the caste to which a person belonged that was to determine his status, but, in the first instance, his own individual significance. We have allowed ourselves this digression in order, from this angle, to focus the attention once again upon the problem of the Gandharvas, and we shall now return, having found it among the Gandharvas, to our discussion of the circular course itself. In the myth which tells of the origin of death it was said that pravritti and nivritti came into existence at the same time. The meaning attached to these words in Indian philosophy is approximately that denoted by the English words, energy and inertia or activity and non-activity. The stem-word is vrit, to turn. So, literally, they mean „turning forwards" and „turning backwards". In 1898 Prof. Caland wrote an interesting article entitled: „An Indo-Germanic Lustration Practice" *), in which he drew attention to the essential differences to be found in these practices according as the ceremony concerned the gods or the manes. The difference is shown in the manner of orientation, in the use of the left or the right hand and foot, in the way in which the sacred cord is worn, in the turning of the body and in the side of the body turned towards the object to be worshipped. As regards the manner of orientation East is front, South right and West behind; but North is not considered left, being a region of the gods, which may not, therefore, be called left, because right is the „good" side, the „lucky" or „favourable" side, left being the „inauspicious", the „sinister" side. The East is pre-eminently the region of the gods, hence the religious usage of marking off the place of sacrifice by means of lines drawn on three sides, the front, facing the East, being left open. If a person turns round in the performance of a religious ceremony, he must also always make a turn back in the opposite direction. If the officiant, for example, has turned round about his right arm as round an axis, he must turn back again round the left arm taken as axis. In the first case he is dakshinavrit, in the second case savyavrit. Vivritya is defined as apradakshinam avritya. Incisions must be made in the sacrificial pole, running from right to left starting from the top. This unusual direction (the „good" direction is always pradakshina, from left to right) cannot, in Caland's opinion, be a mere accident. But the real meaning underlying this peculiar departure from the norm can only be guessed at. When one wishes to make a sacred fire, the fire-stick must first be made toturn three times to the right. In the ritual processions there very frequently follows a retrogressive movement in the opposite direction. In funeral ceremonies it is the custom to move round from right to left in the procession. In the Epic there are some examples of warriors describing circles in their war-chariots round about their enemies, thereby *) Versl. en Meded. Kon. Ak. Wetensch. Amsterdam afd. Letterk. IV reeks dl. II, p. 275. Vide also J. Przyluski-Pradaksina et Prasavya en Indo-Chine-Festschr. Wïnternitz 1933 p. 326. driving around them from right to left. In this manner they were enabled to defeat their foes. The mere fact of Caland's mentioning these data induces us to venture the opinion that he is on the track of an important mythological idea which is very much more than a mere usage in a purificatory ritual, a mere lustral practice, as Caland somewhat superficially terms it. The mythological idea to which we refer is that of the rotatory motion of the cosmos as plastically represented in the ritual for the making of fire by means of the fire drill and in the myth containing the plastic description of the „churning" of the ocean (Amritamanthana). The making of fire has been closely associated in the Indian mind from ancient times with the physical act of generation, and is, as a result, accompanied in the prescribed ritual by a number of traditional sayings bearing upon sexual life. The two firesticks (arani) represent man and woman, being termed, after what they are supposed to be in the ritual, yoni and prajanana, and are identified with the familiar pair, Purüravas and Urvasi. The very bringing of fire to men is connected with the advent of the first man or king. Right down to modern times the making of fire is imagined to be a symbolical coitus; hence the practice, upon the night preceding the celebration of the fire-ritual, of giving the upper fire-stick or arani, the fire-drill, to the Agnihotra („fire-priest") and the lower arani, the hollow fire-stick, to his wife for both of them to sleep with. It is the custom in various districts to perform the rite of fire-making on all sorts of occasions, important events etc. such as births, funerals, housebuilding and the like. The fire-sticks were manipulated with the hands and also by means of the fire-drill, which is still in use. The apparatus consists of the lower arani (adhararani) and the pramantha („churner", „rubber"); the latter constituting the upper arani (uttararani), the lower part of which turns in a hollow of the adhararani, and the upper part is fastened to a spindle (chatra), which by means of a piece of cord (netra) wound around it can be made to turn. The man holds the whole apparatus, while the woman pulls the cord *). *) Hillebrandt-Rituallit. p. 106; Crooke-Rel. and Folkl. of N. India p. 335. These two descriptions do not wholly agree with one a nother, that of Crooke regarding the modern time being more ample and clear. This turning or grinding motion forms part of a number of ceremonies, the mill-stones, for example, being held just as sacred in the nuptial rites of present-day India as in the distant days of the Vedas. In the rite performed for the purpose of procuring the birth of a man-child (Pumsavana), a twig a the Nyagrodha tree *) was crushed some months before birth between mill-stones and put into the nostrils of the pregnant woman 2). This idea of rotator y motion also underlies the Indian's conception of the Universe and the cosmo-plastic laws that governed its moulding into concrete form. The earth is a disc 3), washed upon all sides by the vast ocean of the world; arched over the earth is the bright dome of heaven (naka) with a corresponding dark dome arched beneath it. The domain inhabited by each of two oposing groups of gods is bounded by either dome. In the centre of the wide earth is the mountain called Meru, whose base is often said to be narrower than its summit, representing the ovary of the worldlotus. Near the great mountain is the mighty tree called Jambu 4). Meru is the great spindle, upon which the celestial dome revolves. It is called the sublime mountain, and its peakis the immovable point which sustains the dome of heaven (nakam avritya tishthantam uchchhrayena Mahagirim)5). W. Kirfel also mentions the idea of the axis of the world conceived of as passing through the earth into the waters below, where the serpent, Sesha, which like most serpents, is supposed to inhabit the waters of the underworld, supports the earth (adhastad dharanïm sada dharayate)6). Alongside the tripartite division of the cosmos into heaven, earth and underworld (or the waters under the earth) there is an earlier dual division to be found already in ') The Nyagrodha tree (Ficus Indica) is often represented in postVedic literature as the male fig-tree over against the female fig-tree, Asvattha (Ficus Religiosa). Zimmer — A. i. Leben p. 58. The Gandharvas and the Apsaras are supposed to have their abode in these trees. Oldenberg — R. d. V. p. 255. 2) Crooke-op. cit. p. 324. Hillebrandt-Rituallit. p. 65. 3) Possibly as a number of concentric circles having the Mountain of Meru as their centre. The oldest conception, however, is supposed to be that of a circle divided into four parts, of which the Southern quarter is called Jambudvipa (Oldenberg — Das Mahabharata, p. 87). ') Kirfel-Die Kosmographie der Inder p. 14 ; Oldenberg, Das Mahabharata p. 84 seq. 5) MBh. II. 17. 7. 6) Hopkins-Ep. Myth. p. 23, 26. the Vedas, in which heaven and earth are conceived as opposites. They are thought of as two vast bowls (chamva) turned concavely towards each other, and are compared in one place with the wheels at either end of an axle 1). Earth and underworld are then practically thought of as one. Heaven, as we have already said, is, to a certain extent, the reverse of earth, hence the atmosphere, too, may be called a sea (samudra) in which the clouds are mountains. Now if we imagine these two, heaven and earth, as turning on a spindle, like wheels turning upon an axle, we have two centres of rotation and here we think of the ritual act of firedrilling as the procreative motion through which fire is produced (Agni is born). In accordance with this idea, Agni is continually credited with a twofold birth, a lower and an upper birth. Macdonell takes the lower birth for the making of fire with fire-sticks. But it is exactly the earth which is the underworld and sometimes, indeed, Agni's „celestial" birth is opposed to his birth in the waters 2). The making of fire by means of fire-sticks on earth has undoubtedly been imagined as essentially the same as the making of fire by friction resulting from the rotatory motion of the cosmos. The sun is one of the forms of Agni placed by the gods in the heavens 3). Since the gods are made to act by human ritual, it can also be said that the Angirases caused him to be put there by means of their ritual, or that the sun was brought forth by Soma, the offering of Soma, the principal sacrifice.4) We have already seen that ritual and cosmic order are practically identical. When laying the sacrificial fire the Brahmin had to set a cart or a cart-wheel in motion on the Southern side 1) Macdonell-Ved. Myth. p. g, 10. 2) Macdonell-Ved. Myth. p. 93 seq. . 3) The association of Agni with the sun is, of course, obvious; although, in our opinion, it is starting from the wrong side: to proceed from Agni's identity with the sun, with lightmng, and the like, by way of a nature-myth interpretation of his significance. Nor is it originally a mythologization of the natural phenomenon, that warmth is produced by friction, that makes Agni originate fiom the fire-sticks. It was Vivasvant, the first of men to sacrifice to the gods, whose messenger, the eagle, Matarisvan, stole fire fro m heaven. Ritual, so also the ritual of fire-drilling, is not of man but of God; its source is heaven. The stealing of fire from heaven is not, in our opinion, contradicted by the „birth of Agni out of fire by friction; it is merely that ritual birth given a wider association. Oldenberg — R. d. V. p. 121. 4) Macdonell-Ved. Myth. p. 31. of the place of sacrifice until the wheel had turned round three times. Fire-drilling before sunrise was taboo. Here, then, we see clearly enough the close connection existing between ritual and cosmic order *). Agni, however, is spoken of as being born not only in two, but even in three ways. So in that case we have to do with a birth in heaven, upon earth and in the waters of the underworld. „From heaven first Agni was born, the second time from us (men), thirdly in the waters under the earth" 2). The „celestial" Agni must be made by the frictional action of the gods, which means that they, too, must actively co-operate in keeping the cosmos in motion. This cosmic motion, as we have met with it more especially in the ritual for the making of fire for the sacrifice, proceeds in the Indian mind from Brahma, the impartial father of Devas and Asuras alike, the twin parties, which, in the cosmos created by Brahma, constitute the two rival groups. Thus, too, it is represented in the myth about the churning of the drink of immortality (Amritamanthana) 3). The Devas and the Asuras forget their rivalry, and carry with the help of Ananta (the serpent upon which Vishnu sleeps) the mountain of Mandara (Siva's mountain) to the sea, where it is supported by the turtle (Vishnu). Gods and Asuras, standing on either side, now pull the churning-cord to and fro, the cord being Vasuki (the serpent associated with Siva), its head held by the Asuras and its tail by the Devas 4). A tremendous fire results, which is again extinguished by Indra. As a result of unwearying labour on the part of the churning gods and demons all sorts of desirable and delectable things emerged from the depths of the ocean, such as the moon, Lakshmï, the white horse Uchchaihsravas, sura, the jewel of Narayana, Dhanvantari with the coveted soma, drink of the gods, and, at last, Halahala, *) Oldenberg- R. d. V. p. 109 n. 2. 2) Macdonell-Ved. Myth. p. 93. 3) There it is not Brahma himself who commands the amrita to be churned, but Narayana a manifestation of Vishnu who asks Brahma to give the command, and then himself commands the gods and demons to start churning the sea. In the meantime Vishnu himself has not remained idle, but has actively assisted in lightening the labours of the churning Devas and Asuras by holding up the churn-staff himself in the shape of a turtle. Concerning this state of non-activity of Brahma in the actual work of creation, his not doing things himself, but his transferring of his cosmic functions to another god who does the work of creation proper for him at his injunction, see below, p 223. ') Vogel-Indian Serpent Lore p. 202. a burning poison, which was drunk by Siva in order to save the world from destruction. *) The Asuras, having so far gained nothing from their tireless churning of the ocean, for the gods had appropriated all this untold delight, thus conjured up by their united labours from out the „unfathomable deep", all too ungenerously to themselves for their own private enjoyment, the former became enraged, and a violent struggle between them and the Devas ensued for the possession of the moon and the drink of immortality. The covetous gods shared the delectable drink between them, but were not quick enough; the Asura, Rahu, managed to get the vessel that contained it within his grasp and take a draught of the elixir of life, but it did not pass his throat, for Narayana succeeded just in time in preventing the Asura from swallowing it. But that part of the demon's body that had held the immortalizing nectar was, henceforth,immortal, and from that moment, there was enmity between Rahu's head and sun and moon and he is ever out to devour them. The Asuras, embittered by Vishnu's deceitful behaviour (mayam mohinïm samupasritah), who had played them false in the shape of a woman, now waged fierce but futile war against the crafty gods who had cheated them by their cunning.Thegods were victorious,andfromthat moment the immortalizing drink became their inalienable possession. It is evident that the procreative motion of the universe traces its origin back to the co-operative action of the two rival parties of the Devas and the Asuras, laying aside for the nonce their ancient rivalry, in the performance of an important rite2). i\ a nrobably not very reliable representation of the Amritamanthana, most likelv boriowed from Baldaeus — Afgoderye (Idol:worship) etc. (A T. de Tong's edition) p. 52, has frequently been met with by us, ï.a. in B A Gupte's Hindu Holidays and Ceremomals, pl. 13; and in Winthuis' Das Zweigeschlechterwesen, Bild 11, borrowed by him in turn from Jackson. There is another representation of it to be seen in ^^Th^the"mytl^of the Amritamanthana is to be understood merely as an allegorical description of sea-borne commerce in its early days C. F Öldham, in his work, The Sun and the!Serpent^, ceeded in convincing us; it seerns very unprobablc. HaveU sees m th churning of the ocean a description of the nsing of the sun, wtucn, viewed from a high peak in the Himalayas, looks as though it were being dnüed wi°h^a fire-dnll from outtheclouds (= the sea) - (Handbook of Indian Art p. 169). Gupte in Hindu Holidays and Ceremomals, p. 228, mentions Br^nnand as having said that it is an eclipse of the sun that must here be thought of. In another myth the existing world-order is represented as dependent upon the uninterrupted sequence of life and death, seen as an endless cycle of life fading into death and death constantly blossoining out again into new life, life and death being called into existence simultaneously with the creation by the supreme Brahma of pravritti and nivritti. We presume that the word, pravritti („turning forwards") is suggestive of the turning of the churn-staff from left to right when the churning-cord is pulled by the gods, and the word, nivritti, as implying the „turning back" again, the turning of the churn-staff from right to left when the churning-cord is pulled by the demons. Pravritti is ihen the „good", the „auspicious" turning, just as in the ritual for the making of fire the officiating priest is enjoined to commence the ceremony by giving the fire drill three turns to the right. It is incumbent upon the officiant, having turned round about his arm as round an axis, always to turn back again. This injunction is evidently based upon the same churning movement, in which a movement to one side of the churn is always followed by a re verse movement to the other side. Whether there is any connection between pravritti and the pulling of the churning cord bv the gods on _he one hand and between nivritti and the pulling of the churning-cord by the demons on tne other must remain for the time being a supposition, since we have not succeeded in coming upon any positive evidence that might justify our assuming that a similar connection really existedx). But however that may be, the gods who participate in the performance of the ritual act followthe law of pravritti2). The movement denoted by the word, nivritti, is connected with the idea of death, but it is also the movement of Vishnu himself3), especially of Vishnu Narayana, the sleeping Vishnu, symbol of life that has ceased from activity and become merged in death 4). The essentially philosophical ideasunder- x) In MBh. XIV there is mention of kalachakrasya pravritti and kalachakrasya nivritti. Both Deussen and Strauss (Vier Phil. Texte des Mhbh. p. 73) as well as Protap Chandra Roy render „standstill and movement of the Wheel of Time." Bui can there be any question of a standstill in the process of involution and evolution? 2) MBh. XII. 340. 3. 3) MBh. XII. 337. 4) The modification of contrasts, such as life and death, becomes plausible enough, once we think of a cycle in which life and death succeed one another in such a manner that life contains within it death 10 lying the words, pravritti and nivritti, the real meanings of which cannot perhaps be said to be conveyed with exactness in the English rendering, „activity-inactivity" (energy-mertia), might then be assumed to fit in quite naturally with the conception that we ourselves have of the Amritamanthana. The stemword, math, contained in the compound-word, Amritamanthana, does not only mean the turning of the fire-drill, but also the turning of the churn. It is, therefore, evident, that not only the action of fire-drilling, but also that of churning was connected in the Indian mind with the idea of cosmic motion. Even nowadays in India milk is still churned by means of a churn-staff with a cord wound round it, in exactly the same manner, therefore, as it is described in the myth of the churning of the sea in the MBh.Mindful of the sacred character attaching to the cow in India, this identification of the churning of the sea with the churning of milk becomes easy enough to understand. The cow is identified with Prithivï, the Earth-Mother, and with Aditi, the Mother of the Gods.2) In conclusion we would draw attention to two remarkable words, viz. to the term, dharmachakrapravartana, the term employed in Buddhism to denote the preaching of dharma, and the term, chakravartin. Both words must, in our opinion, be understood as alluding to the idea of rotatoi y motion, being the movement attaching to the making of fire and the churning of milk as religious rites together with that of the entire cosmic world of gods and men. The chakravartin, the wheel-turner by pre-eminence, is the absolute monarch, holding undivided sway, who, being, as a god among men, the preserver of the universal and death bears within it the seed of life without either idea being feit to lose its own peculiar and essential character. Man after death comes back aeain as a Gandharva from the shades of the Pitaras. ') See the picture in The Todas by W. H. R. Rivers, pl. 15. is a peculiar ritual, in which dairy-cattle play a leading part. „The general plan of the dairy operations appears to be much the same as that practised elsewhere in India." (op. cit. p. 52). . , . . 3) Aditi is the vaguely defined being famihar to us as far back as tne Vedas, chiefly as the Mothei. She is the Mother of all the gods, and is identical with Vach (the ritual formula) and with the earth; in her dual form identical with heaven and earth together (Macdonel! Ved. Myth. p. 120: Hopkins — Epic Myth. p. 81). Is she the female counterpart of the Supreme Being himself, the same as the old woman in other mytholoeies? Locher — The Serpent in the Kwakiutl Religion, p. 71; Duyvendak — Het Kakeangenootschap in Seran (The Kakean Society in Seran) p. 163. order, governs the motion of the cosmos 1). In that passage of the Bhagavadgïta, III, 15—16, in which Brahman, the ritual formula, is said to be present in the sacrifice itself, it is said, that he who does not make the wheel roll by means of the sacrificial ritual (pravartitam chakram nanuvartayatïha), lives in vain. * * * After having endeavoured to give an exacter definition of what we ourselves understand by the word, myth, we proceeded to examine the mythical conception of cosmic order as elaborated in a classificatory system, consistent with a mode and trend of thought derived from a state of human society in which the circulatory connubial system prevailed. The cosmic order, preserved and governed by ritual, yea, even more or less identical with the religious ritual, upon which it is dependent, is that fundamental principle of order, permeating, as an impersonal power, the whole of creation, denoted by the word, Brahman, the ritual formula, proceeding from a supreme deity, conceived of as personal, Brahma himself; Brahma, the personal god and creator, Brahman, the impersonal power and ritual word and the entire system of prescribed ritual connected with these two, are all three one and indivisible; the supreme Brahma, then, the impartial father of the warring Devas and Asuras, the dual celebrants in the performance of the cosmic ritual, stands more or less aloof from the world of which he himself is the creator. *) H. Kern derives the word, chakravartin, „world-ruler" in the sense of the „quoit-thrower (discobulus)" or „celestial wheel-turner", i.e. Vishnu, from another usage, according to which the word, chakravartin, really means „a person who governs (waitayati) a district (chakram). Now when this meaning of wartin had become forgotten, the word was taken to mean „wheel-roller" or „a person whose wheel rolls over the earth" and the significance of a „world-ruler" became attached to it, whereas the woid, as has already been said, simply signifies „governor" and nothing else. By way of rhetorical fancy it was also made to refer to the sun, its wheel or disc (chakra) being thought of as rolling daily across the blue expanse of heaven. In extravagant fancies such as these, the result of abusing the science of etymology, Indian mythology is as rich as early Roman law". The History of Buddhism in India, vol. 1. p. 217, nt. 22. III KRISHNA AND ARJUNA We explained in the preceding chapter that the religion of ancient India has been reduced to a substratum of natural phenomena, which is still to be found in the old gods of the Vedas, gods either immediately associated with some natural phenomenon or other or, at least, supposed to have once been associated with such phenomena on the strength of etymological considerations. Pious imagination has made the old Vedic deities so human that their pure and unalloyed naturalistic origins have had to be relegated to the regions of prehistorie obscurity. We shall now proceed to discuss another alleged root-idea underlying the religion of ancient India: the idea of magie which scholars link up with the idea of Brahman, the Absolute, the dynamic power permeating all cosmic forces and phenomena. The arguments of investigators resolve themselves to this: it was thought that by means of the ritual formula this vast dynamical power could be laid hold of and converted to the use of the ritual celebrant or celebrants. The ancient Indians had come to believe that the order of the Cosmos (rita) was inextricably bound up with the efficacious power inherent in the ritual sacrifice. When it came to be considered a perilous thing to continue to entrust the sole care of the universal world-order to a single individual, the gods were burdened with the responsibility of its maintenance and compelled by means of the ritual sacrifice to fulfil the wishes of the sacrificial celebrant. The gods were made to react upon the ritual act of men. This stage in the development of religious thought, the shifting of cosmic responsibility on to the broader shoulders of the celestial occupants, had not yet been arrived at in Vedic times. Men, in that age, were content to make known their wishes and desires to the gods, in the hope of working upon their feelings and evoking their divine generosity.1) That „root-idea" of magie, which investigators think they have here discovered, all at once emerges from obscurity. Scholars admit that there is neither a pure substratum of natural phenomena nor a pure substratum of magie to be found in the literature of the Vedas. The state of things we come up against as far back as the Vedas is such, that the personalization of the gods is beginning more or less to pale beneath the strong light of a „pantheistic" Brahman belief. For in the Rig Veda the inclination is already apparent „die Umrisse jener göttlichen Individualitaten wieder zu verflüchtigen, in ihnen eine Umschleierung zu erkennen, die ein unpersönliches Eines umhült". 2) The word „magie", as employed here, is an extremely dangerous word to use. „Magical" is the term employed to denote the peculiar mentality or mode of thought of peoples whose ideas concerning the cosmic laws governing the forces and phenomena of nature apparently differ from °ur own. The expression „magical thinking" is all too vague, and, as such, misleading. The idea underlying the practice of magie is, of course, that the magical act will have a positive effect. But alongside the sacred world in which the art of magie is practised there is the profane world of common, practical, daily life in which the knowledge of things is adhered to, obtained from observational experience. If the thoughtworld of magie is to be conceived of as underlying the ritual practice of the magie art, there is, of course, nothing preposterous in harbouring such an idea. But we cannot speak of a trend of thought, in a generalizing manner, as belonging essentially to a thought-world of magie conceived of as characteristic of a mentality peculiar to a certain stage of cultural development and from which the practice of magie as such is supposed to have sprung. Alongside ritualistic practice in the world of magie, there is always the ordinary practical life that is lived under non-magical conditions from day to day. Magie, in fact, belongs to the sacred world; the ^ Inde Antique p. 145 sq.; Oldenberg-Lehre der Upanishaden .ïnleitung). 2) Oldenberg — op. cit. p. 13. ordinary life of every day is lived in the profane world. It is naturally inadmissable to regard the religious conscicusness as consisting of a certain devout state of mind, in most cases a form of piety as we moderns understand it, and then to set about enclosing the older religions within the narrow confines of that idea of piety. It behoves us to take a religion as we find it, however far it may be removed from our own conception of what religion ought to be. The sacrificial ritual of Brahminism may strike us as having but very little to do with piety as such (taking that word to imply an intensively developed individual religious feeling of a warm and all embracing nature), but whatever and however we may think of the Brahminical ritual of sacrifice, we have to remember that it is a ritual and as such constitutes part and parcel of religion. The very fact of the Brahmins having subjected the entire ritual of sacrifice to such meticulous and minutely detailed elaboration is cogent proof that their dealings with the world of the sacred were dictated by motives of extreme prudence; it was a world in which they walked with cautious steps, for was it not fraught with dangers for him who did not scrupulonusly follow the prescribed order of performing the ritual acts, and was not the order of the Cosmos dependent upon the faultless celebration of the religious rites? We might say that the sacrificial ritual was of a more negative than positive character, that the very basis of it was the taboo character of the sacred world itself. Now whatever else such a system of ritual may be, it can never per definitionem be called philosophy. The twin „roots", therefore, of Indian religion, the belief in gods (nature-worship) and the sacrificial ritual (magie) both lie deep in sacred soil *); it would even be better to say, to keep to the same metaphor, that they are both roots of one and the same plant. The sacrificial ritual is not onlyinconceivable without the gods, its very existence is dependent upon the gods to influence and control whose actions it is performed; they constitute its „raison d'être". The Standard a religion is n it is really better to restrict the use of the term, Magie, to those ritual practices which are of an essentially anti-social character. A great deal of confusion will thereby be prevented, there being then no possibility of mixing up „Magie" with „magical thinking and such-like. recognized by is not a particular mode or form of thought, but, as we have already seen, the subtle distinction made by Dürkheim between the sacred and the profane worlds. The sacrificial ritual is not a ritual because the idea underlying it is a magical thought-world, but because it is a ritual. By starting from these two elements as separate points of departure, the curious result has been arrived at, that both elements are found to neutralize one another. For the belief in gods is said to proceed from nature worship. By means of ritual, likewise based upon a sense of impotence over against the forces and phenomena of nature, upon a magical thoughtworld 1), the power of man toinfluence and control the Cosmos, who comes more and more to realize his ultimate and utter powerlessness in the things beyond his ken, is transferred to the gods, in other words, man rids himself of his responsibility. And now what should we expect to happen? We should naturally expect that the gods would come to occupy a higher position in the estimate of men, being thus invested with vaster powers. But that is just what does not happen. The personification of the forces of nature becomes weaker and weaker. Magie and nature worship, both result in the belief in gods, each by itself. In India, where they are associated, we perceive no intensification of this process. On the contrary, magie swallows up nature worship, and, so, does not thrive and blossom into religion, but sickens and becomes as lean as the cows in Pharaoh's dream. It is evident, therefore, that magie and nature worship cannot be taken together in one and the same argument without the argumentation itself breaking down. But supposing we are willing to assume the existence of these two „roots", even then we are as far away as ever from the solution we are seeking, for, after religion has pursued the tortuous path, beginning with magie, leading thence to *) Nobody will nowadays contend that it is essentially a magical thought-world which underlies the sacrificial ritual, a thought-world, moreover, which does not even hark back to the consciousness of a supernatural power in the sense in which the famous scholar, f.J* Frazer, understood it. We have purposely ignored all cnticism of the Frazerian idea according to which magie thinking is simply made to proceed from an erroneous association of ideas. Oldenberg, as a matter of fact, also speaks of „magical practices." religion and passing from religion to philosophy, there come into being in Hinduism, so, in the Epic we are studying, the figures of new and other gods. We might then expect that in the persons of these gods the older nature worship would begin to shine out again in all its pristine purity. „Their nature-origin is for the most part lost," is what Hopkins says, however, of the Hindu gods. *) It is more than we can do to walk the road, leading from the belief in the forces of nature of pre-Vedic times (i.e. the forces of nature personalized in the Vedas as gods) through the labyrinth of a pre-Vedic thought-world of magie, and ending in the polytheistic belief of Hinduism (no longer based upon natureworship, the Hindu gods having for the most part lost their nature-origins). In order to acquire a clear idea of epi~ mythology, we have been obliged to touch upon the subject of pre-epical religion. We shall not confine ourselves however to mere criticism, a relatively easy thing in itself, we shall also endeavour to point out another way of arriving at a likely solution. We shall leave the subject of the origin of religion and the belief in gods in ancient India out of consideration, the time not being ripe as yet to hazard any presumptions on this vexed question. In the preceding chapter we set forth what we ourselves understand by the idea of Brahman, viz. the ritual word, the employment of which by the celebrant in the prescribed ceremonial ensures the maintenance and perpetuation of the order of the world as conceived and expressed in the classificatory system. Behind that cosmic order there must be a creator who is thought of as standing aloof from the world he has created, whoever, in the Vedas, that creator may be. In the Epic it is manifestly Brahma. The gods themselves are part and parcel of that cosmic order; hence it is the gods whom the ritual ceremonies concern. To be able to perform the prescribed ritual the celebrant must be initiated into the mysterious workings of the order of the universe, the cosmic order as understood and expressed in the classificatory system. It is only when armed with this cosmic knowledge that one can tread the sacred world with impunity. From this point our argument has to follow two more or 1) Hopkins-Rel. of India p. 355* less diverging roads. The sacred act based upon sacred knowledge becomes elaborated into the detailed sacrificial ritual of the Brahmins. This sacred knowledge by means of which the possessor comes to occupy the place of the establisher of the cosmic order, gradually develops, closely connected with that order, into the philosophical theology of the Upanishads. Knowledge (jnana) comprises the entire classificatory system and is elaborated in the remarkable Samkhya philosophy, which even in its epic form, as we have already seen (p. 122), is found to be closely related to the ancient classificatory system, as the word itself (samkhya = enumeration) clearly shows. The object of all true „knowledge", as we have said, is the cosmic order, the ritual word, Brahman, virtually Brahma himself. He who knows (i.e. he who has come to know the laws that govern the order of the cosmos), knows himself to be an integral part of that order, realizes his oneness with the Absolute; he is himself the Absolute. He knows, too, that all things have their substitutes, that all things are merely rüpas of one another, and ultimately of Brahman itself. He becomes himself Brahman by virtue of his cosmic knowledge. This is the speculative philosophy of Atman-Brahman. In this connection it is better, in our opinion, not to make use of terms like pantheism etc. The speculative philosophy of Atman-Brahman is not immediately based upon the monistic philosophy of the identity of subject and object, upon the idea of the complete identification of the individual with the transcendental, but is simply a product of religious speculations upon the mystical identity of things within the confines of the classificatory system. That is why this cosmic 1) Is it not safer to avoid hete the use of terms like monistic and dualistic? This philosophy is not born of introspection, it does not proceed from an analytical examination of one's thoughts and feelings; it springs from religion. The Creator engenders the Cosmos; to a certain extent, therefore, we are at liberty to speak of a dualism constituted by the Creator and the thing created, the Cosmos, the Maker being distinguished fiom the work of his hands; on the other side, however, there is Brahman, the cosmic order itself, the two being one and indivisible; here we have to do with a purely monistic idea. The Indian mind does not, as a rule, concern itself with the origin of the world as primaeval chaotic matter. When the Creator causes everything to return into himself again, there remains the primeval abyss of waters, upon which he floats. It is also frequently said, that the Creator himself was born; (of an egg, for example). knowledge is also a mystical, a sacred knowledge. The peculiar manner in which the new thought is expressed in the ritual language is evident from the following passage: „Er schaut, zum untern Holz machend den Leib, das Om zum obern Holz, Durch der Versenkung Quirldrehung verborgnem Feuer gleich den Gott".1) Alongside this cosmic knowledge is the holy act (karman). The holy act concerns the performance of the sacrificial ritual. Jnana and karman are, therefore, not mutually exclusive ideas, on the contrary, they are complements of one another. It is not exactly the difference between mysticism and legalism. The man who has attained the higher knowledge, comes to consider himself as being more or less identified with Brahman. He does not occupy any special place in the cosmic order; he is above it. He has, therefore, no place in the caste-system, he is above it. The Brahmin, the man who „acts", occupies a high position in the cosmic order; as the celebrator of the religious rites he is at the same time the upholder of the order of the world. The Brahmin in general attachés very great importance to caste prerogatives. The Sramana, the man who „knows", considers himself, as an individual, above the differences of caste. This transcendental perception of the Sramana should, strictly speaking, not be called Pantheistic Mysticism. It is by occult knowledge and not by the grace of God that things are revealed to the Sramana. Knowledge, not faith is the soul of his religion. So far the jnanamarga is even more individualistic than the karmamarga of ritual. For by means of ritual the Brahmin is at any rate able to do something for his fellow-men. 2) In our opinion the Vedas, the Brahmanas and the Upanishads do not succeed one another in a straight line, the one immediately behind the other, even though they follow upon one another chronologically. 3) In the Brahmanas we find one aspect of religion, a meticulously elaborated ritual, in the Upanishads another aspect, a philosophized theology. Along- ') Oldenberg — Lehre der Upanishaden p. 280. 2) Oldenberg — op. cit. p. aoo. a) See also Holtzmann — Mah. I p. 62. side both of these manifestations of the religious consciousness and certainly closely connected with them there might very well have been the less complicated religion of everyday life of the people. This normal religious life, not particularly elaborated into a circumstantial ritual on the one side, nor into an intricate system of philosophical theology on the other, makes its appearance in the world of the Epic. In the Epic the gods are more personal than they are in the Vedas, not because the imagination has her? been allowed a wider range of creative activity, but because the ancient classificatory system is there characterized by a stronger leaning towards a systematized theology than was earlier the case. Sacred knowledge is now almost entirely detached from the idea of ritual thus becoming more and more a systematized theology accessible to the wider circle of seekers after a more tangible, more personal religion. There is one drawback attaching to our idea of the evolution of Indian religion, when the continuity of Vedas, Brahmanas, Upanishads, Epic is not seen as progressing in a straight lire; and it is this: it is not always possible then to indicate the prototypes of the epical deities, for the simple reason that nothing has been recorded of this prototype in the existing literature. The position of Vishnu-Krishna as the All-God, for example, cannot be inferred from the Brahman doctrine of the Upanishads, that doctrine itself being already a specialization of a distinctly individuil character. There is practically nothing known at all about Vishnu and Krishna in the older literature of India. But we prefer this leap into the dark — for such may be called our conception of things on this point — to any other leap, which, in our own opinion, would be no better than a salto mortale, a jump to the death. For the rest a lack of verifiable data must naturally press not only our own opinion but likewise any other, against the wall. As far as we ourselves are concerned it is by no means out of the question thai the mythology of the Epic is closer to that of the Vedas than to that of the Brahmanas and the Upanishads, even though chronologically the distance between the Epic and the Vedas is much greater than that between the Epic and the Upanishads. As a matter of fact this is Barth's opinion, too, of the mythology of the Epic, which, he says, in contradistinction to the teaching of the Sakyamuni, lays claim to continuing the Vedic past, yea, even claims to be that Vedic past itself. And yet Barth sees in the Sivaite and Vishnuite „sects" the natural outcome of the „insuffissance de la vieille théologie brahmanique".*) It is not so much a reaction that we are inclined to see in Vishnuism and Sivaism, new „sects" (the word is in general use) whose origin is due to a rebellious or protesting mood, a spirit of revolt against a form of religion that had become all too speculative. It would be far better to say, in our opinion, that the ideas familiar to us from the Epic as Sivaite and Vishnuite, in a word as Hinduistic, existed alongside of and simultaneously with the Brahmanas and Upanishads without the existence of such ideas being mentioned in as many words in the Brahmanas and Upanishads. The persons who are speaking in that literature simply did not concern themselves all too much with the everyday religion of the profanum vulgus, the common herd. The myths recorded by them are, therefore, but few in number. They were well enough acquainted with these myths, but there were other things that claimed their attention. We ourselves consider the myth of the Amritamanthana, for example, to be exceedingly old, not judging by its epical form, of course, but by its significance. The myth is intimately connected with an old ritualistic practice, the making of the sacred fire and the churning of the milk of the sacred cow. And yet the myth, as far as we know, is not even alluded to in the Brahmanas, although the ritualistic practice attaching to the myth is constantly spoken of. Nor has the Epic been able to steer clear of the shoals of a theologizing philosophy. The flow of the narrative is also interrupted by that insatiable passion for philosophic speculation so characteristic of the spiritual culture of India. Large portions of the poem are devoted to this sort of religious literature. If we are to see in Hinduism a religious reaction, a striving after a more tangible religious experience, we shall naturally have to assume, that after that reaction there came a revival of *) See p. 99. the old religion again,resulting in other kinds of literature, especially the Upanishads, being accorded a place in the Epic. In that case we should have to speak of a re casting, and, in the spirit of Brahminism, of a reconciliation with those features of the poem which were decidedly Sivaite and Krishnaite, of assimilati°n etc., all this having first been rendered necessary by presupposing a reaction, which can only be imagined to have taken place if the continuity of the older literary data is admitted to have progressed in a straight line. The correctness of our own view upon the subject being granted, and the criticism we have given of the manner in which the ideas expressed by the words, religion (philosophy) and myth, are understood, being considered sound, all this is unnecessary, and there will be an end to that „swaying" of critical research between Philosophy and Religion upon its weary way through the mazes of the Epic. Then the critical student of the Epic will enjoy the great advantage of being able to take the text as it is, and there will be no need for him to see in what he may find the other side of an older stage in the development of religious thought which has not been found. !) If the ancient adage, simplex veri sigillum, be true, our own opinion may also be considered to have its „raison d'être" and to be not unworthy of further consideration. * * * In the two great gods we are now about to discuss, viz. Siva and Vishnu, scholars are out to discover the two main elements, to unearth the two „roots", as we have already called them, of the religion of the Epic. Vishnu is the personalization of that „pantheistic" idea which grew out of the belief in the all-embracing, all-pervading dynamical power of Brahman, while Siva connects with Rudra, of whom it is imagined that he is to be identified with some sort of terrible and destructive natural phenomenon. The Epic, as we have said, is Vishnuite in so far as it is Krishnaite. Krishna is *) We shall purposely avoid discussing the subject of Buddhism, as we do not feel called upon to form an opinion upon this point. It is certainly quite possible that Buddhism is to be seen as a sect, as a reaction against the existing system of religious belief. constantly being referred to as the one, only, universal god, the Absolute; hence the idea which naturally urged itself upon the investigators of this conception being the unmistakable product of a system of monistic philosophy of the Absolute, an idea, the reasonableness of which a mere perusal of the Epic must inevitably confirm, for it must strike the reader, that in that pre-eminently Krishnaite poem, known as the Bhagavadgïta, for example, an inordinately large portion has been given to philosophy. Vishnu-Krishna's divinity, however, is just as theistic as it is pantheistic. In order to explam this the construction of the following process of development is had recourse to: Brahminical pantheism; a reaction from the side of theism, whereby Krishna is thrust intotheforeground; succeeded in turn by a revival of Brahminical pantheism, by means of which an „Ausgleich" (compromise) is brought about between pantheism and theism, resulting in the two erstwhile conflicting systems of religious belief coming to exist alongside one another in a reconciling and levelling Hinduism. Alongside Vishnu-Krishnaite Hinduism and chronologically precedent is Siva-Rudra, originally a mighty and terrible god, who gradually came to occupy a more central position in the cult. Sivaism in more pantheistic form succeeded pantheistic Vishnuism; Siva was at first called the All-God by his worshippers after the example of the All-God, Krishna. Some of the Upanishads unreservedly call Siva the Universal God, but in the Epic, according to epical investigators, Siva was first thought of pantheistically in those passages which (just because Siva is pantheistic in them!) are considered of later date. „At first neither (Siva nor Vishnu) was more than a single god without any philosophy . *) In order to make our argument more intelligible we shall not discuss Siva and Krishhna together, but take Krishna first and then return to Siva. There is, however, one drawback in this mode of treatment: we are obliged to separate what will prove to be inseparable. Krishnaite and Sivaite belief are inextricably bound up with one another. When attempting an explanation of the figure of Vishnu-Krishna we immediately find ourselves up against an almost ïnsurmounti) fTnpkins — Rel. of India p. 460. able obstacle: in pre-epical literature there is practically nothing known of either Vishnu or Krishna. In Vedic literature the one thing by which Vishnu is characterized is the fact of his having taken as a young man three tremendous strides, of which the third was the highest. Within these three strides of the god who is said to be no longer a child is contained all and everything that has fallen to the shire of men. Now without indulging in all kinds of conjectures as to the obscure significance of these three strides of the „wide-striding" god (which are generally associated with the daily journey of the sun across the heaven in its rising, its ascension into the mid-sky, and its setting), we might mention, that it is more evident from the Brahmanas, that Vishnu was a dwarf, whom the demons imprudently allowed to take as much of the earth as he was able to cover in three strides. It is also constantly and emphatically stated to his renown, that in these three strides he covered the whole, wide earth, thereby taking possession of it and giving it to men for a dwelling-place. 1) In the Epic this god, who rides upon the bird, Garuda, is, nevertheless, a mighty deity, composing together with Siva and Brahma the trinity of the Indian pantheon.2) He is especially associated with the sun. Although Brahma is the creator, Vishnu repeatedly acts as a demiurge, taking upon himself the actual work of creation while Brahma remains „absorbed in the contemplation of his own inward musings". He it was who urged Brahma during the Amritamanthana to cause the Gods and Asuras to prepare the drink of immortality, he himself not disdaining to render them actual assistance in their churning operations by assuming the shape of a turtle and holding up the churning-staff. When the Praia ya, the consummation of the world, comes about and all the elements have become merged into one another, only Brahman is left. This Brahman proves to be Vishnu. Vishnu then falls asleep and on being awakened by Brahma again causes everything ) Oldenberg is of opinion, that we must not see in the number, three, someconcrete natural phenomenon or other, but merely a preference for the number in ïtself. The idea would then be to express, that the e®'I'e ea"h was won for men by Vishnu by his three strides. (K.a.v. 231). 2) As to Vishnu in the Epic see the rich work of Hopkins-Ep. Myth. to evolve anew. Brahma arouses him from his sleep to set things moving again upon the All-Father being disturbed by demons in his recital of the Veda. Brahma himself is then seated within the lotus, proceeding from Vishnu's navel. It will be noted how strangely their various functions are divided among the gods. Brahma, the creator, holds himself so aloof from the work of his hands, that the process of evolution and involution threatens to come to an end without his deigning to interfere. Brahma is even said to proceed from Vishnu. Vishnu is the Universal God, not only in so far as he is srashtri, but also as samhartri, in which quality he is identical with Rudra or with Yama. It is of importance to remark here, not only, that Brahma, as a result of his identifying himself with the cosmic order was compelled a priori to remain more or less apart from the world he had created, but also, that in a circulatory system now the one side now the other side is uppermost. A proceeds from B, B from C, and C in turn from A again. Vishnu's consort is Lakshmi or Sri, Good Fortune, whether it be wealth or happiness, to possess whom the Devas and Asuras quarrelled upon her emerging from the depths of the agitated ocean during the Amritamanthana. Concerning Krishna there is even less to be found in pre-epical literature. *) There is mention of a certain Krishna, to whom is ascribed the authorship of several hymns in the Vedas. Further there is a Krishna in the Chhand. Up., the son of Devaki, who is there the disciple of Ghora Angirasa. In the Epic Krishna is entirely identified with Vishnu. In the Harivamsa he is Vishnu in his incarnate anthropoleonine form, the man-lion that came out of a column and tore Hiranyakasipu to pieces. He is also Vishnu in dwarfish shape, whom Aditi, the Mother of the Gods gave birth to after a thousand years of pregnancy that he might overpower Bali. And here we have the story of the three strides, in which the dwarf covers the entire universe, when Bali, prince of demons unsuspectingly allows him to take as much of the earth for himself as he is able to measure in three strides. The measure of Vishnu, the dwarf's three steps was the whole universe, and there being no room left now for Bali, the demon, he was 1) Hopkins — Ep. Myth. p. 217. taken, bound and cast down into the netherworld. Hopkins further makes mention of Krishna, the „god, without end, unborn, born in the house of Yadu", whose mother is Devakï. As such he is identified with Narayana, who in turn is identified with Vishnu. Beside Narayana there is Nara, denoting Arjuna. And Krishna is also one with this same Arjuna, the mighty hero of the MBh. They are one being in two persons (sattvam ekam dvidha kritam). And then, finally, there is Krishna, the man, the son of Vasudeva and Devakï. His elder brother is Balarama or Baladeva together with whom he does all manner of feats of strength and skill. One of his greatest deeds is the slaying of his mother's brother, Kamsa, who had slain Krishna's six brothers when they were but new-born babes, a mysterious voice having prophesied to him in the chariot when he was driving away his sister and her husband in a wild outburst of affection for the former, that he would be slain by her eighth child. Kamsa's command to slaughter all children born of his own sister's womb, is compared with the massacre of the innocents of Bethlehem at the command of King Herod. In the Epic Kamsa is the son of his mother's brother, who is slain by Krishna, because during the lifetime of Krishna's uncle, his authority had been usurped by Kamsa. Arjuna, with whom he is one as Narayana is with Nara, is also a cousin cf his, viz. his father's sister's son. Later on Arjuna married Subhadra, Krishna's sister. The two cousins are complements the one of the other, considering they are boih invincible, when they are together, but neither of them when each is alone. This Krishna grows up originally as a cowherd. Many a page in the Puranas is filled with Krishna's pastoral adventures, which are largely of an amorous character. Krishna is as fond of women as his brother, Balarama, of the intoxicating delights of wine. His is the enjoyment of the embraces of 16000 women. Although Krishna is identified with Vishnu, he does not seem to have been a man of importance in every respect, being afraid of his relatives (jnati) and even being obliged to flee from Mathura, his native city, because he is afraid. Nor is it at all an easy thing to locate the cult of Krishna. For, although Mathura, his native city, may be identified with the city of the same name on the Yamuna, where, according to Megasthenes, Heracles was accorded divine honours, the great hero in whom one recognized Krishna it is not so easy to locate Dvaraka, which, it is true, reminds us of another Dvaraka in Gujarat. In the sixteenth book, in which is described the ruin of Krishna s people, mention is made of Dvaraka being swallowed by the ocean. It is wellnigh impossible to attempt to locate the Krishnaite cult at all, when we think, that Krishna's Purï, lies at the bottom of the sea. Mathura and Dvaraka are his cities, but only the City of the Gates (Dvaravatï) is sacred. x) Besides his amorousness there is another trait of character which makes this remarkable figure stand out in a strong light: he is entirely unreliable. As a warrior he is censured for his ignoble actions (anaryena jihmamargena). He is continually represented as an impostor, who manages to carry his point by employing all kinds of fraudulent devices, at the same time endeavouring to make it appear as if his deeds were the immediate and logical consequence of the subtleties of Dharma. In this exposition we have folio wed Hopkins. But is must not be imagined that it is possible in itself to make out, where Krishna acts as an incarnation of Vishnu, where as All-God, and where as a mere human being, as Kamsa's arch-enemy and as the crafty head of the Vrishnis. Always and everywhere it is the selfsame Krishna who is spoken of, and the various distinctions which have been in this manner introduced are simply due to the complicated structure of the figure of Krishna as such. Not all his traits are worked out in equal detail in the Epic. Krishna is also worshipped in the shape of a child or babe, a form of worship, which, at a later date, was even to become of great consequence. Concerning his doings as a child and, likewise, concerning his weakness for women, his dallyings with cow-girls,the Epic has not so much to teil as the Puranas. This leads Hopkins to conclude again that „like the lover-god the child-god develops later".2) *) Barth — Oeuvres I p. 157: „la Ville des Portes . 2) Ep. Myth. p. 216. Then shall Narayana destroy the city with the Satvatas and return again to the heaven of Brahma. Bhandarkar sees in the Satvatas the votaries of Vasudeva-Krishna, the human embodiment of Narayana. It must be assumed, according to Bhandarkar, that Krishna-Vasudeva, the human manifestation of Narayana, is the proclaimer of a system of monistic philosophy (Ekantika Dharma), which even before Krishna must have grown up round about the figure of Narayana, the AllGod. Krishna, the preacher of the new monistic belief in Narayana, was later paid divine honours, a case, according to Garbe, analogous to that of the Buddha, who was also raised after death to the eminence of a god by the over-zealous devotion of his disciples.1) Upon one point, however, the line of thought does not run smoothly. The argument seems to us to break down where the identification of Krishna with Arjuna is concerned. We can understand that a person who propagates a religious system should in course of time come to be elevated to divinity. But the quintessence of the figure of Narayana does not lie in the first instance in the identification of Krishna with Narayana but in that of Nara with Narayana, and this couple is analogous to that of Arjuna and Krishna, each couple being the essential complement of the other. This identification seems to have been of great importance, the attention being drawn to it not only in a number of places in the text but also at the commencement of each separate Parvan. S. Lévi was of opinion, as we have already said, that it is this very identification which must be considered as constituting the outstanding feature of the entire Epic. Krishna is certainly a teacher. That has been definitely established by the Epic. But we can see nothing of the teacher in the Arjuna of the Epic. He is simply one of the most doughty heroes of the Pandavas, this „nelmetwearing" Arjuna. It does not do, then, in our opinion, to see, in turn, in the hero, Arjuna, the deification of the disciple of the preacher of the new Narayana doctrine, in itself a really all to naive way of looking at things, and one, moreover, which is not borne out by the facts of the case themselves. We should then be obliged, as a matter of fact, to assume, that *) Die Bhagavadgïta 1905. the Epic was written at a time when Arjuna's apotheosis had not yet become an accomplished fact, Arjuna never showing the essentiaUy „divine" qualities of a „full-fledged" god like Krishna. We have only to look at him, so to speak, to see that he is but an ordinary human being, however favoured bis position among men may be. Now if we are to assume that the Epic was really written at 3 time when Arjuna's elevation to the divine dignity, by way of Krishna as a stepping-stone, was only yet in its inceptive stage, then we are compelled to come to the conclusion that the Epic must have been composed before the time of Panini, it being unanimously agreed upon in the light of a passage in Panini, that in nis time there existed something like a worship of Krishna and Arjuna. There will be no one who would think of according so venerable an age to the Epic in its entirety. The identification of Ar juna with Krishna does not, therefore, in the hypothesis, that Krishna is the deified founder of the new monistic religion, occupy the place it obviously deserves. Then, moreover, there is difficulty in dating the Epic. A less serious objection is, though an objection all the same, that it is said of the Narayanas in the Drona Parvan, that they are the sworn enemies of Knshna-Govinda, in spite of the fact that they are also called the followers of Vasudeva. x) So in that case we find it necessary to distinguish not on'y various Knshnas, but also various Vasudevas, whereas the Epic itself, it seems, is not concerned tvitn distinguishing either group at all, whether together or apart. The Epic, as we have already said, merely teaches that Krishna, Krishna-Govinda, Narayana etc. are all one and the same person. There have been other investigators who have assumed the existence of a separate Krishna as the local god of some group or other, e.g. a fire-god, a sun-god or god of lightning. Hopkins, for one, in his „Religions of India" (1895), following, perchance, in the footsteps of von Schröder, seems to have seen in Krishna also a solar deity as the god elect of an anti-Brahmimcal group of theistical devotees, whose struggle against the stronghold of Brahminism, in the vicinity of the present Delhi, is supposed to have constituted the historical back*) Hopkins — Ep. Myth. p. 214. ground of the epical narrative. It was apparently Hopkins' intention to raise this Krishna to the dignity of All-God by identifying him with another Krishna, who is supposed to have preached the abolition of the old religion of sacrifices, proclaiming in its stead the worship of the solar disc, the supreme light (Vishnu). He does not, therefore, represent Krishna's apotheosis as the result of an extreme veneration for the person of the founder of the new Narayana faith, but as proceeding from the worship of an existing god of an antiBrahminical, later pro-Brahminical tribe.1) In an essay, written in iyo8 J. Kennedy also distinguished several Krishnas.2) A Krishna, „hero of many a solar myth, the slayer of demons, who (in his combat with the daitya Salva) dives under the sea (the war between sun and darkness) and slays Kamsa and Ke?i and Madhu. This semi-agaricultural, semi-solar, or atmospheric god is evidently connected with the dark sun and the storms of the rainy season, and his shrine is at Dvaraka on the sea-shore, where the sun dips into the boundless Western ocean". His brother is the famous Balarama, famous in the first place for his drinking-bouts, a plough-beaiing harvest-god. Another Krishna is „the original Krishna of the Indus Valley, non-Aryan hero and semi-Aryan god, Aryan only in name". It was with these two Krishna figures that the later idea of Vishnu-Narayana became linked up under the influence of the growth of a striving after a monotheistic form of religious belief. It was not until 300 A.D., however, that Krishna's identification with the cosmo-philosophic Vishnu became complete. In the controversy that ensued upon the appearance of Kennedy's work, A. Berridale Keith disputed Krishna's solar character, assigning to him the status of a demon of vegetation. Now, however different between them the various theories may be in their conception and elaboration of the figure of Krishna, whether we think of the founder of a religion raised to the status of a god, or of some tribal god or other associated As far as we can remember Hopkins did not return again to the idea of the Pandavas having originally been an anü-Erahrmmcal tribe with Krishna as their tribal god. His „Epic Mythology is a work of a more descriptive character. 2) J.R.A.S. 1907/08. with the sun or something else, or of a wily tribal chieftain, or whether we think of all these possibilities combined — one thing is certain, viz. that Krishna's proud entry into the Brahminic pantheon is thought of as a mighty wave of theistic belief that swept in upon the still waters of pantheism, causing it to surge and swell. This disturbing wave of theistic belief, sweeping in upon a perfect pantheistic calm, came from elsewhere. There is but little of this theistic stir to be discerned in the Epic; the reactionary wave of theism has worked itself out, and the worried waters of pantheism have come to rest again. Here we really see no deified figure, no personal god; here Krishna is the Supreme Lord, the God of All, in the absolute sense. It is repeatedly said that Krishna is not only Vishnu, but that he is also Siva, the god of that other „sect", which we shall discuss more fully later on. Arjuna is even represented as paying homage to Siva in the shape of Vishnu and to Vishnu in the shape of Siva (Sivaya Vishnurüpaya Vishnave Sivarüpaya namah), and Krishna, too, makes sacrifice to Siva — surely the last word in religious toleration! „Onemayrepresent",saysHopkins,„theattitudeofaKrishnaworshipper in the Epic somewhat in this way: „Krishna is a modern incarnation of Vishnu, the form which is taken in this age by the Supreme Lord. You who worship £iva should know that your £iva is really my Krishna, and that the chief point is to recognize my Krishna as the Supreme Lord. The man Krishna is the Supreme Lord in human form. Of course, as such, being the One God in whom are all things and beings, he is also all the gods known by names which designate his special functions. Thus he is the head of the gods, the FatherGod, as our ancestors called him, Brahma; and he is all the gods known by still older names, who are the children of the secondary creator, Brahma, viz., Agni, lndra, Sürya, etc. All gods are active manifestations of the Supreme God called Vishnu, who is born on earth to-day as Krishna." And the Civaite says: „£iva is the manifestation of the All-God," and repeats what the Vishnuite says, substituting £iva for Vishnu, but with the- difference that the £iva-sect has no incarnation to which to point, as has the Vishnuite." x) But if *) Rel. of India p. 405. this idea of Hopkins is really rooted in fact, we are bound to ask why on earth there should be any talk at all of two different sects. Once it is conceded, that one's own god is, at the same time, an other, altogether different kind of god, it would hardly be logical to speak any longer in a differentiating manner of a strange god and one's own. There can no longer be any further talk, then, of either Vishnuism or Sivaism. There can be only one thing in that case. Then we cannot speak any longer of a theistic religion, merely of a system of philosophy. Not thus is the state of things. Siva is not Vishnu, nor is Vishnu Si va. We are not here concerned with the question of a logical similarity and equality of the nature of: A = B, therefore B = A; it is a religious identity that has our concern. Vishnu is Siva, but that does not imply, reasoning upon a religious basis, that Vishnu is one and coincident with Siva; he is neither more nor less than Siva; the fact is, he is a different being. It all sounds very strange, but we must simply take things as they are. And further — is it in the nature of sectarianism to be so tolerant? On the whole the Hindus are tolerant enough, but once they are drawn into the maelstrom of religious controversy, they are no more tolerant than any other religious „sect", Eastern or Western.1) No, philosophy and religion are certainly not one and the same, however closely they may be related. Religion is a matter of life, of action and of ethics above all things, thinking and pilosophizing being only a secondary concern. Moreover there is practically nothing that points to that older stage of religious thought, when Krishna-Vishnu was still „a single god without any philosophy". We have to do with the same apparently absurd and yet unmistakable conception of Vishnu, who is Siva and nevertheless continues to be Vishnu, and of Siva, who is Vishnu and yet remains none the less „his own indivisible and inalienable self", but then upon a much larger scale, when the Devas as a group are opposed to the Asuras as a group. In the ancient religion of Iran the corresponding words are employed in exactly the opposite sense. The Supieme Deity of ancient ') See Dubois-Manners and Customs p. 295. Persia bears the name of Ahura-Mazda. In the religion of India the two terms are not discriminated with any degree of precision1). Often enough Asura is a name given to a god, while Deva may just as well be used to designate demons of any kind. This is evident from the use of the word, devajana, to denote serpents and demons. ,,Das Wort hat offenbar eine neutrale Bedeutung, man könnte damit ebenso wohl den Begriff „Damon" wie die Vorstellung „Gott" verbinden".2) Asura, then, may just as well be employed to denote God. In the Brahmanas the Asuras are associated with darkness. The day belongs to the gods, the night to the Asuras. The latter, nevertheless, are as much the children of Prajapati, the creative deity, as are the gods, according to the Taittiriya Brahmana, where he is represented as creating in succession Asuras, Pitris, men and gods. Hence the Asuras are the elder brothers of the Devas. Gods and demons are constantly at war for the hegemony (rajyahetoh) of the world, as, for example, the Amritamanthana clearly shows. „Perhaps the earlier usage shows that no great difference was feit between gods and demons. When good, a god; when destructive, a demon". 3) Here we obviously have the same thing again: the Devas are the same as the Asuras, but they are not identical with them. If we are to understand this apparently illogical relationship, the orly thing to do is again to call up before the mind the organic structure of the classificatory system. Our discussicn of this system in the preceding chapter was based upon a form of tribal relationship after the pattern of a circulatory connubial organization, i.e. a tribal organization composed of l) According to Oldenberg Asura originally meant „possessor of occult power ' in a neutral sense, as belonging both to good and evil beings. He supposes the word, Asura, to have undergone in India in the long run a change of meaning, it coming finally to be used in a bad sense only. We would observe in this connection, however, that magie is not a neutral act, always being distinguished as having a pro-social over against an anti-social aspect. In its pro-social aspect we prefer to call it religion and not magie. The celebrant in the sacrificial ritual does not ,jdeal in magie", unless the sacrifice be one of a magical character. It behoves us to bear in mind that there is no such thing as an „a-moral" religion. In Iran, then, furthermore, according to Oldenberg, the older gods were cast down from their eminence and transformed into devils by the all too enthusiastic devotees of a new faith. This negative demonstration is obviously unacceptable. (Oldenberg — R.d.V. p. 158 ssq.). !) E. Arbman-Rudra. Uppsala 1922 p. 149—156. 3) Hopkins — Ep. Myth. p. 52; Macdonell-Ved. Myth. p. 156. three intermarrying groups. In the chapter on social grouping, however, we explained that a circulatory system is perfectly well conceivable in conjunction with a division into two exagamous groups called phratries, the so-called dual organization. Let us leave the circulatory system for a moment and devote our undivided attention to the potent influence that a division of a tribe into two phratries may ultimately exercise upon the classificatory system. Our purpose in ignoring for the time being the question of the circulatory system, is merely to avoid making our argument unnecessarily involved. We shall later on (p. 223) return to a discussion of the circulatory system in connection with the dual organization prevalent within it, bringing each into proper relation with the other again. The two exogamous phratries, the two tribal moieties, come together at the gatherings of the clans, the great tribal meetings. They live in a state of constant rivalry, although they are entirely dependent upon one another, one phratry not only being in connubium with the other, but each phratry also having to assist the other in the celebration of the tribal ritual, for instance, in the initiation ceremonies. In the classificatory system the two rival phratries are usually identified with all kinds of things constituting each other's natural complements, e.g. with male and female, light and dark, heaven and earth and such-like. x) In India, where, as we think we have made phusible, a patrilineal grouping is intersected by a latent matrilineal grouping, each phratry by itself is composed of a male and a female half. In this respect, therefore, each phratry constitutes in itself a perfect whole, which, in the classificatory system is expressed, as we have already seen, by male-female. So it can be said, that the two phratries together constitute the tribe as a whole; it can also be maintained, that the whole is also constituted by each phratry in itself, by each half or moiety of the tribe. The two patrilineal phratries are to one another as man to woman. But when we turn to matrilineal descent we are equally justified in saying, that the two phratries are identical, the patrilineal dual division being intersected *) See p. 122. by the matrilineal. *) Here, then, with the literal facts of the case before us we are confronted with the to all appearances illogical logic of A being the same as but not identical with B. Now the contrast between Devas and Asuras is, in our opinion, merely the contrast in another form between the two moieties of a tribe. Heaven is the same as earth; with one reservation ofparamount importance: the same as the sacred earth. The social organization which prevails in heaven is, therefore, conceived of as being the same as that upon earth. Society in the sacred world exhibits the selfsame structural characteristics as human society in the profane world. The classificatory system according to which the Devas are associated with heaven and the Asuras with the earth is, therefore, merely an example of the ordinary divasion of things between two phratries, each phratry being associated with two complementary contrasts. Enough has been said concerning the classificatory system to show, that it is not an accidental contrast which has here been made, and which the Devas and Asuras have inadvertently become involved in. That is the reason why we shall not delve any deeper into the multitude of other contrasts of varying nature attaching to the system in question (contrasts such as light and dark; day and night etc.). In the ingenious scholastic system of the Brahmanas we should undoubtedly discover any number of extremely curious contrasts. Let us leave it at that, only remarking, that the true nature of the Devas and Asuras was not belied in the practice of ritual either; we have seen them co-operating with one another in the performance of the important ritual of the Amritamanthana, literally working together hand in hand; at the same time their typical rivalry is also clearly shown in the same ritualistic ceremony. We might very well be tempted to ascribe all this — this attempt at bringing about a unity of opposites — to a fantastic whim of the Indian mind, which all manner of seeming oddities can so easily be attributed to, things we are at a loss what to make of or what to do with; it behoves us, however, to be extremely careful; we have here to do with a system, not with a mere coincidence, with design, not with ') See above p. 54, 95. chance, a design which becomes all the more comprehensible once we come to identify the contrast between Devas and Asuras with that existing between twin tribal phratries or moieties; oneness and dividedness, with the completeness of the whole in the part, and the necessity of the coming together of the parts to make the tribal whole. We shall now retrace our steps and return to Krishna, following the same path by which we left him. And here, once more, we shall have to do with the idea of phratry division as it is clearly expressed in the division between Devas and Asuras. For in Krishna and his brother, Balarama, we see before us the two brothers representing the two tribal moieties; they constitute the couple, of whom one is the benefactor, the other the arch-deceiver or trickster, the two culture heroes that are to be met with in a number of religions as two friends or brothers. They usually live together in a state of constant antagonism. Frequently one is wise, the other foolish. In the Epic Balarama is the elder and Krishna the younger brother. Balarama is white; Krishna, as his name implies, is black.1) The benefactor-trickster couple can always easily be distinguished from other mythical figures, occupying, as they do, in mythology a place peculiarly their own. It is always in the interests or in the service of men that they are occupied, and they participate in some way or other in the actual work of creation. They are a kind of mediators between the Primal Being who engendered the world, the cosmic order, and human beings, the living work of his hands, coming more into prominence and developing greater activity as the Primal Being becomes more closely associated and identified with his handiwork, the order of the Cosmos.2) ') Krishna, though it mean „black", does not, on that account, of course, mean „negroid"; it does not, therefore, imply Dravidic ongin. Krishna has not only a „white" brother, but is also identical with the „white" Ariuna. „Le magnanime Héros de 1'epopee appartient-il a la même race que ce „blanc" Arjuna?" (Masson — Oursel in L Inde Antique p* 202). There are no racial differences m heaven. s) For our knowledge concerning the culture-heroes we need not only lay under contribution Prof. de Josselin de Jong s lectures at the University of Leyden, containing, as they do, data enough to draw from: there is also a paper by the same scholar, published in the Meded. Kon. Akad. v. Wetenschap. Afd. Letterk. Deel 68. Serie B No. 1. (1929). Contr. Royal Acad. of Sciences. Lit. Sec. Vol. 68. Series B No. x. (1929) and entitled: The Origin of the Divine Trickster. As we wish to restrict ourselves to the Epic, being unable to discuss the Puranas from want of time, our interest and attention will be concentrated not upon the two brothers, Krishna and Balarama, but upon Krishna alone. It need not be feared that in so doing we shall be leaving half of the figure of Krishna out of consideration, Krishna being a complete mythical figure in himself. He is not the trickster only, his brother being the benefactor; he himself is benefactor and trickster in one and the same person. The idea of brotherhood is merely an expansion of Krishna's doublé personality; his contrasting qualities being embodied in two separate personalizations, a common enough thing in the world of mythology. Krishna, then, represents the totality of the two phratries. The state of sexual tension existing between the two tribal groups reaches its culminating point in the person of Krishna; witness his pronouncedly sexual nature. There is no end to his amorous adventures; and the number of women he may call his own is estimated in the Epic at 16000. His worshippers pay him divine honours as Rudra with the phallus.x) He is also identified with death. As far as we know, Krishna is not an Hermaphrodite, although the being androgynous is frequently enough characteristic of the benefactor-trickster elsewhere. He leads a doublé life between the upper and the lower world. As Narayana he is white, as Krishna he is black. Krishna is a benefactor. It is he who takes the reins as Arjuna's charioteer, and assists him in battle, slaying a number of demons. Even as helper and protector he does not belie the innate duplicity of his nature, not only aiding the Pandavas by giving them good advice, but also assisting their foes, the Kauravas, by furnishing them with warriors. Krishna is a culture-hero. It is he, for instance, who teaches Arjuna the Lord's Song or Bhagavadgïta. As Vishnu-Narayana he is also the demiurge, the creative god himself, who, aroused from sleep, starts the Cosmos „turning again". That does not mean to say, of course, that we think of Vishnu and Krishna as never separated from each other. The fact is, there is but little known of Vishnu and it is by no means improbable that we shall never succeed in determining what part Vishnu really *) MBh. XII. 47. 56. played in the ancient Brahminical pantheon. For us it is only of importance to remark, that Krishna as Vishnu is the creator. Obviously, then, Krishna may also manifest himself as a culture-hero. Krishna is a trickster. His deceit is not downrighta criminal. It is more of the nature of a subtle cunning, a crafty attempt at sparing the goat as well as the cabbage. One of his artifices, for example, is to transform a weapon into a garland of flowers; another, to make it appear as if the sun is setting. Let us take a typieal specimen of his deceptive dealing, viz. in the passage describing the famous mace-fight. The situation is familiar. The tremendous eighteen days' battle is terminated by the titanic mace-fight between the warriors, Duryodhana and Bhima. The duel is full of a tense and dramatic excitement; Bhïma is by far the stronger of the two, but Duryodhana engages his antagonist more dexterously and tactically. Which of the two will be victorious ? This single combat between two great fighters can never be terminated in the ordinary manner. There is Herculean strength over against Ulyssean skill, but strength will not slacken nor skill fail. If the combat is to end a deus ex machina must intervene. He does in the shape of Ar juna, who, at Krishna's instigation, signs to the mighty Bhïma to lay low the proud Duryodhana by giving him an unexpected and, therefore, an unfair and cowardly blow, viz. by striking him upon the thigh.1) At first sight, perhaps, that is not as it should be; but the way out of the moral cul-de-sac was this: Bhïma had once sworn to shatter Duryodhana's thigh, thus to avenge the gross insult put upon the noble Draupadï by his insolently baring that part of the body in her presence. Bhïma kept his vow and laid low by a cowardly stroke his hated foe. Duryodhana, as so many of his fellow-warriors, is made to bite the dust by treachery (adharmena hatah). Baladeva became so incensed at this perfidious act, that, had Krishna not restrained him, he would have rushed upon Bhïma himself and slain him in his rage. Nothing and nobody, however, can prevent him from mounting his war-chariot and returning in !) It was for bidden in ancient India in single combat with maces to strike one's opponent upon the thigh, thereby depriving him of his balance, just as in modern boxing it is forbidden to strike below the belt. disgust to Dvaraka. Here we have the fraternal couple KrishnaBahdeva in its two distinct manifestations of benefactor and deceiver. Is Duryodhana now defeated? He is — and he is not; on the one hand, practically speaking, he did succumb; on the other hand, he was, to all intents and purposes, unconquerable. And what about Krishna with his tricks? Undoubtedly a deceiver from one point of view: a man shall not be made to sin. But he may also be looked at from another point of view and no fault be found in hem: he must insist upon a man's keeping a once made vow. Krishna, then, stands in between right and wrong. It is exactly this ambiguous position, however, this standing „between the devil and the deep sea", so to speak, which accounts for the deceitful side of Krishna's character. „His nature"—says de Josselin de Jong, speaking of the divine trickster — „is still divine, though bordering upon the human: he is the mediator between gods and men, standing upon an equally familiar footing with both parties alike, a central master of ceremonies, in whom can be discerned, as circumstances shall dictate, either a servant or a god. It is exactly this indeterminateness in his dealings with gods and men which constitutes one of the outstanding features of the culture-hero. In myth, to be sure, it is not seldom that Creator and Benefactor become merged within one another, although, as a rule, the latter is manifestly and even more or less emphatically depicted or characterized by his actions as a being of a lower order than the former." * * * The person of the god-deceiver is admirably adapted to being allied with all kinds of other mythological figures, for instance, with the Primal Being or with a divinity of the upper or nether world. In each individual case as it presents itself it will behove the investigator to consider the various possibilities. The reader will have been struck by the fact, that, so far, there is one facet of Krishna's composite character upon which no light has been thrown, that of his infantine manifestation. On the mental plane he has a dual nature, on the physical *) The Divine Trickster p. 24. plane he has a dual form; he is also a child, and as a child-god, especially in later times, he was accorded a measure of divine homage which was destined to assume greater and greater proportions, his cult as divine infant becoming almost as popular as the Christian cult of the child Jesus. Now the reader of the Epic will remark, that Krishna's doings as a child-god are not given that detailed treatment which is accorded to his doings as a god-man. It is exactly in connection with the increasing significance of Krishna-worship as the cult of the divine child in later times, that it has been concluded, that the mention of Krishna's doings as a child-god is to be considered of later origin. We have reason to fear that we are becoming monotonous, but it behoves us, however that may be, to remark once again, that the discriminating between a number of different elements in the composition of the Epic must be devoid of all sense unless the student gets to know how it was that the various elements thus discriminated could have been brought together and united so as to constitute such a uniform whole. If a general, synthetic idea of a thing is to be acquired, one must be possessed of some analytical insight and vice versa. And otherwise the entire Krishna problem must be doomed like the proverbial snowball to go on rolling along its twofold path of theologico-philosophical speculation, githering substance in the sense of solidity and losing in clarity in the sense of lightness of texture as it goes, finally to assume the alarming proportions of an avalanche of incoherent and disconnected fa ets sweeping through the Epic and burying the primitive poem beneath a mass of débris. Albrecht Weber and lastly J. Kennedy have endeavoured to account for and explain Krishna's infant-god manifestation by assuming early Christian influences and such with especial reference to the narratives of Christ's early years, especially those found in the Apocryphal Gospels.*) Kennedy traces that historical contact with painstaking accuracy and arrivés at the conclusion that Culture spread from West to East and that that eastward moving Culture carried with it after the third century of our era elements of Christian thought and x) J.R.A.S. 1907 p. 51; ib. 1908 p. 169; ib. 1917 p. 209, 469. religious belief. The majority of invesrigators are by no means satisfied with thïs explanation of Krishna's infantine godship, simply because there is not the slightest reason for affixing to just these stories of Krishna's early years rather than to other Krishnaite narratives a date as late as 200 A.D. and later. x) It is, on the contrary, very much more likely, that these stories are much older. Take, for example, the antagonism between the youthful Krishna and his maternal uncle, Kamsa, who had the offspring of DevakI, Krishna's mother, done away with while they were yet young, a deed for which, as we have already said, an elucidative analogy is sought in the Massacre of the Innocents of Bethlehem at the command of King Herod. It is cut of the question that this event should be dated as late as the Christian era, mention being made of the Kamsavadha as early as Patanjali. 2) As to the Kamsavadha, the struggle between Krishna and his mother's brother or with his mother's brother's son, Kamsa, 3) we would make a few observations over and above what we have already said. This ceremony should be performed upon Krishna's birthday, which is celebrated upon the 8th of the month of Bhadrapada, i.e. at the height of the rainy season. That is the reason why Krishna is so frequently associated with the sun and given the status of a solar deity. He is the sun obscured by clouds (krishna), the personalized solar disc, who does battle with Kamsa, the demon of the dark, and after a desperate struggle slays him. Concerning the passage referred to above from Patanjali's Mahabhashya, Weber observes, that the Kamsavadha is apparently a kind of dramatic representation of Krishna's slaying of the villainous Kamsa. 4) In the Kamsavadha ritual the parties were divided into two *) Jacobi — E.R.E. s.v. Incarnation. l) This was of course not unknown to Kennedy. He sees, however, in the pre-Christian Kamsa episode the continued conflict between the sun and the demons of darkness. It must be admitted, however, that this method of inquiry is altogether too arbitrary. s) Apparently Kamsa may be either Krishna's mother's brother or his mother's brother's son (Dubois — Hxndu Manners and Customs p. 624; Hopkins — Epic Mythology p. 214). *) Beside the Kamsavadha mention is also made of the Balibadha i.e. Vishnu's conflict in his dwarfish avatar with the demon, Bali. This might lead us to conclude, that even then Krishna was connected in some way with Vishnu. 12 groups, one party, that of Kamsa, having black faces (kalamukhah) and the other, that of Krishna, having red faces (raktamukhah). 1) Weber is surprised that Krishna, the black one's party should have red faces. But such is not at all peculiar in considerition of Krishna's twofold nature, red over against black being merely an elaboration of the selfsame classification according to contrast obtaining in a society with a dual phratry organization, a system of classification which opposes the „white" Balarama to the dusky Krishna. Another instance in which we have to do not with natural but with mythological colours. When we think of these colours as being indicative of a dual organization, it is not because it happens to fit nicely into our own particular theorv, but because Kamsa and Krishna are positively representative of the two phratries.2) Kamsa is either the maternal uncle or the mother's brother's son of Krishna. The dramatic representation of a ritual antagonism between the two phratries, in which one party is the mother's brother and the other the sister's son of the former, (in the beginning the pass we party), or maternal uncle and nephew respectively, at once naturally calls to mind the ceremony of initiation in which the mother's brother is the initiator designate. 3) We would bring some points to the fore which may be J) Ind. Stud. XIII p. 487 ssq. 2) It is by no means unusual for the two phratries to be opposed to one another both as brothers and as cousins, just as they arerepresented as being brother and sister or as man and wife. In his interesting essay on Buddha and Devadatta Hocart comes to the conclusion, that the antagonism between these two cross-cousins is to be understood in the same sense as the antagonism between Krishna and Kamsa nas been explained by us. Their antagonism is that of „Good and Evil , „based on the rivalry of two intermarrying groups." The genealogical tree reconstructed by Hocart (see p. 78) shows that he, too, connected this antagonism with a dual phratry-division. , ... 3) For particulars concerning the employment and significance 01 the initiation motive in Javanese myth the reader may be referred to the works of Dr. Rassers, of Leyden. De Pandji roman (1922); Over den zin van het Javaansche drama. (Concerning the meamng of the Javanese drama). — Bijdr. tot de Taal-, Land- en Volkenkunde Nederlandsch-Indië 81 (1925); Over den oorsprong van het Javaansche tooneel. (Concerning the Origins of the Javanese Drama) ïb. Vol. 88 (1931). A short outline of the first two works in the Arch. f.Rel. Wiss. XXX p. 174. Religionen der Naturvölker Indonesiens — J. P. B. de Josselin de Jong. The new light thrown upon the subject by Rassers hasshownustheway in which we ourselves should go upon our journey through the labyrinth of the MBh. deemed of importance in this connection. The ceremony of initiation is the occasion of the gathering of the entire tribe for the purpose, in a succession of festivities which frequently continue for a considerable space of time, of admitting the young lads of the tribe, as they cease to be children, to adult life and to full membership of the community. The young man has now come to man's estate and is prepared to take upon himself the responsibilities of manhood and to enter upon the duties of such as are actively concerned with the we'1 and woe of the community; and not only the care for the well-being of his fellow-tribesmen has become his concern but even the maintenance of the very order of the Cosmos in so far as that Order is dependent upon the tribal ritual to be performed, directly or indirectly, by the male adults of the tribe.x) Upon that occasion the initiate is submitted to a number of ordeals for the purpose of testing his mental and physical endurance, the manner in which he undergoes his sufferings, hardships or abstinences being evidence of his „manliness". As a rule it is enjoined upon him to spend a certain time in seclusion from his own; he lives for a space of time in a forest or some other solitary spot, where he is gradually initiated into the mysteries of cosmic existence. Prior to initiation he is, practically speaking and from the sacral point of view, an unborn being. He has not yet been admitted to the performance of the sacred rites and ceremonies of the tribe and it is therefore impossible for him to participate in its sacred life. In the ceremony of initiation the initiates are supposed to suffer death followed by rebirth. When the young man returns to the common everyday world from his state of seclusion, he not infrequently affects to be a new-born child, babbling a meaningless and unintelligible jargon and having yet everything to learn.2) ') The following observation in Crooke's Rel. and Folkl. of North. India p. 186 is noteworthy: „Initiation marks the admission of the child to caste privileges and obligations, and children, dying before this rite, are usually buried, not cremated." High caste children in Bengal are permitted before initiation to eat with other children of any pure caste. !) The consecratory ceremony which has to be gone through prior to the offering of sacrifice, the Dïksha, is also a dramatic representation of human birth. The Dïksha hut represents the maternal womb; the robes in which the man who is to be consecrated (dïkshita) is clad are theinvestingmembranes. The Dïkshita cannot even speak articulately or intelligibly, but babbles like a little child (Oldenberg — R.d.V. p. 407). All this can be seen clearly in India in the Upanayana, the beginning of the probationary period that the novice must pass through under the supervision of a Brahmin. The state of the initiate throughout the period of probation is considered identical with that of one who has been ritually set apart as sacred for the sacrifice (dïkshita). The dïksha, as we have seen, is a mythical birth.1) But however that may be, he who has passed through the Upanayana, has been born again (dvija). There being no doubt as to the Upanayana really being the process of initiation, we here strike up against the idea, therefore, that the initiate experiences a spiritual rebirth through the ceremony of initiation, as it were; 2) this being born again is generally imagined as a being swallowed up by a terrifying demon, the spirit of initiation, which carries the initiate within it. The initiatory birth, is conceived of in exactly the same manner as the dïksha, the mythical birth. The sanctuary or sacred place, in which the rites of initiation are performed, represents the body of the initiatory genius. As such then, the sanctuary also represents the idea of the totality of things, the ceremony of initiation being, at the same time, the recognition of the initiate as a fully responsible member of the community. This explains why the initiatory daemon is not associated with any one clan in particular; it symbolizes the completeness of the whole, the fullness and maturity of life. It has repeatedly been said, that the distinction made between gods and men must not be understood as representing a contrast in the sense, for example, that mind is the opposite of matter. The world of the gods is exactly the same as that of men, the only difference being, that the world of the gods is concewed of as being a sacred world in contradistin ction to the profane world of the sons of men. As we have already said, it is the same world, and yet different; it is a ritual world. For the rest, the conditions prevailing in the sacred world are more or less the same as in the commonplace, non-ritual world; the selfsame contrasts, the selfsame rivalries and antagonisms are to be met with in the upper as in the lower world; the rivalry between the Devas and the Asuras is simply a phratry-rivalry. Both 1) See p. 179 n. 2. *) Oldenberg — R.d.V. p. 468. Gods and Asuras are compelled to render their active and co-operative assistance in the celebration of the sacred rites. We have seen that e.g. in the case of the mythical Amritamanthana ritual. Not men only but the gods, too, have to do sacrifice; they are as dependent upon the efficacy of ritualistic practice as are human beings, their very existence being determined by the sacred ritual's cosmic effectiveness. We even see, mainly in the Vedas, how the Gods as well as men are denoted by the selfsame term, Nara. Gods and men are inextricably bound up with one another through ritual just as are ritual and cosmic order through the ritual formula, Brahman. Now, it being incumbent upon the gods to take an active part in the celebration of the sacrificial ritual, we might naturally expect to find some indication, however hesitating, of an initiatory ritual among the gods. We employ the word, „hesitating", because the initiation rites practically constitute the right of admission to the ceremonies of the tribe, the sacred world, the world of the gods themselves. On the other hand, however, it constitutes an all-important rite, and, as such, can hardly be conceived of as pertaining only to the world of men, as lying outside the domain of the gods. The fact is, it could not and was not thought of as such. The initiatory genius itself was made to pass through a process of initiation, as it were, a process of initiation running parallel to the probationary ordeal undergone by the human initiate, with whom it was considered as being one in point of fact. The initiate is able to say with the Dïkshita at the moment of the sacrificial act: „That which is my body is in thee; and that which is thy body is in me 1), for the initiate is swallowed up by the initiatory daemon, and is born of the consuming spirit anew. This analogy between the initiatory genius and the initiate holds good to such an extent, that the initiatory daemon itself is also not infrequently represented as a little child. It is pictured as floating upon the primaeval waters, i.e. withui a world where the classification of created things by the assignment to those things of naman and rüpa has not yet been effected. It is in this selfsame stage „of being in and yet not one with life" that the youthful initiate finds himself. He has ') Sat. Br. III. i. i. 8; Oldenberg — R.d.V. p. 406 n. 1. not yet been initiated into the mysteries of his tribe's cosmic ritual; he has not yet learned from the men who are his initiators the indispensable knowledge of the existitig order of things; he has been taught nothing of sex and sexual intercourse ; he has not yet entered upon real life. That is why he belongs to the other side of the circle of existence, which is non-existence; he belongs to the passive, the female side of life. He still sleeps like a child in the cradle, unconscious and ignorant. But although we are justified in saying of the initiate, that he is even as a little child sleeping within the still world of the unclassified, to say the same thing of the genius of initiation itself would be bordering upon rank profanation. The initiatory daemon, itself a god, can hardly be considered to be livmg in complete ignorance of all it is about to reveal of itself to its other self, the youthful initiate. In order to link these two apparently contradictory elements together and establish an harmonious connection between them, the initiatory genius is imagined as a little child, drifting upon the primaeval abyss of waters, but also, at the same time, as the all-embracing daemon of initiation itself. There is another feature, too, by which the initiatory genius is usually characterized. He strikes terror into the heart and his invisib'e presence is chili with the fea*- of impending doom. This of itself need not surprise us all too much, for the mystery of initiation spells death; and the fearful ordeals are such, that this symbolic dying of a second death must needs take on in the eyes of the physically and mentally exhausted initiate the proportions of a grim reality. Here, then, there is a second reason for assigning to the initiatory daemon the abyss of waters for a dwelling-place, it being the waters by which the underworld is indicated. It will now be clear with what end in view we launched out upon our argument. It seems to us, that the recognizable motive underlying the antagonism between Krishna and his uncle, Kamsa, is the idea of initiation. And now to continue the same argument: In Krishna we are primarily concerned not with the genius of initiation itself, but in greater measure with the divine initiate. As such, he is also at the same time the genius of initiation. Indeed, how could it ever be said of Krishna, the divine trickster and benefactor in one, in whose person the two rival phratries are united into one who'e, how could it ever be said of this deity, that he was not also at the same time omniscient, thereby ccmprehending within himself the initiatory genius also? To speak thus of the Lord, Krishna, the All-enfolding, would hardly fall short of sheer profanation. And it will have given Krishna no difficulty to play this doublé part to perfection. He combined within his person the „easy" virtues of artful deception and a conveniently elastic moral consciousness. The great benefit redounding to the culture-hero from the duplicity of his character is that he can look both ways at the same time. Krishna, being a culture-hero, was given a sphere of action within the cosmic order in between the two worlds of gods and men, acting in his capacity of conciliatory mediator for the Common Good of the citizens both of the sacred and the profane world. And well, in this capacity, becomes him his innate twofold mentality. His very name, Krishna, the Black One, accords with the terrifying appearance assumed by the genius of initiation (he is generally called the initiatory demon); which is no contradiction of the more aimable side of his character in his passive röle of „suffering" initiate over against the terrible Kamsa. The colour of the actors in the Kamsavadha, as far as Krishna's party is concerned, is, on that account, red, i.e. „good" as opposed to the black colour of the partisans of Kamsa. The accounts of the child Krishna's tricks as we come up against them in the Epic are not worked out in such minute detail as they are in the Puranas. But there are also Mahabharata myths in which he is manifestly featured as the initiate-god. Worthy of mention in the first place is Markandeya's vision, which, at his request, he describes to Yudhishthira; he tells him what the world's appearance was at the time of its consummation, the pralaya, when the Creator lay outstretched upon the serpent, Sesha.J) The wise seer relates how the periods of the world's existence succeeded one another and how gradually the cosmic order that came from God disappeared from creation. Then was the hour come when gods and demons, man and beast and all living things should return to the source ») MBh. III. 188 sq. from which they had come — God. Naught remained but the vast primaeval waters; upon this endless ocean Markandeya drifted in aimless solitude, weary of seeking for a place of shelter. In the mldst of the desolation of waters, however, there stood a Nyagrodha tree, a tall and stately tree, with broad and strong branches, in which there was a canopied bed. And in that bed the saint, Markandeya, discovered a little child; and the child looked upon him kindly with eyes open wide like to a lotos-flower. And greatly did Markandeya marvel, that amidst the abomination of desolation only this one child remained of a whole world that had gone the way of Pralaya. And the wise man thought and thought, but his thinking, with all his wisdom, was in vain; he could not imagine who this child might be. He was a wise man, so wise that he knew what was, what is and that which is yet to be. But this one thing in all his wisdom he knew not. And the child, marked with signs of good omen, spoke to Markandeya; for it knew that the wise man was overcome with exceeding great weariness, desiring to find a place where he might find shelter. And lo! it offered him its own body to be a place of shelter. It opened wide its mouth, that Markandeya might enter in and take up his abode within its body; and Markandeya entered the child's body through its mouth and took up his abode therein. And within the body of the child Markandeya saw a new heaven and a new earth, and in the new heaven and the new earth were new men and new gods and new beasts. And there was no end to that new world, however far Markandeya went into the body of the child. And Markandeya knew now that it was a god that had given him his own body to be a dwelling place and a refuge from the horror of loneliness and the darkness of desolation. Then did the wise man take refuge with this god with heart and mouth and tongue; and through the mouth by which he had entered in did he come out again. And now he acknowledged that he had found true knowledge (jnana) and sure salvation. And Markandeya praised the god in whom he had dwelt, magnifying him. And lo! the child manifested itself to the seeking saint as the Lord Krishna whom the whole creation seeks. And there was not one 'in all the world that could comprehend this unfathomable mystery. For it was revealed to the saint in that hour that this helpless child in all its helplessness was the Be-All and the End-All of all creation. Not even to the high gods themselves was it given to fathom this mystery. Such was the vision that the rishi, Markandeya, told to the dharmaraja, Yudhishthira in his banishment in the solitude of the forest, the vision of the mystery of the little child that was the great Lord Krishna, the Lotos-eyed.1) From this beautiful tale we clearly see, that Krishna, the All-God, was also the divine initiate. His sleeping upon the primaeval waters is here made to bear upon the cosmic sleep of Vishnu, although it is evident, that it was in this instance an initiatory god that was thought of, Markandeya, by entering into the child's body winning salvation and acquiring knowledge (jnana). As we have already said before, it is extremely difficult, if not altogether impossible perhaps to entirely separate Vishnu and Krishna. The dwarf, Vishnu, who also occupies the entire Cosmos, is likewise the child of the Mother of the Gods (devamatri), Aditi, she being the same as Krishna's mother, Devakï. This fits in admirably with the conjecture we have proposed, that Aditi, the Great Mother, might very well be seen as the female counterpart of Brahma. Now once we assume Krishna to be the divine initiate, it immediately becomes self-evident why he has the reputation of being weak and despairing as a result of his relatives making things uncomfortably warm for him, why he becomes so afraid as to have to flee from his city, Mathura and why the Narayanas were his sworn enemies. Such is exactly the situation in which the initiate finds himself placed, obliged as he is to spend a certain time in the seclusion of the woods, afterwards returning a new man. Such, at least, is the construction we are inclined to put upon Krishna's sojourning among the herdsmen in the woods and in the mountains, where they temporarily picched their camps.2) *) The fact of the child's identification with Krishna occurring only at the end of this narrative leads Hopkins to conclude that this identification must be of a later date (Hopkins — Epic Myth. p. 213). This is pushing the interpolation theory a little too far. 2) The cowherds were called Abhïra, a name evidently denoting their occupation. It seems to us most unlikely that the word should originally have denoted a definite race and have had nothing to do with „cowherd" at all (Bhandakar — Vaisnavism etc. p. 37). Why it is that Now however important all this may be, of paramount importance for the study of the Epic it is not. For the study of the Epic it is essentially the identification of Krishna and Arjuna with Narayana and Nara which must be considered of paramount importance, as Sylvain Lévi has already observed. This identification concerns the Epic in its entirety. Let us take a single characteristic feature in itself which will clearly show what we are here concerned with. In the Bhagavadglta Krishna instructs Arjuna in the way in which he shall go, finally manifesting himself to his cousin as the All-God. Arjuna is given an other-worldly vision by means of which he is enabled to see what Markandeya saw within the child Krishna's body: the entire Cosmos, gods, demons, men and all manner of beasts. It was as the highest in each separate class of living beings and lifeless things that Krishna manifested himself — a typical manner in which to give material expression to the essential absoluteness of his All-Being, and one, moreover, from which it becomes evident that the term, pantheism,must be considered out of place in the present connection.The term, All-God, must not be understood as denoting that he is all things at the same time; it merely means to say that in a series of classified things and living beings he himself constitutes the highest form in each separate class. Understood in this sense there are manifold things which may be said to constitute his rüpa. But it was given to Arjuna to see even more. He is also permitted to look upon Krishna in his appalling manifestation of the „terror-inspiring", in his shape of horror (ghoram rüpam), for such as are not gifted with the higher vision a shape „difficult to see and one which even the gods would wish to look upon." Arjuna, sick at heart, sees his foes, the Kauravas swarm into Krishna's vast, extended jaws, there to be crunched to death. It follows, therefore, that Arjuna is the ideal initiate, being one with and yet not the same as the god-initiate, Krishna. There are various episodes in the Epic, which the space at our Krishna is spoken of as sojourning among cowherds of all people during the years of his exile is not at all clear to us. Have we perhaps to do with the sacred significance attached to the cow? We are also not able to assume any tribal indications in other names of peoples among whom Krishna dwells. Vrishni signifies strong, powerful. Satvatas is derived from sattva, meaning ènergy? disposal does not allow of our discussing in any way comprehensively, episodes, which, in our opinion, throw back to early initiation myths. We need only mention a passage such as that in which it is related how Arjuna in his banishment, after having come upon his brother and co-mate in exile together with the fair Draupadi, undertakes a journey to the East, to the South, and finally to the West, where he marries Krishna's sister, his mother's brother's daughter (Book I, 214, ssq.); and the episode in Book III, 37, ssq. in which it is related how the same Arjuna journeys to the far North, where he shall be initiated into the lore of arms. There, he is submitted to a trying ordeal by the mighty Siva in the shape of a woodsman. As the result of a heated altercation about the shooting of a wild boar which Siva had claimed as his target, a tremendous fight takes place between Arjuna and the disguised god. At length the mighty Siva tires of the unequal fight and fells the exhausted Arjuna with one powerful blow. Arjuna comes to, and sends up a prayer to Siva, supplicating the god to help him in his most dire need. And Siva is moved with compassion, and manifests himself in all his glory together with his divine consort Uma to his amazed but happy „victim". Arjuna had won his celestid weapons; and in the palace of Indra he is initiated into the right use of arms (i.e. the careful employment in the ritual of the formulas; the fight itself is also a ritual) and instructed in the art of dancing (the ritual dance) by the celestial musicians.L) Dr. Rassers, upon whose works our exposition is based, was also struck by the figuring of the initiation motive in the epic narratives. He sees initiates in „these five sons of Pandu, who, the offspring of divine parents, slew numbers of evil monsters, and who, at a certain stage of their „heroic" career, were obliged to conceal their real persons behind false names and disguises, and who had to win back in battle that half of their lawful kingdom which had been usurped by their brothers, and one of whom, Arjuna, the hero of manifold amorous adventures, secured in a contest l) It is remarkable that in Book I Arjuna travels only in thiee directions; it being in Book III that mention is made of a fourth. Is there any natural connection between these two portions of the poem and have they perhaps become accidentally disconnected ? rainy season naturally stands out most clearly of the three, the dry and barren earth then rapidly becoming covered, as though by magie, with a luxuriant vegetation. But the rainy season is not only the season of budding life; it is also a season of death, the rotting vegetation as a result of the sweltering heat constituting a very hot-bed of disease and preparing the way for the outbreak of virulent epidemics. In modern India the belief obtains, that during the varsha, the rainy season, Vishnu sleeps upon the serpent, Sesha. The rainy season is the ominous period, the time of bad omen, in which it is forbidden to marry.1) Vishnu awakes from his serpentine sleep in the month of Karttika, i.e. four months after the rainy season has set in, from which time the marriage taboo is lifted. In Gujarat it is forbidden by caste-law to dye during the four months of the rainy season.2) The Kumhar (potter) is not allowed in this period to bake his pots. 3) This inauspicious period, this time of evil omen, from the middle of July until the middle of November is accompanied by a number of religious festivals. In the first month there is the feast of the Nagapanchamï. Upon this day the serpents are worshipped and represented. It is forbidden to walk behind the plough, to cut food, to cook food or to eat cooked food. 4) The feast generally falls upon the fifth day of the month of Sravana, but sometimes, too, in the second month. Sometimes the ceremonies are commenced in the month of Sravana and continued into the second month. In Nepal it is the common belief of the people that upon this day a terrible fight is raging between the sun-bird Garuda, Vishnu's bird, and the Nagas.5) In the month of Bhadrapada, the second month of this period of ill omen, is celebrated the feast of Gaurï, Siva's consort. An image of the goddess is made, consisting of a number of plants tied together, enveloped in clothes and to which is attached a swastika or fylfot. It is the custom in Southern India for everybody to pay homage to the special tooi, imple- 1) Moor — Hindu Pantheon p. 16; Crooke — Rel. and Folkl. of N.-India p. 314; Russell — op. cit. II p. 295. 2) Russell II p. 331. 3) Crooke — Tribes and Castes III p. 340. 4) Gupte — Hindu Holidays and Ceremonials p. 178. 5) Vogel — Indian Serpent Lore p. 275 ssq. ment, or utensil that he or she employs in his or her particular occupation.*) In this month or at the end of the month preceding falls also the feast of Krishna's nativity, the Krishnajanmashtamï. Weber, who wrote a most interesting study of this particular feast, found mention of it made for the first time in a text dating from the I3th. century. For this feast a special festive house is built, of which we shall have more to say later on. The symbolic birth takes place at midnight. The night is spent in singing and dancing and in the reciting of passages from the MBh. and concluded with the paying of divine honours to Krishna's mother, Devakï, and the giving of presents to a Brahmin. In the opinion of some scholars the festive celebrations are also attended with the worship of the lunar disc as an essential concomitant. 2) The second half of the same month is set apart as being an especially auspicious time for the sraddhaceremonial, the sacrificial offerings to the departed ancestors. In the third month of the Pitriyana there is a special episode of ten days. The first nine nights, the Navaratra, are dedicated to Dev! or Kali, the horrific female counterpart of Siva, represented with protruding tongue, wide-open mouth, having in her hand a decapitated human head, a sword and a pan of fire. 3) The stories from the Mark. Pur. are recited in which she is described as: „Kali of the terrible countenance, armed with a sword and noose, bearing a many-coloured skull-topped staff, decorated with a garland of sculls, clad in a tiger's skin, very appalling because of her emaciated flesh, exceedingly wide of mouth, lolling out her tongue terribly, having deep-sunk reddish eyes, (and) filling the regions of the sky with her roars." 4) He who reads the story of the combat bet ween this cannibalistic goddess and the Asuras „at the great annual worship that is performed in autumn time", will be preserved from all evil and, enjoying her especial favour and protection, be blessed with riches, fullness of corn and the *) Dubois — Manners and Customs p. 568. 2) Ueber die Krishnajanmashtamï. Abh. Kön. Ak. Wiss. Berlin 1867 p. 217 ssq. 3) Arbman — Rudra p. 266. Picture in Russell I p. 344; further IV p. 575; Moor — Hindu Pantheon. ') Maik. Pur. (Translation Pargiter) adhy. 87. possession of children.x) Upon the tenth day follows the feast of Dasara, the most important day of the whole year for the Rajpüts and Marathas. Then there is a great army parade. The royal Standard and the war-drum are carried in stite in a procession of great pomp and splendour. The king is seated beneath a Sami-tree, it being the traditional belief that the sons of Pandu had concealed their weapons in just such a tree. This tree is hung with branches supposed to be of gold. After doing sacrifice the king returns to his dweiling, where his consort moves a lamp about his face. He gives her a piece of money, a neckchain, a ring and other presents. The prominent citizens regale the people with all manner of public entertainments, consisting more often than not of single combats, which Dubois compares with the gladiatorial combats of imperial Rome. It is also usual throughout a great part of India for a buffalo to be slaughtered upon this day which is meant for the entire community of the village, the first cut generally being given by the „proprietor of the village", and the flesh being consumed by the lower castes. Russell will be right in connecting this usage with the eating of the totem animal by the whole tribe.2) The day on which the feast of Dasara is celebrated is held to be an especially auspicious one for starting upon a military undertaking. „Many of the historical expeditions will be found to have started on the Vijaya, or Success day." It is also a pre-eminently favourable day on which to begin the instruction of boys. The pupils of the schools recite verses at the houses of their relatives, and pass in procession through the town. 3) Upon the first day of the fourth month of the rainy season is celebrated the Dïvalï, the Lampadedromy or Feast of Lamps. This day is looked upon as a New Year's Day, being the day on which the demon, Bali was compelled to yield up the earth to Vishnu in his dwarfish shape. The nights during this feast are fraught with manifold dangers and terrors, being haunted. The disembodied spirits of childless persons walk abroad. The mothers pass with their children on their arms !) Ib. 91. 41 ssq. 2) Russell I p. 173. 3) Gupte — op. cit. p. 185; Dubois — op. cit. p. 509. from house to house, craving a few drops of ghee or clarified butter for their lamps, it being the child that gives light to the souls in the Pitriloka. In Bengal sacrifice is made to the departed ancestors. It is the time for the fields to be new-sown. It is also the day set apart for the bankers for the balancing of their books, the close of the financial year. It is universally held to be an exceptionally lucky day to dice or gamble upon, especially if the head of the family chances to win. It is said, that Siva, playing at dice upon that day with his consort, Parvatï, forfeited to her all that he possessed. x) Upon the Ekadashi of this month, the most momentous of the monthly eleventh day feasts of the year, a day of great religious significance especially for the Vishnuites, is celebrated Krishna's marriage to Rukminï, the Princess of Vidarbha. This divine marriage is symbolically commemorated by the marriage between the Salagrama (a black stone, considered a family possession of great value) and the Tulasi plant. The „marriage" is solemnized in exactly the same manner as an ordinary civil marriage between persons, being attended with festivities, merry-making and everything that an ordinary wedding brings in its train to add lustre and mirth to the occasion. 2) In these symbolical nuptial rites the owner of the Éalagrama stone represents the bridegroom and one of his female relatives the Tulasi plant. Four months later, in the last month of the second, the cool season, is celebrated the great feast of Siva, the Sivaratri, the most important of the nights devoted to Siva, when the cult of the linga enjoys prominence in the ceremonial revels. The Pongol feast is also celebrated in this month, being the feast of the winter solstice. One of the most important feasts of the eight months outside of the rainy season is the Ho li feast,3) a Spring Festival, and one especially popular among the lower castes, by whom 1) Crooke — Rel. and Folkl. etc. p. 344 ssq. 2) The „marriage" between Salagrama and Tulasi also represents the marriage of Mother Earth or of a pond or the like, with a male being, put in the form of a stick into the middle of it, there being otherwise no virtue or „goodness" in the water. (Crooke — Rel. and Folkl. of North. India, p. 64). 3) It is not quite clear whether the Pongol and Holi feasts are both celebrated throughout India or whether the Pongol feast is more espe- 13 it is celebrated as a species of carnaval.') Large bonfires are lighted in which the effigy of an animal or a human being is burnt. The effigy, bearing a decidedly phallic character, is supposed to be a demon, erstwhile slain by Krishna. The revellers add mirth and merriment to the occasion by throwing red paint at one another. The licentiousness which is given free rein in the singing of songs of an obscene nature and such-like upon the occasion is a striking and characteristic element of this festival. In some districts a sham fight takes place between the male and female participants in the ceremonial revels. Frequently connected with this festival is the so-called Dola Yatra, the rocking of Krishna in a swing. Here and there during this Spring Festival the Brahmins and other members of the higher castes are said to deign to touch the members of the lower castes, a thing they certainly would not think of doing under ordinary circumstances. It is a traditional belief, that a special main annual festival day has been instituted by Vishnu for each of the four castes: the high-day set apart for the Brahmins, receivers of alms, is the day of the full moon of Sravana, upon which day money is collected by the distribution of amulets or charms against evil among their masters, which evil-averting trinkets are tied round the wrists of the alms-givers; for the Kshatriyas there is the Dasara festival; Dïvalï for the Vaisyas and Holi for the Südras.2) We must beware, of course, of attaching undue significance to this tradition, but it is at any rate a tradition which tells us that we are justified in seeing in these four festivals the principal high-days of the four castes and furthermore that of these four religious holidays the three main ones fall within the rainy season. The rainy season is a time of war and of military expeditions, a period of antagonism, too, between the bird and the serpent, i.e. between the upper and the netherworld. It is the time in which sacrifices are made to the dead; it is the time of death, during which Vishnu sleeps upon the cially restricted to the Southern provinces. Gupte (op. cit.) and Crooke (Rel. and Folkl. of North. India) make no mention of the Pongol feast at all. It is mentioned, however, by Dubois (Manners and Customs, p. 571), and Hopkins speaks of both (Rel. of India, p. 448 ssq.). 1) Mirmal Kumar Basu — The Spring Festival in India. Man in India VII, 1927 p. 112. 2) Gupte — op. cit. p. 36. him to study trom morn tui nigni ana iu wum uuimg the hottest part of the day, for he is to be compared with the sun, whose heat is fiercest and his light brightest when his golden disc is highest in the heavens.1) It is just such an injunction which so clearly shows how closely linked up with one another are sacred knowledge, ritual practice and cosmic order. It would even seem as if the mighty sun himself were more or less dependent upon the recitation of the Vedas. This self-tuition is designated by the remarkable name of Brahmayajna. Now it seems very likely, on the strength of all the various data, that there once used to be a period of great and important feasts in ancient India in attendance upon a time of serious social crisis and involving also the entire ritual of initiation. This was the antagonism between Krishna and Kamsa. Weber wondered how it was, that the feast of Krishna s nativity could ever have come to be celebrated at the end of July whereas the feast of the birth of Christ which, in his opinion, was the fons et origo of the Indian celebration, occur red so much later in the year in the Christian ecclesiastical calendar. Kennedy, who also thought of early Christian influences, was inclined to see in this natal feast vestiges and reminiscences of an ancient solar myth connected with the Kamsavadha, thus practically thrusting a spoke in the wheel of his own attempted solution of the problem. It is by no means entirely out of the question that the Kamsavadha is some kind of symbolical dramatization of the continued contest between the sun and the darkness. It is exactly to the gathering up of contrasts of such a conflicting nature as that between sun and darkness, light and dark moon into a unifying whole that such a manifest form of rivalry as that existing between one phratry and another is so admirably adapted. But such is merely of secondary importance; of primary importance is the significance this natal feast formerly possessed, and perhaps, in a more modified form, even now possesses, its importance, that .) Weber. I. S. X. p. 115; 130. That the rainy season should have been set apart for the initiatory studies just because that time ol the year is so cool, as Weber supposes, is not, in our opinion, a very plausible conjecture, seeing the student had to study hardest during hottest part of the day. is, as a feast of the celebration of the rites of initiation. The very fact of this feast being celebrated at this particular time in contradistinction to the period of the year in which the Christian feast of Christ's nativity was celebrated constitutes in itself evidence of the erroneousness of the derivationhypothesis. While Vishnu sleeps upon the serpent, Krishna, who is entirely identified with Vishnu, is born, and contends with Kamsa. As we shall yet see, the rainy season has always been essentially the time during which Rudra-Siva, the destroyer, was uppermost. Now the very fact of KrishnaVishnu's being born in the very time of Rudra-Siva's ascendancy is new and confirmative evidence of that dualism so peculiarly characteristic of this culture-hero. In the rainy season Krishna assumes his terror-inspiring shape (ghoram rüpam), that of the initiation demon. But his terrifying character is also counteracted by his being born at the same time as the divine initiate. He has still to pass from the profane world into the sacred, to which he by divine right belongs. It is exactly this very character of contradictoriness, which suited him so admirably in his taking upon himself the doubling of the röle of divine trickster. x) The day of the Upanayana, of the entering upon the probationary period of initiateship, is still universally commemorated in Southern India in the month of Sravana. To mention but one of the many noteworthy customs prevailing in our own times as part of the modern ritual of initiation, it is the usual thing for a young men's feast to be given to those who have just entered upon their status of initiateship together with others who have been through the initiatory ceremonies a short time before.2) Everything is still reminiscent of the fact, that once upon a time the feast itself was a „cérémonie totale". As far as the age of the Epic is concerned, we may perhaps be justified in ascribing a certain amount of demonstrative force to the fact, that Krishna himself should have chosen of all the months of the year the month of Margasïrsha wherewith to identify himself, i.e. the first month after the rainy season. It M In modern India the Lord of Initiation is called Ganesa, the son of Siva by Durga, the warlike goddess of destruction. Even now this formidable god enjoys the selfsame universality as Krishna. 2) Dubois — Manners and Customs p. 166, 170, 161. is not improbable that this was the month in which the ritual of initiation was considered definitely completed by a final ceremony. Now the older texts contain a few remarkable passages, which might seem, at first sight, to shake the structure of our argument to its very foundations. It is written in Ap. I. i. 18, that the initiation of a Brahmin shall take place in the Springtime; that of a Kshatriya in the Summertime; and that of a Vaisya in Autumn.x) It is likewise written in Asv. G.S. 1.4. 1. that the initiatory rites shall be celebrated during the Devayana, although, according to others, the period of the year in which the initiatory ceremonies are performed is altogether immaterial.2) Such injunctions are in flat contradiction with the other prescriptions, according to which the young initiate's course of probationary instruction shall be made to take place in the rainy season. As a matter of fact, it must be deemed most improbable in itself, that the ritual of initiation, which has left such clear traces of its existence in so many different places, and set its unmistakable seal upon the whole conception and structure of Indian mythology, should ever have been entirely free from all restriction as to the time of its performance. We, for ourselves, therefore, are convinced, that whatever may point in the older texts to the absence of any prescribing of a definite time of the year for the fixture of the Upanayana must be considered devoid of conclusive force when weighed in the balance with other references in which positive allusion is made to a specified time for the celebration of the initiation rites of which the Upanayana constitutes the inception. On the other hand such passages in the older texts need not of course be considered absolutely meaningless. They may in turn be explained not as the result of any indeterminateness in fixing an invariable date for the initiation period but rather as the consequence of a decided aversion, prevalent in certain branches of the community, to the rainy season as such. It is also not unlikely that the very fact of that most important ceremonial, the Soma sacrifice, not being performed upon an invariable and specified date, but being by preference x) S.B.E. II p. 3, 174') S.B.E. XXIX p. 164. made to take place in the Springtime, must be considered in the same light. We have here to do with the selfsame aversion to the rainy season which induced the theological author of the S.B. to classify the rainy season with common people, Summer being associated with the Kshatriyas and Spring with the Brahmins.1) The literal facts, however, are in flat defiance of such a classification, the rainy season, of all times, being exactly the period in which Krishna himself saw the light of day and in which many a great religious feast is celebrated. We shall not concern ourselves for the nonce with further endeavotiring to ascertain why or how it was that a certain inclination came to prevail to shift the preference for the rainy season to another period of the year; we shall return to the subject at a later stage in the present work, approaching it from another angle. That will be when the other side of Krishna's Proteus-like nature has been examined. * * * Krishna now stands before us as the divine trickster, at the same time associated with the idea and ceremonial of initiation as the divine initiate, in which capacity he is identical with Arjuna, the mighty hero of the Pandavas. But this idea of contrast or antagonism in Krishna's mental make-up, which we have also come to see as a characteristic feature inherent in a Dual Organization of Society, has yet another side to which the student's attention must be devoted: starting from Vishnu, it is his opposite Si va; the antagonistic Asuras when we start from the Devas; it is the dark and terror-inspiring side of which the contrast is formed by the opposing ideas of light and peace: contrasts, which, in the synthetic figure of the al1embracing Krishna, have been gathered up into a neutralizing unity, although it is decidedly more to the side of „good" that Krishna inclines than to the side of „evil". In the following Chapter, therefore, it will be the figure of Rudra-Siva upon whom we shall now presume to focus the attention. l) Oldenberg (Weltanschauung der Brahm, p. 113). Yet even in the time of the Brahmanas the rainy season was still held in honour, it being written in S.B. II. 2. 3. 7. that the rainy season is a rüpa of all the seasons (op. cit. p. 106). IV RUDRA AND THE SABHA In the preceding chapter it has been expounded that a typical change of phases is also perceptible in Indian society with a changing „densité dynamique", as Dürkheim called it. A period of stronger cultural tension showed itself during and about the rainy season, which also involves a period of intenser life for the whole of Nature. In that time, in which probably the tribe assembled, the great festivals were held, a fact that may still be traced in the modern calendar. The immediate motive to these great tribal-festivals, the initiation, proved also to be connected in a certain way with the rainy season. Not only the time at which, but also the place where the great festivals are held is remarkable. When we wish to consider what the data show us in this respect, we do not aim at completeness, but restrict ourselves to marking some features, in order that our sketch of the Epic may gain in perspective.x) The choice of the sacral grounds does not depend upon chance, but is prescribed by various factors. It is not as if the tribe as a whole chooses suitable grounds somewhere, where anybody may come, but usually one of the clans, of course not always the same, acts as host and sends the invitations. Therefore there is no permanent tribe-sanctuary. Every time new grounds for the celebration of the great event are fitted up by the clan which invites the others, or existing grounds are restored and extended. If we consider the nature of these grounds on which the festival is held, we should in the first place pay attention to the significance of the events which take place there. The ceremonies performed on the sacral grounds are partly imitations of important episodes from the life of the ancestors. The officiants in the ceremonial are at the moment of the sacral action no ') Here we follow closely the treatises of Rassers e.g. in Contr. Geog. and Ethnogn. Mag. Neth. East Indies vol. 88 p. 371. longer the people familiar from everyday life. They are the ancestors themselves, representatives of the clan, in whom the officiant loses his own individuality. He therefore wears the sacred clan-emblems, paintings or masks. In India this has not been otherwise. The word Nara, which indicates a god as well as a human being, the word purusha with the same doublé meaning point to the fact that we need not take the, moreover unnecessary, trouble to make a precise distinction between pitaras and devas, even though a pitri is not always a deva. The place where the sacred tribal-ritual is performed must of course possess the suitable surroundings for the action of the ancestors. So the grounds represent the sacral world. As the system of classification is closely connected with the clanorganization, care is usually taken that in the laying-out of the grounds the place of the various clans is indicated. Where more particularly the place of the feast is a building this is done by assigning the two sides of this house to the two phratries. At several places therefore the two phratries are called „the two sides of the house". Not everywhere is the sacral meeting-place a building. At one place people confine themselves to clearing a spot in the woods and enclosing it, at another they deem it an honour to erect a huge building provided with beautiful adornments. One feature is naturally found nearly everywhere: the indication of the particular sacredness of the meeting-place. This requisite is often practically acquired by choosing a spot in the woods or in the mountains which is already different from the profane world owing to its remoteness from the main camp. If the isolation is not sufficiënt in itself the grounds are fenced off. Only the fully initiated are admitted, that is to say, only grown-up men. The people admitted to the sacral grounds form in this respect a more or less secret society of men. Women and uninitiated men are not allowed to know the details of what actually goes on there, and may onno account be present at the ritual proceedings proper or see the sacred objects of the clan except on pain of death. The separation of the sacral world is on the other hand no definite and absolute negation of common society. The sacral world is the same as the profane world, but quite different. People speak, act and live there just as in the profane world, and yet there is a difference. In some cases the sacral grounds consist of two parts, of which one part is open to women and the other, the sanctuary proper, is not. On the sacral grounds the two phratries are the acting parties. They help each other and take turns in the performance of the myths. At the initiation the duty of instructing the initiates rest on the members of the other phratry than the one to which the novices belong. As has been repeatedly remarked a certain differentiation is perceptible in the competence and task of the two phratries as regards this reciprocal rendering of services, of which we have a beautiful example in the organization of the Omaha-Indians. There one phraty, the phratry of the celestial people, is especially charged with what is specifically religious, whereas to the other phratry the group belongs which sees to what is more of a social nature, fighting and the provision of food. The same is seen in the organization of the great festivals. Attention may be paid to the economie, but also to the religious side of the question. Neither do the two spheres stand entirely apart in this case, but partly over-lap. The clan inviting the others, in whose territory the tribe assembles and whose sacral grounds temporarily become the tribal-sanctuary, does not only reciprocally render services to the other clans, especially that of the other phratry, as far as religious rites are concerned, but also as regards its social duty as host. People vie with each other in munificence and openhandedness, not exclusively to brag, but especially because they can set off their own importance and eminence through a brilliant reception, in the happy certainty that another clan will not misunderstand the requirements of this do-ut-des, when the opportunity presents itself. The clan acting as host does not think that it has acquitted itself of its task, if it restricts itself to fitting up a space for the reception, but in addition to this it does not let an opportunity slip by to show distinctly that the blessing of the gods also clearly rests on it as regards material things. This is done by excessively providing its guests with food, and preferably with so much that it far exceeds the capacity of everybody to consume it, and by putting up magnificent ornaments. The trees are loaded with exquisite things. When, as often happens, a house has been erected in the village to serve as a meeting-place, people try to make an epicurean heaven of it, with the doublé purpose of expressing the fact that they are in the wood away from profane society and that the guests are staying with a distinguished host blessed by the gods. This is the first point we want to stress: the house for the festival is the richly adorned heaven (for this is the place where the ancestors really appear), actually and virtually situated in the woods or mountains, so away from the profane world (that is in the first instance the world of women), an isolation which if necessary is especially indicated by an enclosure. After this concise introduction we shall consider some of the Indian data. First the modern house for the festival: the pandal. This is a kind of arbour of branches and leaves, erected before the main entrance of the house. „The top or ceiling is ornamented with paintings or costly stuffs, while the whole is hung with garlands of flowers, foliage, and many other decorations. The pillars are painted in alternate bands of red and white. The pandals of rich people are often exquisitely decorated. A propitious day, hour, and star are always chosen on which to erect these canopies. Then the relations and friends all assemble to set up the centre pillar, which is called the muhurta-kal, and to which they offer puja to the accompaniment of music".2) The pandal is usually decorated on festive occasions with toranas, a kind of triumphal arch of flowers. When the Südras erect a pandal, they place in the centre a green branch of the sacred fig-tree, in which they worship Vishnu. The Kshatriyas hang their weapons in it at festivals. At festivals: for rich people often have permanent pandals, in which they receive their guests, just as the clan-sanctuary may be a hostel, also when it is not used for ritual services. If one moreover knows that the pandal is erected „with much pomp and care" for all the more important Hindu ceremonies (initiation, marriage etc.), then it will not be thought too rash, if we doubt whether Dubois is right this time, when he calls such a pandal a picturesque alternative for the little accommodation that the Hindu house can offer its guests. In our opinion it is not easy to make a mistake, when one tries ') Dubois: Manners and Customs p. 154, 162, 232, 233. to determine after what example tradition has built this pandal in India. We now quote without comment the description which Weber gives of the house in which Krishna's birth-day feast is celebrated in the month of Bhadrapada. We only point out the fact that the maternity home (sütikagriha) is considered to be a home for shepherds and must as such have lain in the field. The richly decorated house for the birth of the divine initiate does not differ in this respect from the initiationgrounds which ethnography describes us. „Die aus Pisangstammen bestehenden Pfeiler sind mit lotusfarbigen (rubinrothen) bunten, resp. mit weissen, gelben, rothen, gestreiften, grünen Zeugen zu umhüllen. Es ist mit Sandelkranzen, Perlen und Juwelen (mit Schutzamuletten), resp. mit neuen wassergefüllten Krügen (kalaga), mit manichfachen Blumen und Früchten, mit Lampenreihen zu verzieren, mit Blumenkranzen, Sandel, und Agallochun zu durchduften. Es ist ganz wie eine gokulam, Kuhstall, resp. Hirtenhaus herzurichten und mit Hirtinnen zu erfüllen. Fesseln für Frauen, eiserne Schwerter, ein Opferpfosten nebst schwarzem Ziegenbock gehören hinein; ein Keule ist an der Thür nieder zu legen; Wacher sind hinein Zu stellen, desgl. die Göttin Shashthi und allerlei erlesene Speisen für die Götter(— Idole) in reicher Fülle.... die Wande (sind) mit Bildern aus der „heiligen Geschichte" (ausgeziert). Auch ist die Herstellung des sütikagrha von Musik aller Art, Tanz und Gesang zu begleiten". Other sourcesspeak of a mandapa, according to one „ist derselbe über einem geweihten Kreise, der seinerseits die Mitte eines durch eine Kuhhaut abgemessenen Raumes bildet, aus Pisangstammen zu errichten, mit vier Thüren zu versehen, mit Früchten, Blumen etc. zu verzieren, und ein schönes, buntes Zeltdach darüber zu breiten". Or: „(wird) anstatt des durch die Kuhhaut abgemessenen Raumes ebenfalls bereits ein mit allerlei mythischer Zutat nach der Art des Tantrarituals geweihter, heiliger Kreis, Namens sarvatobhadramandala vorgeschrieben".*) We want to go into a few items which are material to our argument. This relates in the first place to the posts of the i) Ueber die Krishnajanmashtamï. Abh. Kön. Ak. Wiss. Berlin 1867. p. 268, 270. pandal. Abbé Dubois states that the central post of the pandalJ) in particular enjoys a special interest. At the initiation the two central posts of the pandal are draped with a new woman's dress. It will not be necessary to collect much material about the post and therefore suffice it to state some of the data mentioned by Crooke. 2) In Bengal the Malïs worship the main post of the house under the name of Gumo Gosain. When they start building a new village, they first erect a post, under which rice, betel, sugar and a piece of red cloth is buried. Instead of this they often take a branch of the main tree of the mother village. The Nagas erect poles in memory of their dead, which often have phallic motifs. A pole is set upright in a pond, through which that pond „marries" that pole, that the water may be good and living. Newly-married couples worship the central pole of the pandal by walking round it. The Bhavads of Gujarat call the pole Manikstambha („rubby pillar"), because it is smeared with human blood to appease the mamo („maternal uncle"). This is a maliaous spirit, who haunts the pole. Images of the mother-goddess and of other gods are cut into the pole. After the marriage-ceremony the pole is thrown into the river and the place where it washes ashore is appointed for the meeting-place of the next tribal meeting. Very probably the central pole is the „mantapam deity, that is to say, of the pandal itself", the Mahurta-Kal, because the Südras place a symbol of Vishnu in the centre of this little building. Now it would not have been possible to teil by the poles themselves, which were erected in the houses used for the festivals in the time from which the epic poems date, even if they had been preserved for us by time, whether they had a ritual meaning and if so what that meaning was. Yet Indian architecture still yields a thing or two which is of importance in this respect. The shape of the pillar had a special meaning. To the worship of Brahma e.g. a pillar belongs, with a square shaft; to that of Vishnu one with an octagonal and to that of Rudra one with an hexagonal shaft. More remarkable still however, is the fact that the pillars often have at the foot the shape of a pot, so that it seems as if the shaft of the pillar stands Muhurta-kal (the god Kala-of-the-decisive-moment?) ") Crooke-Rel. and Folkl. of North. India p. 89, 90, 325, 404. in that pot. The shape of the capital suggests a pot with its mouth downwards. The explanation for this queer decoration is sought in the demands which building with wood made upon the architects. It is said that the pot was placed upon the wooden shaft to prevent the latter from getting soaked. Nor was this pot forgotten later on, when stone was used for building purposes and people wanted to give in stone an exact copy of the older oonstructions in wood. We must admit that this explanation does not entirely satisfy us. The usefulness of a pot on the top of a pole is evident. But what about the pot at the foot of the pole? It is said: to protect it against noxious insects. 1) But can a pot carry a pillar without breaking? If that is possible, we must take for granted that the pot was in the ground. But in that case the builders of later times who used stone were astonishingly diligent indeed when they placed the pot round the foot of the pillar above ground. Philippe Stern ascribes it to the influences of foreign art.2) The oldest pillar was according to him the mere trunk of a tree on which afterwards a bell-shaped capital was placed after foreign style, just as on the pillar of Asoka. This capital, he thinks, must afterwards have returned upside down at the foot of the pillar and have been assimilated to a pot. Without denying the possibility of a historie influence or of a techmcal explanation, it must be admitted, we think, that E. B. Havell s opinion is better founded. He sees in the lotus-shaped capital of the pillar „an adaptation to structural purposes of the same lotus-and-vase motive, which with different application, served for the symbol of the Buddha's nativity".3) That structural purpose then is the application of the pot in order to protect the pillar from the detrimental action of the climate. This pot was afterwards formalized to a lotus-flower. „The open flower with turned down petals.... suggested to them the heavenly vault supported by the pivot of the universe. The vase forming the base of the pillar stood for the cosmic waters, the shaft was the stalk of the mystic flower.... the bell- shaped capital '•! gïïd?'JoS: See the suggestive «O dis.inct pic,„« ' f'' j E. Havell-A Handboot of Indian Art. p. 4'- was the world itself enfolded by the petals of the sky". At another place he says of the pillar in the pot: „The pole or axis of the Universe.... was Vishnu's pillar, or churningstick or Siva's lingam." 1) The pot mentioned here is the Kalasa. It is not difficult to determine the ritual significance of the pot. It will suffice to read what Dubois writes about the use of pots at the upanayana. To inaugurate the „ishta-devata (or tutelary deity)" for this ceremony a pot is taken which is filled with water and then decorated with women's clothes and ornaments worn by women. The priest now prays the „ishta-devata" to go into the pot, „which becomes from that moment a female deity". In addition to this six earthen vessels are wanted for this ceremony, five small ones and one big one. The five small ones are filled with water and emptied into the big one. The big one is placed against the central pillar of the pandal in such a way that a garland of flowers covers the mouth of the pot. The five small ones, which are now called Brahma, Vishnu, Varuna, Rudra and Devendra, are placed against five other pillars of the pandal.2) Because gods sojourn in the pot religious homage is rendered to them, a usage which is so universal that it is not necessary to illustrate it fully with references. 3) In front of a Brahmin nuptial procession a kalasa with water is carried, which represents the universe with the chief gods and mothers of the gods, whereas the water in it is the seven seas. 4) The kalasa is also the pot wherein the milk is churned. At the Amritamanthana the words: „Mathyatam kalasodadhih" are spoken.5) So Havell rightly sees in the kalasa the underworld represented as a woman, in which Siva's linga turns round as a churningstick. If one rightly regards the capital as a pot turned upside down, this kalasa might be the representation of the celestial ocean, the pillar in the two kalasas then being a perfect representation of the cosmic motion; at the base the pillar revolves in the waters of the underworld, at the top in the world turned upside down, the celestial ocean. The kalasa in x) Havell-Ancient and Medieval architecture of India p. 53. 2) Dubois — op. cit. p. 164. 3) Crooke — op. cit. sub „Pot". 4) Russell II p. 366, 380. 6) MBh. I. 17.12. formalized shape is heaven as a lotus-flower. The two centres of rotation should represent the two births of Agni, in heaven and on earth, or the Amritamanthana. It is possible that an attempt has been made to express the idea of the revolution in the carvings on the pillar. This was suggested, we think, by an incision which had to be made in the sacrificial pillar from top to bottom and that counter-clockwise (so prasavya). x) In this connection we should like to return for a moment to a typical ceremony on the occasion of Krishna's birth-day feast. Krishna is not only worshipped then in his Mother's lap, a representation which shows a striking similarity to that of the Madonna lactans. By the side of this, however, Weber mentions in his essay on the Krishnajanmashtami the following representation: Krishna worshipped as a child on the top of a kalasa. Weber thought (perhaps not wrongly) of a connection with the matsyavatara of Vishnu, while suggesting a magical rite as another possibility.2) It is probable in our opinion that we are to think of Krishna as the divine initiate, floating on the waters, the idea of coming into existence, the birth of the initiate being suggested by the cosmic motion of the pillar (by which also a god is represented; see below p. 216) in the kalasa. From our reasoning it does not follow that the pot was not used for technical purposes. 3) It is always very difficult, if not impossible, to decide in how far one has to do with the exigences of technics, in how far with those of ritual or of aesthetics. On the other hand, however, we do not intend to look upon the ritual aspect as something accidental that is of no great importance to the matter itself. It may be brought forward that a pot is a pot and remains a pot and is not a sea or an underworld, but on close examination this is not nearly so matter-of-fact as it seems, because no justice is then done to the facts. The distinction between the symbolism and the essence of the kalasa may be as dangerous as that between religion and fancy in myth. It is a pity that even Havell has not a decided opinion 1) Kaus. Br. X. 2. Caland. Een I. g. lustratiegebruik p. 290. 2) op. cit. p. 276 n. 1. . , 3) Apart from the pot at the foot of the column, one also gets impression from the description by Dubois, that the pole should be considered standing in the pot. on this subject and frequently appeals to the „purely utilitarian uses" evidently to make his ingenious, but necessarily not always provable explanations more sound. This is most dangerous. For nothing is more obvious now than that the whole of the work of Havell will be looked upon as a description of symbolism in Indian art, with which reality has only an indirect contact. The student may take note of Havell's work, he may even think it important and accept his results — and yet lay it aside to return to „reality". This matter is also of importance for our inquiry, so that we prefer to take sides openly. The conceptions „symbolism, syncretism and magie" should be taboo, to be used only with circumspection. Just because this symbolism is so vague Havell's work is a little unsettled here and there. The pot is not only a symbol of the nether world. It may also be said that the nether world is a kalasa. It is just as if a man should call a fence, by which he separated his own field from that of another, a symbol of his right to that enclosed field and to settle afterwards that it is really a construction of planks of such and such a shape. When we now read, to return to our starting-point, in Macdonell's work that the Vedic cosmogony hesitates between two theories: that the universe is a technical edifice; and that it is the result of natural production, we think of the significance of the house of assembly which represents the whole world, even if that should be the sacral world. „As the Vedic house was built of wood, the material is once or twice spoken of as timber. Thus the poet asks: „What was the wood, what the tree out of which they fashioned heaven and earth?" The answer given to this question in a Brahmana is that Brahma was the wood and tree.... Heaven and earth are very often described as having been supported (skabh or stabh) with posts (skambha or skambhana), but the sky is said to be rafterless..., and that it never falls is a source of wonder The framework of a door is called ata; in such a frame of heaven Indra fixed the air... The doors of the cosmic house are the portals of the east through which the morning light enters... Foundations are sometimes alluded to. Thus Savitr made fast the earth with bands Visnu fixed it with pegs, and Brhaspati supports its ends.... The agents in the construction of the world are either the gods in general or various individual gods; but where special professional skill seemed to be required in details, Tvastr, the divine carpenter, or the deft-handed Rbhus are mentioned. Little is said as to their motive; but as man builds his house to live, so of Visnu at last it isindicated that he measured or stretched out the regions as an abode for man...." 1) Here again the post is spoken of, of which the central post was certainly the most important. That was the Skambha, of which it is asked in A. V. 7- 8.: „Who is the Skambha on whom Prajapati founded all things ?" 2) The building of a house was usually attended by various ceremonies.3) In A. V. III. 12. 6 the vamsa 4) is urged to climb the sthanu (= skambha) and to keep away the enemies. In the painting of the cosmic house not only the supporting post draws our attention, but also the door. The doors which opened into the Vedic place of sacrifice are often called goddesses (devïr dvarah)5). To contrast the extremes: even up to modern times the door is worshipped. Even now a special „door-god" is sometimes known.6) It is not difficult to ascertain why the door should get a special sacral significance. The entrance to the initiation-grounds is the entrance to the body of the initiation-demon itself. His mask, a wide open mouth, is fastened to the front of the house for the festival. The front of this house is conceived as the gate to heaven, the most central point in the separation between heaven and earth. In the Javanese wayang-show there is a separate object, which is always shown and in which Dr. Rassers recognized the reduced representation of the front of the sacred festal-house. Rassers has proved that the Javanese wayang is genetically connected with the dramatic performances z) Whitney makes the following note: „Skambha, lit. „prop, siipport, pillar strangely usedin this hymn as frame of the umverse, or half personified as its soul". 4) ^emas^cJogiMlly^his word is remarkable. According to the Pe*ershure Dictionary it means: bamboo, reed; the joists and laths, lymg on the beams especTally those lying longitudinally and indicaüng the beanngs of the house"?; be, generation, pedigree; flute, spine; upper part of the sword Knowledge of the social milieu will be necessary to understand this complex. M Macdonell-Ved. myth. p. 154, 171- •) Crooke — op. cit. p. 98; Arbman-Rudra p. 9». of the clan-myths in the sacral festal-house. The object we refer to, indicated by the remarkable name of Kayon (grove) is raised to remind the women (the profane world) of the fact that the wayang scene is really carried out in the sacral world. On the Kayon we find among other things the head of a monster over a closed doublé door. The head of the monster on the Kayon is the same as the mask over the entrance of the sacral meeting-hall. To illustrate his theory Rassers proves with suggestive photos and sketches that the Kayon, the reduced representation of the sacral house, has the same outline as the front of the ceremonial house. The outline of the Kayon is a little more liquid. It has the shape of an isosceles triangle with close above the base a typical constriction. But though it will be admitted that the front of a sacral meeting-hall is important yet one does not immediately expect that on the whole it is possible to speak of the shape of the sacral house front in general. At first sight it is somewhat strange when Rassers puts the form of the Kayon beside a men's house of New Guinea. Yet it is possible to do so. There is certainly a difference in the various festal-houses, but the number of types is restricted after all. Besides, in building the houses the builders are bound to a severe symmetry, so that in all the cases in which bamboo is used to build houses the builders have to confine themselves to the comparatively few possibilities which are then left. Rassers lays the photo of a meeting-hall in the neighbourhood of the river KaiserinAugusta (formerly German New Guinea) beside the Kayon.x) The resemblance is indeed striking. The theory of Rassers implies that also elsewhere we may expect „kayon-ish" representations of the sacral meeting-hall, also in India therefore, where bamboo is used in building houses. As for India the test of Rassers's theory has a favourable result. We are alluding to the well-known horseshoe-shaped opening, which is generally known in Indian architecture. „Le motif le plus important de (l')architecture du bois, motif essentiel de toutes les fagades des édifices, est 1'ouverture en fer a cheval formant auvent, placée au-dessus d'une entrée ou d'une fenêtre comme une gigantesque lucarne. Cette lucarne Contr. Geog. and Ethnog. Mag. Neth. East Indies vol 88 p. 408. est souvent, peut-être même toujours, 1'aboutissement d'une voute de même forme soutenue par des poutres dont 1'extrémité carrée se voit, sur la fagade, dans 1'intrados de 1'ouverture en fer a cheval. Avec le temps, cette ouverture en fer è cheval se modifie peu k peu; elle tend h se refermer et k s'orner aux extrémités. En outre, des ouvertures plus petites accompagnent 1'ouverture principale. Sur la fagade des cavernes, dès les plus anciens exemples, se voient de fausses fenêtres avec de fausses lucarnes de toutes tailles. L'art évoluant, ces petites lucarnes se multiplient toujours davantage et leur taille diminue. C'est un motif d'architecture qui, peu a peu, devient un motif de décoration. Les fausses fenêtres, que surmontaient les fausses lucarnes dans les exemples les plus anciens, disparaissent tout d'abord. Soutenues seulement alors par des balustrades et des corniches en gradins, ces fausses lucarnes tendent k se ranger en ligne. Plus tard, balustrades et corniches en gradins disparaissent k leur tour et sont remplacées par des corniches courbes sur lesquelles viennent s'aligner les fausses lucarnes devenues toutes petites: ce sont les Kudu (Ajanta); une tête apparait alors dans ces minuscules lucarnes, mais ne se maintient pas trés longtemps." *) Jouveau-Dubreuil and Havell a.o. have brought this horseshoe-shaped opening into connection with the shape of the roof that has been used throughout India and is still found among the Todas.2) That opening, a representation of the front of a house built of wood, has also held its own in the houses built of stone. We may compare the front of the temple of Chezarla; there we have the exact model of the kayon.3) It cannot be accidental that that horseshoe-shaped opening, the front of a house built of wood has so often been fixed on the temples. Even the mask is not missing: „The makara was carved at the springing of the arch, and a lion's or dragon's mask, with gaping mouth, as if swallowing the whole arch, was carved at the crown." 4) In this we see the mask of the initiation-demon in the front of the sacral house. Itis also unacceptable in this !) L'Inde Antique p. 405. 2) Jouveau-Dubreuil — Archéologie du Sud de 1 Inde p. 20. 8) Ann. Rep. Arch. Dep. South. Circle. Madras 1917— 18 Also 1'Inde antique fig. 20. 4) Havell — op. cit. p. 24. light that the head of the monster in the horseshoe-shaped opening should only be a decorative filling of this now useless opening. x) The unprejudiced testimony of Jouveau-Dubreuil and Havell makes us strongly suspect that the horseshoe-shaped opening is exactly the same thing as the Javanese kayon: the reduced representation of the sacral house for the festival. The form, the frequency and the place of the application and the head of the monster make it almost impossible to make a mistake. Architecture testifies here rather distinctly to the great importante of the initiation. Gradually the image of Krishna is outlined more and more clearly. Krishna, also the threatening god of the initiation, who swallows everything, has not by accident his domicile in Dvaraka, the Town of Doors. There he indeed still lives in a primitive society where his mask has been put up menaaingly over the door of the sacral house. There is still one point on which we want to give the work of Rassers an Indian character so to say. That is namely the question of the enclosure which divides the sacral from the profane world. The door is the central part of the enclosure. It is the entrance through which the initiate comes out of the profane into the sacral world. In this door the idea of sacralization must, according to primitive art, be strongly expressed. In general the sacral character of the temple is indicated by the fence as a whole. The fence, found round sacred places and sacred objects, is deduced from the fences found round the most ancient structures in India, the stüpas. But the fence has a much wider application than the stüpa only. The space at our disposal does not allow of a detailed exposition of what we could but explain in a few words. Anyhow it is clear that there used to be a sacral house in India, representing the whole world, and of whicii we have tried to explain three important parts, the pillar, the door, and the fence. * * * It is not our intention to dweil upon various other interesting points, connected with our argument. Our point at issue is the ') »b. p. 174. Epic. The pillar has been paid special attenrion to, because in the Epic the name Sthanu (pillar) is one of the most familiar epithets of Rudra-Siva. Sthanu is the son of Brahma. The Epic states that Rudra bears the name of „pillar", because properly speaking the pillar is a phallic symbol.1) The pillar is Siva's linga, whicn stands upright in the earth. Rudra-Siva is actually called mahalinga, ürdhvalinga, sthitalinga, etc.2) The name sthanu does not refer to an imaginary, but to a real wooden pillar. Rudra-Siva's body is enveloped by trees (vrikshair avritakayah) 3). It willbe clear that the otherwise senseless name of „pillar" cannot have been given to a god if the word pillar had not a special sacral meaning, which we fancy we can still detect in later stone-structures. Who was this god who, in the shape of a pillar, was the centre of worship? In other words, who was Rudra-Siva? Doubtlessly he was a central figure in the Indian Pantheon. In the Svetasvatara Up. Rudra-Siva is acknowledged as the highest deity. Hence it follows directly that it is not easy to explain the epic Vishnu-Krishna as a local god (ev. complex of local gods) raised to the highest godhead, after whose example Siva was later on ilso promoted to the highest godhead. This is the less probable because this Upanishad is rather older than younger than the Epic. 4) Indeed, already in the Vedic texts Oldenberg detects the tendency to indicate Rudra by large numbers of names, to give him numerous dwelling-places, through which he obtains an everywhere perceivable omnipresence, and his shape is multiplied indefinitely.5) The coming in of the Krishna-Vishnu-creed has been visualized as a transfusion of spontaneous popular religion into an exhausted Brahminism, given to over-subtle and sophistic philosophical speculations. In Rudra-Siva's case this is not so easy, there not being a possibility of an innovation here. The 1) The question whether the phallusworship was Aryan or not, has given rise in our opinion, to rather unprofitable, debates. Whatever one thinks about this, the MBh. is familiar with the phallus-cult. (Rhys Davids-Buddhist India p. 165). Or B. C. Mazumdar — J. R. A. S. 1907 p. 337- „ „ , 2) Hopkins-Ep. Myth. p. 212. 3) MBh. VII. 203.34. *) Bhandarkar-Vaisnavism etc. p. 110. 6) Oldenberg-R. d. V. p. 220. Vedic texts already depict Rudra as a terrifying and formidable god. It is therefore an established fact that in spite of Brahminism, there were in olden times already pure „Aryan" gods, whose character and savageness can hold their own with any other „popular gods" like Krishna, who are, in their origin, said to be entirely or partly outside the pale of Brahminism. It once more appears difficult to consider Sivaism and Vishnuism as sects, arisen as a reaction against Brahminism. RudraSiva may be considered to have possessed a strongly anthropomorphisized popular character from of old. In the Svet. Up. he is just as well an AU-god as Krishna in the Bhagavadgïta, and in the Epic he is a „popular god" again. It is much more plausible simply to assume that no gods have ever existed who had not been anthropomorphisized. We can then simply assume that gods have existed with „popular" characteristics in the times of the Vedas as well as in those of the Brahmanas and those of the Epic. The picture of the relation of the religious-literary tradition becomes somewhat more complicated then, Upanishads and Brahmanas may then be considered a much specialized style of literature. At any rate one has booked the advantage then of not being referred from philosophy to religion and vice versa. In his treatise on Rudra Arbman points out that it is not easy to regard the „popular" characteristics in the figure of Rudra as accidental circumstances of a later date. Rudra is never a complex of all kinds of figures, he is always himself: Rudra. x) The thesis that Rudra is a more and more humanized natural phenomenon is apparently supported by the fact that Rudra in the Rig Veda is more of a „celestial" god than in the later texts. Arbman started from this contrast; but came to the conclusion that the „volkstümliche" Rudra of the later texts must be, not younger, but older than the „celestial" Rudra of the older Rig Veda. He justly remarks that it will not do to make the younger Vedic texts, to which a more and more exclusive character is ascribed, accessible again to the thesis of the „volkstümliche" Rudra. Arbman regards the original Rudra as „eine erdgebundene, damonische, grausige, den *) Just as Krishna did not appear to break up into all kinds of other Krishnas. primitiven Vorstellungen vom Tode und dessen Schrecken entsprungene Gestalt", from whom an unbroken process of growth goes to the later Si va. Side by side with this original Rudra a „celestial" Rudra arises who should be regarded as a specific product of the religious idiosyncrasy of the Rig Vedic literary circles. , In our opinion Arbman is right in endowing the Vedio gods also with antecedents. The differentiation into a „celestial" and a „volkstümliche" Rudra seems less happy to us. Without further proof the thesis that demoniacal traits in a god are more in place among the lower classes, cannot be regarded as proved. It is, for instance, unwarranted to represent Rudra's dwelling in the mountains as more „volkstümlich" than his living in a heaven. Mountains and forests, assigned to him by the people, are not the dwellingplace of the formidable god of storms, less sacred than the celestial dweiling, which the fancy of the Rig Vedic poets has built up for the celestial Rudra. In the mountains and forests lays the place for the sacral tribal meetings, which per definitionem is heaven itself. In old Indian theology heaven and earth are not two locally separated, but rather religiously differentiated worlds. Also the starry sky is this sacral world, but the focus of the sacral world is the place for the great ceremonies, where also the gods themselves dweil. The „heaven" on earth is simply the sacral trysting-place of the whole community, just as the sadas not merely represents Vishnu's body, but really is this body, although Vishnu may at the same time be said to be all. Moreover Arbman has strained the contrast between the celestial Rudra and the „popular" one too much, for in the Rig Veda the „demoniacal" traits are by no means absent. Oldenberg says about Rudra's place in the Rig Veda: „Der Inhalt der rgvedischen Rudralieder ist durchweg die Angst vor den Geschossen des furchtbaren Bogenschützen, die Bitte, dasz er Mensch und Vieh mit Seuchen und Sterben verschone, seine wunderbaren Heilmittel spende".1) And on the other hand the popular god Rudra is by no means wholly demoniacal, for he is repeatedly addressed as the great healer. He is indeed also called Siva, the Propitious One, a name which can hardly, in sober ') R. d. V. p. 2x6. nt. 2. earnest, be said to be explained as a lucus a non lucendo. Roughly speaking we may therefore say, that investigators discover in Rudra a certain ambivalency which, in the Rig Veda, conformably to the peculiar character of this Veda, has been illustrated from a divine point of view, whereas some (a.o. Hillebrandt) think, that in the literature of more recent times the demoniacal aspect of Rudra comes more to the fore. „Hillebrandt (V.M. II. 203) findet dasz die altere Zeit mehr als die spatere Rudra's Heilmittel, dagegen weniger den von ihm ausgehenden Schrecken betont. Mir s~heint dieser Schrecken doch in den rgvedischen Dichtungen durchweg sehr deutlich durch zu blieken. Aber der Stil dieser Poesie brachte es mit sich, dasz eine gewisse Ruhe bewahrt, der sonst gewohnte Ton der Anrufungen auch hier festgehalten wurde." ]) After the Brahmanas, the later literature, in which Rudra is more conceived as a terrifying being, epic literature appears, in which his divinity comes more to the fore. The most important ritual, consecrated to Rudra, is the Trayambakahoma, the sacrifice to Rudra Tryambaka, an epithet sometimes explained as „Rudra-with-the-three-mothers". At a cross-road sacrificial cakes are sacrificed, one for every inmate; moreover one sacrificial cake is buried in the hill of a mole, the animal of Rudra. One walks three times round the fire, to the left, and three times to the right. The remaining cakes are tossed up into the air, are caught again, and hung on a tree. After this the participants say three times: „The archer has unbent his bow." This Trayambakahoma coincides with the sacrifice of the fourth month after the rainy season, when also the great sacrifice to the dead is celebrated. This is of course most important. Hillebrandt says indeed: „Rudra's Wirksamkeit zeigt sich am grössten in den Krankheiten der Regenperiode und des Herbstanfanges. Dies ist die schöpferichste, aber auch die gefahrlichste Jahreszeit." Hillebrandt very properly thinks of the influence of the climate in India where the rainy season is very conducive to epidemics. He sees in Rudra „einen Gott der Schrecken des tropischen Klima's vom Beginn der heiszen Zeit an bis zum Uebergang zum Herbst. Der Herbst ist seine *) Oldenberg-R. d. V. p. 216. nt. 2. Schwester Ambika. Zusammen mit ihr tötet er, wenn er tötet, zusammen mit ihr findet man ihn ab."1) The Trayambakasacrifice is obviously mesnt as a desacralization-ritual. One wants to get rid of the demon of the rainy season. He has now unbent his bow; the period of dangerous action is past. This was the period of the great feasts in memory of the dead and of the general crisis in Nature. One walks three times to the left round the fire, just as one does at the great ceremonies in memory of the dead, and three times to the right.2) One wants to get rid of Rudra. This also appears from the term nir-avada, which is more especially used for sacrifices to Rudra. Oldenberg translates it by „ab- finden", i.e. ihm seinen Anteil geben, damit er sich entferne." 3) The colour of this terrifying god is red, the colour which is very often pre-eminently the taboo-colour. Further strikmg points in his appearance are the bluish-black hairknots, and his blue ihroat. He is adorned with ornaments of gold and he wears a necklace. His clothing consists of the skin of an animal. He is not merely the evil archer, to whom one prays not to injure his worshippers, their parents, children, or horses in his wrath, and of whom even the gods are afraid. He is also the kind-hearted, for he is the father of the world, who arringes all, who makes the rivers flow and who, roaring, moistens all. He is therefore also the kind god, Siva, the healer among the gods, pre-eminently the possessor of healing medicine. 4) In the Epic some more characteristics show more clearly, viz.: his phallic character. The linga is exclusively worshipped through Sivaïsm. His murderousness and bloodthirstiness are further explained in the Epic as cannibalism. During the Asvamedha the blood of the horse is sacrificed to Rudra. The battlefield full of corpses is his playground. He is fond of raw meat and eats flesh, marrow and blood (mamsasonitamajjada). A wreath ï) Ved. Myth. II p. 195, 196 — Oldenberg objects: „lm Zentrum der Rudrakonzeption steht, meine ich, sein schadliches Wirken, nicht die Jahreszeit, in der dies Wirken kulminiert." (R.d.V. p. 223 nt. 3.) In this case, however, we must defend the natural-mythological view, against Oldenberg's. The significance of the rainy season is of course closely bound up with the ekological milieu. 2) Caland — Een I. G. lustratiegebraik p. 298. 3) Oldenberg-R.d.V. p. 217. *) Macdonnell-Ved. Myth. p. 74. of skulls round his neck is very well in keeping with this Siva. In art Siva as well as his spouse Durga, are sometimes represented with a severed head in their hands. Durga lifts the severed head of her antagonist by the hair, and Iets the blood trickle into a cup. x) Undoubtedly Arbman is right when he sees in all kinds of epithets, which speak of Rudra's formidable teeth, his great belly, and his tongue, allusions to this cannibalistic inclination. „Er wird im MBh. als mahavaktra, mahodara, mahamukha, vikrtavaktra, damstrin, mahadamstra, mahadanta, simhadamstra, sutïksnadagana, khadgajihva dargestellt". 2) To this probably also the other epithets will refer which characterize him of old, viz.: cowkiller (goghna) and mankiller (nrighna). Rudra-Siva is an All-god, just like Krishna-Vishnu. He is androgynous. Therefore, according to the Epic, the whole universe belongs to Siva or to his spouse, because the whole is masculine or feminine, pullinga or strïlinga. This does not mean that these two figures cannot be kept apart, for Si va is indeed Vishnu, but yet he is not in the first place srashtri, but samhartri. Vishnu is indeed Siva in being an All-god, but nevertheless he is chiefly srashtri, whereas the samhara belongs to Siva. They both preserve their own character, though they presuppose each other; the same relation which exists among the phratries. One might say: Vishnu is the All-god, viewed from the side of life; Siva the same, but viewed from the side of death. Again the well-known division into two. There is another characteristic in the figure of Rudra which is already very old. His dweiling in mountains and forests. His spouse has many names in which her dwelling in the mountains is referred to. Together with her he lives on Himalaya, where he makes creation go through its various phases, and where he continuously lives ascetically. So he dwells far from men, in the mountains and forests, and in lonely places. There is a rite, named the Sülagava, in which a buil is sacrificed to Rudra to appease him. This rite must also be performed outside the *) Arbman — op cit. p. 270. 3) ib. p. 274- precincts of the village. The remnants of the sacrifice must not be brought home.*) If therefore one looks for Rudra, one should look for a figure which is worshipped in the central pole (sthanu) outside the village, in the solitude of forests and mountains as the redoubtable cannibal of the rainy season. Already in the Vedas he possesses a certain omnipresence, which is once more emphatically explained in the Svet. Up. The drift of our argument will be clear now. Rudra is the central figure of the initiation. He is the initiation-demon himself. Just as he is Vishnu, his complexion is also black, and he is both krishna and asita. But this does not mean that Krishna-Vishnu is identical with Rudra-Siva. Krishna is the divine trickster, who is at the same time the divine initiate. Should one wish to contrast them, one might say that Krishna is the initiate and Rudra the initiation-demon. This is therefore the same phratrycontrast as between Krishna and Kamsa. After having followed our argument so far, our readers may object that we are doing the same thing as the old naturemythology, merely substituting initiation for moonphases and suchlike. But in the first place the association of Krishna, the divine trickster,with the initiation isaphenomenonnotoccurring everywhere else. The divine trickster is a mythical character, which, as such, can exist independently. Secondly Rudra is more than an initiation-demon. Thirdly our exposition shows itself in conformity with reality: that the initiation is preeminently the great tribal ritual. But besides this our readers will allow, we presume, that it is not the initiation-ritual as such which gave us an insight in the society of the gods, but the world-order, with which the classification-system, or the clansystem and the social organization made us acquainted. The gods cannot be explained, it is true, from the classificationsystem, but we can understand them through it. It was the phratry-relation, as forming part of the classification-system, which has so far engaged our interest. Siva is the god of the samhara, Vishnu the god of the srishti. Other phratry-contrasts may be substituted for this. The two phratries cross each other, according to one's starting from the husband or the wife. We !) Bhandarkar-Vaisnavism. etc. p. 105. may also say: the figures of Vishnu and Siva cross each other according to one's starting from the celestial - or from the netherworld point of view.J) As we explained in our chapter on social organization, the marriage-system is a circulating one: i.e. minimally three clans are wanted to render a complete system possible. Well, if one wants to create a phratry-division with this minimum trio, the outcome will be that there is always a group between the two phratries, which, as a whole, belongs to neither of the phratries. We made the acquaintance of Brahma as the supreme god, the creator of world-order and ritual, and of mutual service between the phratries. In our opinion the mutual relation of the three gods in the trimürti can best be represented by a circulating relation, with a phratry-division which we adjoin here. 1 Brahma 2 vishnu 3 Siva Siva and Vishnu represent the two phratries. Brahma is the group of the centre. As our readers wili have noticed, this relation may also be approached from another point of view, by placing Si va or Vishnu in the centre, for without Siva e.g. Brahma cannot exist. Each figure in this trimürti can also act as a universal god, because he comprises the other two. Each figure may be substituted for the others. Still, they retain their own specific significance and function. One will also notice that it was not such a matter of course *) A good thesis of this is Havell's. (Hist. of Aryan Rule p. 30). jn "o O o Saiva Schools > CS C tn- 'c« > in India, that Brahma, once taken up into the trimürti, should entirely recede into the background as deus otiosus. He is master of life and death, father of devas and asuras. Yet Brahma is not the chief actor in practical ritual, as it is the two phratries which maintain cosmic order by means of ritual. The cosmic motion especially concerns Siva and Vishnu, for they are the two phratries. Siva especially is the god of involution, Vishnu of evolution. When Rudra-Siva, the god of the rainy season, is active, Vishnu sleeps. Now one should not say that this schematic representation is no more than a little scheme, devised by us to explain the relation among the gods. It is a little scheme, indeed, but one not entirely arising from the desire for ïllustration, of our fancy. One should remember that the classification-system, which we here represent in nuce, is closely connected with the vision of the world. The peculiar relation existing among the gods ol the Indian pantheon, has, indeed, long ago already, brought Max Müller to his theory of Henotheism, the doctrine, that every god at the moment of invocation, showed all the character- istics of the highest divine being. Doubtlessly Max Mul er overlooked, that 'n the myth, indeed, every god may, seemingly, be taken for all the other gods collectively, but that in the ritua an undoubted difference existed between the various gods. In the ritual Rudra has a different place from Krishna. Yet the phenomenon described by Max Müller, is so evident, that it cannot be dismissed with a few words. Viewed in the classification-system Henotheism is indeed real and plausible After, with an appeal to the authority of Max Muller, having tried for the present to exonerate ourselves from the pretence of bringing forward again, and now in the figure of Rudra, what was a moment ago observed in that of Krishna, we should 1 e to specify a little more in details. Rudra, namely, has a special liking for the smasana (cemetery for burning or burying t e dead). He is smasanavasin, smasanagrihasevaka, smasanabhaj, and likes to dweil on the smasanas. x) Having assumed Rudra-Siva to be the central god of the initiation-sanctuary, we ought to face the question whether any J) Arbman — op. cit. p. 267. parasols. The whole structure is within an enclosure, a stone reproduction of an earlier wooden prototype, which on the four cardinal points is provided with gateways (torana). In the oldest stüpas sculpture is found exclusively on the gates and the enclosure. x) Although historically the stüpa is closely allied to Buddhism and Jainism, yet archeologists assign a prae-Buddhistic origin to it. For it is now generally assumed that the tumulus was the prototype of the stüpa. In our opinion, however, a few objections might still be raised. Apart from the question why the tumulus should rise to such significance precisely in Buddhism and Jainism, we are moreover somewhat perplexed by the difficulty of explaining how exactly a tumulus might pass into a stüpa. Several archeologists, therefore, suggest not a simple barrow becoming higher and higher, but a kind of building which might serve as prototype of the stüpa. Jouveau-Dubreuil suggests a reproduction of the round house of the Aryan chieftain, a hollow „stüpa", built of wood and covered with clay. 2) Havell speaks of a „memorial chapel of the dead Aryan chieftain". 3) In order to support this hypothesis, according to which the stüpa had of old been some sacral building, Havell points out terms such as vedika (enclosure), the same word with wich in Vedic ritual the altar was described.4) We would ask, whether this particular chaitya, the stüpa, does not fit into a similar complex as that described by Heine-Geldern. The stüpa is then still related to a tumulus, but to a special kind of tumulus, in the first place a sepulchral monument. The stüpa is then erected in memory of a deceased worthy. From this point of view it is no longer strange that there should be stüpas that are not actually barrows, but merely monuments in memory of a certain person or a certain event. This second kind need not be considered as a secundary adaptation of the „real" stüpa. It is moreover remarkable that according to che Manasara l) Pischel-Geldern also gives stüpa = vamsa, beam in a house — Ved. Stud. p. 114, 137. *) G. Jouveau-Dubreuil-Vedic Antiquities — 1922. Only known to us from a discussion in Ind. Ant. 1923 p. 370 by S. M. Edwardes. 3) Handb. of lndian art. p. 72. 4) op. cit. p. 15. the sacred places of the Buddhists and Jainists were relegated to grounds outside the village; no doubt this is due to the fact that the stüpa originally played some part in the cult of the dead. Outside the village stood also the temple of Durga and further the temple of Vishnu in his avatara of Man-Lion had its place there; in this shape Vishnu emerged from a pillar to devour the unbelieving Hiranyakasipu. The burial-ground lies North or East, according to some South of the village.*) The four doors which are often found in the stüpa might be interpreted as illustrating the saying that Rudra likes to dweil not only on the smasanas, but also on cross-roads. The enclosure around the stüpa we associate with the enclosure which should fence off the sacral world from the profane. The stüpa can doubtless be best characterized as the „artificial reproduction of a miniature heaven", in which a central column representing the mountain of the gods.2) On the „punch-marked coin , wich will be dealt with more fully later on, one very frequent figure is that of a small semi-circle, very often surmounted by a dog, a peacock or a tree. The marks of this very ancient archeological material have been recognized as representations of the stüpa. No doubt Sir Bhandarkar is right, when he remarks that stüpas with a dog and such-like on it, can hardly be Buddhistic. The tree will be briefly discussed again below. The dog, however, surmounting the chaitya is the more suggestive of the god whom we think we are justified in recognizing in the central column of this structure (or its prototype), as howling dogs with large mouths belong to the attendance of the terrifying god of wood and mountain. Dr. Arbman who collected much interesting material about these strange companions of Rudra, describes the dog as an animal which functions in religion as a special taboo-being. ) The dog has some other peculiarities which render it suitable to function as the particular animal of Rudra-Siva; it is namely not only a terrifying creature, but it is at the same time occasionally regarded as friendly.The dog is Rudra ssatellite, which swallows its prey without masticating. A dog, however, was i) Dutt-Townplanning in anc. India p. 151, 155, 159* ■) K. B. Raja-Quart. Journ. of Myth. Soc. XXI p. 106 — The Buddhist Stüpa. 3) Rudra p. 257 seq. also the incarnation of Dharma in which shape it was the faithful attendant of Yudhishthira on his journey to heaven. Arbman suggests two „Kulturschichten" here. The dog as a terrifying animal answers the ancient notions that lived among the people. When, however, in a later period, in certain circles the gloomy underworld of popular belief was transformed into a heavenly Elysium, the dog became the peaceful doorkeeper. Here again, it is in our opinion practically impossible to divide the two aspects between two „Kulturschichten". The dog has those particular amphibious qualities which have supplied Rudra, the terrible, with the additional name of Siva, the kind. Another animal that is found in the entourage of Rudra is the serpent, the animal that is particularly associated with the rainy season. The serpent in India is, on the one hand the malitious creature of the underworld and the waters, on the other it is possessed of jewels, of the miraculous jewel, that fulfills people's wishes or can restore to life. It possesses an invigatory elixir and affords people wealth.The serpent lives in perpetual enmity with the Garuda, an enmity which goes back as far as a quarrel between their mothers, in which the serpent-mother was victorious, it is true, but only by unfair means. The conflict between the bird and the serpent is a conflict between relations. 2) In the rainy season, according to popular belief, there rages a fight between the bird and the serpent. The rainy season is a period of struggle, struggle in nature between light and darkness (or whatever mythical contrast on may want to substitute for it within the system of classification), struggle among mankind, — the military Dasara-festival testifies to it. The struggle between the bird and the serpent is in many mythologies one of the most striking representations of the rivalry of the two phratries, which will be more fully discussed later on.3) Vishnu, Siva's opponent in common phrarry-relation is, according to Johannson, the sun-bird. The serpent belongs to Siva. The serpent, in particular, has that aspect of totality, which is also a distinguishing feature of Krishna, the divine trickster and of Rudra, the *) Vogel-Indian Serpent Lore. 2) Sesha calls Garuda his brother (bratri) MBh. 1.36.10. 3) See G. W. Locher — The serpent in Kwakiutl mythology. initiation-demon. Not a single divimty, indeed, is entirely without it, but both Krishna and Rudra exhibit it in a very striking shape. The serpent, like Rudra, is a pronouncedly phallic being, but again like Rudra it is at the same time hybrid. *) Moreover the serpent is not only related to the Garuda, its enemy, and also a benefient creature, the bird lends to the serpent as it were its characteristic features. Serpents with two distinct wings insteed of a hood are by no means rare. ) What a remarkable shape the tension during the rainy season, the struggle between the bird and the serpent, has assumed in the ritual, appears from the festival which was celebrated in ancient India during the rainy season, the Indradhvajafestival. Indra's Standard (dhvaja) was set up seven days before the full moon of the month of Asvina, the third month of the rainy season. It consisted of a pole, decorated with flower garlands and little baskets, it was surmounted by Indra in the shape of a Hamsa. The whole was held upright by means of ropes and pulled down on the day of the full moon.3) This is very probabiy the same ceremony about which Dr. Barbara Renz writes: „Die Schlange als das zum Coitus bewegende Prinzip wurde in Indien noch zu Forlongs Zeit durch Bander dargestellt welche die lebendige Schlange vertraten. Wenn angezogen, versetzten diese Bander einen auf hoher Stange entsprechend angebrachten Lingam in Bewegung, und dann fielen alle Umstehenden anbetend auf ihre Kniee. Dieser Lingam hatte die Gestalt einer Flamingo (Henza), der die Sonne (den Sonnengott Vishnu) vertrat Bei der obigen Vorrichtung durften nur junge Madchen und Burschen an den Bandern ziehen. Das Ganze stützte sich auf einen am Boden liegenden Drachen. Aehnliche Vorrichtungen sah Forlong auch in buddhistischen Tempeln." 4) It will be seen, how strangely the values of the various parts may shift without the whole losing its own character. Here we find in the rainy season a bird-phallus, representation of Vishnu, whereas this !) Vogel — op. cit. p. 271. 3) MBh°I.Ö3. J. LMeyer — Das Weib im altindischen Epos-Leipzig 1915 p. 210 nt 3 and the passages given there Yajn. I.149; Agni Pur. . aqo. Me ver sees a fecundity-ceremony in it. i) Der orientalische Schlangendrache — 193° P* ö9- period is pre-eminently that of the serpent, the birthdayfeast of Krishna-Vishnu being likewise celebrated in this period of Rudra-Siva. So serpent and dog, Rudra's animals, both have that peculiar ambiguity which Arbman wants to divide between two „Kulturschichten", a division which in our opinion is entirely unsatisfactor y. This ambiguity is the very characteristic of Rudra, and cannotpossiblybeeliminated without eliminating Rudra himself. Archeologists have assured us that images did not exist in ancient India. If this does not refer to carved wooden poles, sufficiënt phallic poles can be found in ethnographic museums from which one could form an idea of the one by which Rudra was represented.x) Rudra acquired his central position in his capacity of terrifying initiation demon, in which capacity he is identical with Krishna. His cannibalism, perhaps to be viewed in the light of human sacrifices, is at any rate in a mythical sense: the devouring of the initiate. 2) Various guilds whose profession demands a certain secret initiation, as those of robbers and professional rogues and of others whose activities are related to the ritual objects necessary for the tribal feasts (carpenters, cartwrights, makers of bows and arrows, potters, smiths and hunters) acknowledge Rudra as their patron. In epic poetry Siva is therefore called sarvasilpapravartaka. 3) „Dans £iva, a date historique, se concentre et se reflète tout le paganisme, tout le mysticisme, tout 1'ascétisme. H devint le dieu de la science et de la grammaire." 4) Now it also becomes clear why Si va is represented as the great ascetic of the mountains. The initiate during his period of seclusion is subjected to numerous taboos. Moreover Siva is well-known as the great dancer. Here we *) The sacral places for the performance of the ritual not being permanent should not be adduced as proof of the lack of images or sanctuaries. From their very beginning the sacral places have never been permanent (see p. 202); and moreover, value is especially attached to ritual objects in so far as and as long as they are used in the ritual. Desecrated objects are of not much account. (Bath-Oeuvres I p. 65). 2) Is the name Rudra, often explained as „the Howler", to be connected with the threatening voice of the initiation-demon, imitated in Australia by the droning of the bull-roarer? Rudra-Siva sometimes emphatically bears the name of initiate (brahmacharin). See, e.g. MBh. XIII. 140.48. J. Charpentier — W.Z.K.M.G. XXIII p. 154. 3) Hopkins — Ep. Myth. p. 223. *) de la Vallée-Poussin — op. cit. p. 322. think, as in the case of Krishna, of ritual dances which formed part of the dramatic performances of myths. That dancing and dramatic performance are inseparable is seen from the word nartaka which means dan eer as well as actor. The typical representation of éiva as lord of dancing (natesvara) is found in stone, probably not until the sixth century after Christ, but this does not imply that the image of the dancing god had been entirely unknown before. The same image in wood may have been common at a much earlier date, as Havell justly remarks. The dancing &iva stands on the top of a dwarf, as to whose meaning we dare not pronounce an opinion. The Rudra-Siva figure has been explained by different scholars as wind-god, tempest- or thunder-god, death-god, leader of the host of souls (Maruts), wood- or mountain-faun, etc. If we propose to see in Rudra the central figure of the rainy season, the initiation-demon, this does not automatically imply the exclusion of all other hypotheses. Nature is not excluded from the system of classification; and it is quite possible that lunar phase, the struggle between sun and darkness were assimilated into this mythology, as indeed they often were. But the nature-mythologi cal method considers only one (oftentimes, indeed,veryprominent)side of the mythical beings. The connection with really religious life, with myth and rite, however, is more or less neglected. Our interpretation of Rudra is an attempt to elude this one-sidedness. * * * „But", the reader may ask, „what exactly has all this got to do with the MBh? There has been a great deal of talk about Siva who plays such an important part in the Epic that Barth could write: „Dans le Mahabharata, qui pourtant dans sa rédaction actuelle, est plutöt vishnouite, le culte le plus répandu est celui de £iva"; but, what is the sense of bringing in all kinds of intricate archeological problems, when dealing with an Epic in which e.g. the stüpa is hardly ever mentioned?" In reply we state that we do not mean to keep walking round the MBh. without touching it. But it is also partly owing to the nature of the material known to us, that we are occasionally compelled to deal with all kinds of side-issues because the essentials are for the moment not to be attained. When dealing with an existing primitive community one may try to elucidate obscure points by means of closer inspecture of existing conditions, which is indeed frequently done. But this is naturally impossible in the case of the Epic. So it is, in our opinion, entirely in keeping with the method we pursue, if we are sometimes obliged to go beyond the limits prescribed by our epic text. It would for instance be very convenient, if the MBh. stated in detail, how, when, why and where the sabha was erected. For in the Epic the sabha is the sacral building where festivities are held. Unfortunately the amount of informative material concerning the sabha is too scant to conceive a complete picture of it. What we do know about the sabha, however, renders it all but certain that it was the sanctuary which has so far engaged our attention. In the Epic the sabha is a festive building equipped with every pomp and splendour. The ownership of some beautiful sabha entitles a man to consideration and glory. The Pandavas possessed an uncommonly fine sabha built of heavenly material by the heavenly architect Maya.The inauguration of this clubhouse was celebrated with brilliant festivities. The Kauravas too built a magnificient sabha on the occasion of the great gambling-match against the Pandavas. So it becomes obvious at once that the sabha is a richly decorated festive hall, temporarily constructed. We are not surprised that the poet should speak highly of fine colums and splendid gateways. The sabha built for Arjuna had „golden trees" (satakumbhamayadruma), an enclosure and gateways made of jewels (ratnaprakaratorana).*) The sabha of the Kauravas has as many as one thousand columns (sahasrastambha) and one hundred doors (satadvara) and gates like crystal (toranasphatikakhya).2) The sabhas of the gods are described as equally splendid. That of Varuna has blue, black, yellow and red pillars, decorated with garlands, while Brahma's sabha is notable through the entire absence of columns. Brahma's heaven is of course the Universe, and as such his sabha defies every *) MBh. II. 3.22, 26. 2) MBh. II. 56.26. description. A festive building, richly ornamented, whose doors, columns and decorations continually held the poet's attention, a building which at the same time constitutes a residence for the gods, this picture of the sabha is immediately suggestive. The sabha according to the P.W. is: i. Gemeindehaus; 2. Spielhaus; 3. Gesellschaftsraum im Wohnhaus; 4. Asyl; Zuflucht für Reisende; 5. Vorhalle in einem Tempel. Regarding the sabha Dr. Zimmer states: „Die Versammlung der Dorfgemeinde hiesz sabha; weiterhin bezeichnete das Wort auch das Gemeindehaus, wo diese Versammlungen statt fanden, und dann algemein geselliges Local für die Manner, Spielhaus." *) The sabha, therefore, we must regard as a building which could serve many purposes. Into this complex of functions we have also drawn the chaitya, a structure which primarily served as memorial for the dead. The modern temple still plays a much more important place in Indian sociery than the church in the present-day West. Not each temple, of course, is used for exactly the same functions. There are differentiations. Sometimes there are various buildings at the same time in the same place, each for the performance of a different function. Sometimes temporary structures are erected for special purposes. Abbé Dubois has noted that one and the same building may be used for the following functions: guest-house for strangers, village council-house, temple, common meeting-place for the men, dancing-place for the dancing-girls connected with the temple festal ground.2) Also in Mohammedan villages in the North the village place of worship may sometimes serve as a guest-house for strangers, a building which in its turn does service again as the centre of social life of the village. „Here, says Crooke, „the police officer conducts his investigations when a crime is committed; the land-lord collects his rent; the village-council sits to deal with breaches of morality or violation of caste-rules." 3) There even seems to be some connection between the temple and the cult of the dead, which in olden x) A.i. Leben p. 172. „ , „ 11 a) Manners and Customs p. 325, 567» 579/ 584- See also HavellHandbook etc. p. 66. 3) Natives of North.-India p. 154, 155* All these go very well together, for the potlatch is a ceremony attended with grave risks. One should not take the word „risk in a business-sense, but as a religious risk. The risk of the potlatch may be systematized in tourneys and games of chance of some kind or other. So one may expect that in game of chance or in strife, when one comes across them in a community with a system of „societies", one has to do with some form of the potlatch, the ritual exchange-tourney. For games of chance we have the disposal of the circumstan tial essay of Stewart Culin, a rich collection of material with many pictures.1) Culin started from the „divinatory use of the arrow". The basis of this divination appeared again and again to be a classification according to the points of the compass. A simple and at the same time very universal way of gambhng is the one played with the aid of a diagram in the nature of a game of goose, a number of squares pegged out along a circle or a cross. Somehow (by throwing little sticks (chips) or something like that) a number is fixed which is counted out along the line on the squares. The point where one ends determines the results. As the game of chance is a divination-method (presently proofs of the sacral nature of the game will be given) it will be understood that the place where the piece ends may indicate a spot in the classification-system. On the diagram a definite point has indeed been marked, namely on the spot where the i) Chess and Playing Cards. Ann. Rep. Smithsonian Inst. 1896 p. 665—942. piece came out and the piece has consequently been classified in that part of the cosmos, associated with the point of the compass attained. In India as we saw, the northerndirectionofthe wind e.g. bears the same relation to the southern direction as light to darkness, so, like all kinds of other contrasts implicitly associated with a phratry-division x). Together the four quarters of heaven form the basis of a division into four. Now if one marches one's piece through all the points of the compass along a circle, the piece passes through all the clans or castes, associated with the various quarters. On its way along the points in the circle the piece consequently passes through the entire cosmos. Usually a piece is said to be free when it lands at a certain point in the circle, or when it reaches the centre. In order to indicate on the diagram (which should mostly be symmetrical, because it represents the cosmos, with its phratry-divisions, and so often has the shape of a circle or cross or combination of these two), the place and consequently the direction, i. e. the place in the classificationsystem, simple, little wooden chips are employed2). Usually three so that at one throw one has four chances, according as one, two, three, or no chip turns up its convex side. These little chips have a convex and a concave side, or in other words, a male and a female side 3). The two sides of the gamblingchips are often distinguished by contrasting colours (white-black; pink-dark red) or by definite marks (serpent, bird, club). Remember the model of the dual phratry-division. How closely gambling, owing to its potlatch-character, is connected with strife, appears from the dice actually being arrows, or being considered as such 4). These arrows appear, on a closer examination, to be provided with marks, indicating to which clan the owner of the arrow belonged. These cosmic marks were occasionally replaced by the name of the owner. *£hat South is the land of death, was rightly connected by Kern with the fact that the sun goes tothe south in the Pitriyana (Oldenbers — R.d.V. p. 544 nt. 3). 5 2) op. cit. p.681. , ') Among the Zuni Indians: „The canes are split with reference to the notion that one side is masculine the other feminine fib. o. 780 nt. 1; 854). * ' ') Culin considered the entire game of dice as a „divinatory use of the arrow . ïb. p. 679, 780. As in the cosmological system the totem-mark is identical with the member of the clan of that totem, the throwing of the gamblingchip determines the place of one's own clan or one's own club in the cosmos. The dice are consequently of the utmost importance in the game, and possess much more than technical significance, as is the case in modern games of chance. The significance of the serpent in gambling appears again from the resemblance there often is between the gamblingstick and the serpentx). The diagram and the pieces used, are often very simple makeshifts which can eventually very well be dispensed with. Inagreatmanygamesitis considered sufficiënt to shuffle a number of chips, shards or things like these, to divide them and determine the place in this way. The ritual significance of the game should not only be deduced from the significance of gambling-utensils. It appears from other facts as well. Gambling is found side by side with weddingceremonies and sporting-matches. The Wanabaki Indians accompanied it with the serpent-dance 2). Among the Zuni the Sho-li-we-game was played by the esoterie club of the Priestsof-the-Bow. This war game was played half-yearly at the feasts to the „Twain-Gods of War patrons of the game by virtue of the vanquishment of the Creational God of Gambling". The feast was a divination for the tribe, by means of which it was settled whether there was to be prosperity or adversity, peace or war. Elsewhere the serpent is referred to as the „doublé chief";the,,war-chief and the potlatch-chief"3). Like all rituals, games of chance are often said to originate from the gods, or to have been created by a culture hero. There are special gambling-deities 4). To have success one should propitiate the gods 5). The game of dice is also organized against diseases 6). The potlatch, indeed, is a universal compensation when one has suffered moral or material damage. To the ritual several taboo s apply. The participators fast a few days and sing sacred hymns. The plate in which the dice lie, is held by someone possessing a *) ib. p. 704. . 1 u 2) Locher has explained what the serpent signifies in the potlatch. See also p 229. 3) ib. p. 782. Locher — op. cit. p. 87. <■) Culin — op. cit. p. 730; p. 715 nt. 5. 6) ib. p. 705. 6) ib. p. 722. „charm". In visions and dreams the participators are informed beforehand of the issue. The tourney itself is attended by prayers and exorcisms. In the game the dignity of the contending parties is practically the stake. On poles put up for this purpose, great quantities of money are hung, for the potlatch is a demonstration of wealth. One does not play the game in the first place for oneself, but as a representative of one's entire group. We shall now give a circumstantial quotation to enliven the picture of the game of dice. It concerns the game of the Iroquois. „Oftentimes one village plays against another. Each party chooses a marker; but he withdraws when he pleases, which never happens but when his party loses. At every throw, especially if it happens to be decisive, they make great shouts. The players appear like people possessed, and the spectators are not more calm. They aH make 3 thousand contortions, talk to the bones, load the spirits of the adverse party with imprecations, and the whole village echoes with howlings. If all this does not recover their luck, the losers may put off the party to the next day. It costs them only a small treat to the company. Then they prepare to return to the engagement. Each invokes his genius, and throws some tobacco in the fire to his honour. They ask him above all things for lucky dreams. As soon as day appears they go again to play; but if the losers fancy the goods in their cabins made them unlucky the first thing they do is to change them all. The great parties commonly last five or six days and often continue all night. In the meantime, as all the persons present, at least those who are concerned in the game, are in agitation that deprives them of reason, as they quarrel and fight, which never happens among savages but on these occasions and in drunkenness, one mayjudgeif, when they have done playing, they do not want rest"1). In the course of the game the stakes rise continually and are pushed to and fro between the contending parties. Suppose that A puts up a stake called x against B and that he loses. After this A puts up a stake x + 1. If he loses x + 1, he gets back x from B. Now he puts up x + 2 against the stake x + 1 which he has lost, and which B has won. The process repeats itself. *) ib. p. 721. Assuming that A loses all the time, his last stake will be his wife against all the property he has lost. If A loses again, B gets his wife. A gets back all his property won by B, with the obligation to put up all this against his wife whom he has lost. If A loses again, he is ruined. If A wins now, he gets back his wife besides a part of the other property he had already lost to B. With this latter part he must make another effort to win the game. And B has received back all that A had lost to him before, minus the balance which A got back minus his wife. So the system comes to this: each throw renders a counter-throw obligatory. If one loses a stake, the loser must raise the next stake. During the game the tension increases continually and culminates when the game is on the point of taking a decisive turn. „At this stage of the game the excitement is v e r y great. The spectators crowd around and intense fierceness prevails. Few words are exchanged, and no remarks made by those looking on. If the loser be completely ruined, and a desperate man, it is more than likely he will quarrel, and endeavour to repossess himself of some of his property, but they are generally wellmatched in this respect, though bloody struggles are often the consequence. We have known Indians to lose everything, horse, dog, cooking utensils, lodge, wife, even to his wearing apparel, and be obliged to beg an old skin from someone to cover himself, and seek shelter in the lodge of one of his relations. It is, however, considered a mark of manliness to suffer no discomposure to be perceptible on account of the loss, but in most cases we imagine this a restraint forced upon the loser by the character of his adversary. Suicide is never committed on these occasions. His vengeance seeks some other outlet in war expeditions, or some way to acquire property that he may again play and retrieve his losses. There are some who invariably lose and are poor all their lives".1) So the game of dice is an expression of the potlatch, which appears very clearly from the data supplied by Culin. The term „gambling" is however slightly misleading. This gambling is not merely a game of chance, played for profit as such. One plays rather for one's prestige, and by gambling one invites a divine verdict, to confirm the claim one has to a high position. >) ib. p. 753- Gambling is indeed not a lottery in which everyone has a chance. A man of standing, or a representative of an important club would not think of entering into a gambling-duel with the first comer who wanted to have a go. One must be rich and have credit to participate in a gamble of any importance. Moreover it is always more or less a trial of skill, and never purely a game of chance. Also in this respect it appears that one should not think of it as of a smart club with accomodation for playing games of chance. „Indeed, no really aboriginal game is a true game of chance; the invention of that dangerous and delusive plaything was reserved for civilized ingenuity. An expert player will throw the number he desires with almost infailing certainty.... It is a dexterity which any one may acquire by sufficiënt practise, and only thus".1) * # * Before discussing the data of ancient India, we should like to make some remarks about modern times. Firstly we mention the game bearxng the curious name of Pachisi (twentyfive). The requisites are a gambling-cloth, shells, pieces. The gambling-cloth strikes us at once as remarkable. It is a crossshaped cloth, each arm of which is divided into three times eight squares2). Undoubtedly Culinis right in saying: — „The board itself represents the Four Quarters of the World". With regard to the pieces he says: — „The colors of the men agree with those assigned to the seasons of the year and the four quarters of the world to which they correspond, in Asia. When four persons play, the red and green, and black and yellow play partners. This relation is indicated on the men used in the Burmese game,which are painted with the complementary colors, the red men having green tips, and vice versa .This corresponds with the relation assumed to exist between the seasons and the world quarters and their corresponding colors. As each quarter of the world has four quarters, each player in turn has four men.... the men or pieces may be regarded in the cosmical game as actually representing men..."3). And finally the shells. *) ib. p. 764. 2) See p 248. ') ib. p. 854 vlg. Culin is almost certainly right when, in view of their two sides, he thinks of female (concave) and male (convex). He rightly says: — „A feminine significance is widely attributed to the aperture of the cowrieshell". Culin suspects the whole game to be „sacred and divinatory in its origin". On the same board dice are often used consisting of four cubes fixed to a little copper staff, knobshaped, sticking out on two sidesx). Sometimes the four cubes can turn round the staff. Of course only four sides of each cube are visible. Culin mentions Ramala Pasa (Dice for Fortune Telling) and Pasa (Long Dice). The long pasa (Skrt. pasaka?) he considers as „the more or less direct outcome of the divining staves"2). Finally he mentions a little gambling-top, the chukree (Skrt. chakra?)3). Alas about the „entourage" of the game (place, time, sex, occasion, extent, age etc. of the players) in India very little is known to us. We can only give some interesting data from Further India. Nguyen van Huyen tells in his book „Les Chants alternés des gar^ons et des fillesen Annam" (Paris 1934) about the singing-competitions taking place between boys and girls from different villages or belonging to the same village. These literary duels are sometimes organized by the village-authorities, in front of the temple of the village god, for instance, sometimes byrichpeople at births or marriages which purpose: „1'achat d'un certain rang, d'une certaine place sur la natte du temple du génie du village". Consequently potlatch feasts. These feasts serve at the same time to enable the boys and girls to make each other's acquaintance. On these feasts, however subtle and interesting they may be, we shall dweil no longer here. We should only like to say something about the feast which the wealthy people orgamze to enable the boys and girls to make each other's acquaintance, namely in playing chess. In Tonkin the games of chess are played in spring or in autumn 1) ib. p. 824, 835. 2) A peculiarity of many dice is, that the total of the pips of two opposite sides is seven (6 + 1; 2 + 5 etc.). With the Pasa the two and the five are painted red. This is also noticed elsewhere. 3) We shall not go further into the more complicated games like Chaturanga which resembles our game of chess (for the rest also founded on cosmic classification!) (op. cit. p, 857). Various Indian games have been described by Hem Chandra Das-Gupta (J. A. S. Bengal N. S. XX p. 165, XXII p. 143, 211; XXIII p. 291). See further: Man in India II p. 143, 244- in the court of honour („cour d'honneur") of the village temple. The village men of note create a complicated ritual decorum Boys and girls, resp. the red and the black camp, carry sticks' on which the names of the chessmen they represent, are given. The boys and girls are selected by a committee from the most aristocratie families. The feast is a genuine potlatch-feast, bear- ing much outward display of wealth. Moreover it is a ritual game, which can only be played after certain ablutions and much calling on the village-deity. After this game of chess the girls who have been introduced at the feast become marriageable. Just as after the singing-competition, a marriage usuallv follows. Of course this remarkable Further-Indian custom cannot serve as a conclusrve proof for the ritual origin of this Hindustani game. We can only conclude from it that in Further India this game has without any doubt been a ritual potlatch. Nguyen van Huyen points to this himself»). At any rate primitive games, like playing at chess and dicing, appear not to be allround games or games of chance. The entourage is not that of the social evemng or of the gaming-room, but of the sacral world. This applies not only to America, but also to South Eastern Asra. Nor has the significance of gambling in India escaped the hvely interest of a von Schröder. Yet he is of opinion that Culin's assertions are beside the mark as far as India is concerned, because gambling, also ritual gambling, gives »den Eindruck eines Würfelspiels in unserm Sinne.... resp. der sacralen Stilierung eines solchen Spiels" 3).To a certain extent von Schröder is nght. Culin saw especially one side of gambling, the divinatory one, and this simply because in 1896 when he wrote his essay, the study of the potlatch had not yet been started. Gambling as a divination-method was known in India indeed, (only remember the Pasakakevalï) but for the rest, gambling was practised there simply for material gain So von Schröder is right in saying that gambling in India stands ») Nguyen van Huyen op. tit. p. i9, 146. tra.'fc mterestlng t£at the term ma'yong (the game which still shows th£ nam* f nmitlVu classification), is, in the Malayan Peninsula also " ®,e. for masked performances (Ene. Ned. Indïë s.v. Tooneel) a& sKaTST" «"*** "■m w,,h for much more than a divination-method. Only the „sacrale Stilierung" he has not taken seriously enough. We will now endeavour to arrange the data known to us, on Indian dicing, and will open with the die (aksha). Heinrich Lüders, the author of a circumstantial essay on „Das Wüifelspiel im alten Indien" *), sums up under the collective name of aksha: — pasaka (a rectangular prism of 1 X 7 cM); nuts of the Vibhïdaka- (or Vibhïtaka-) tree; shells (kaparda); salaka (little stick, chip) bradhna 2). The Petersburg Dictionary gives two words aksha, viz. I aksha-Würfel zum Spielen, II. aksha-with, a. o. these meanings: 1 — Achse am Wagen; 2 — Name einer Pflanze, Terminalia Bellerica. Die Synonyme kali und vibhïdaka (vibhïtaka) bedeuten gleichfalls Würfel (aksha), da dazu die Nüsse der Terminalia Bellerica gebraucht wurden; 3 — die Nuss der Terminalia Bellerica; 4 — Elaeocarpus Ganitrus; 5 — der Same dieser Pflanze, der zum Rosenkranz gebraucht wird; 6 — Name eines Gewichts, ein Karsha =16 Mashaka; 7 — Schlange; 8 — Garuda 3). Lüders opines, that Vedic ritual only knew Vibhïdaka nuts. This can hardly be correct, seeing that in Sankh. Ar. XII. 5- 3 and A. V. XIV. 1. 35 also gold akshas are referred to 4). The Epic moreover speaks of gold, ivory akshas, and akshas made of gems (vaidürya), black and red akshas. That in the cases where older literature speaks of gold akshas, one would have to think of gold imitations of the Vibhïdaka nuts is hardly probable, because it makes the matter unnecessarily complicated 5). In order to fix the place of the akshas we start from the meaning aksha = karsha. The weight karsha 16 mashas is equal to 1 karsha or pana (the copper karshapana) to 1 purana or dharana (the silver karshapana) 6). So one gets aksha = purana or dharana or karshapana7). Now karshapana, ' !) Abh. Kön. Gesellsch. Göttingen Phil. Hist. klasse N.F. Bd. IX, IQ°)7 Meyer still mentions the aralas, a word which he translates by „Krümmling". Register to Kautilya). crhmi'Ht n .u» 3) Of meanings 7 and 8, there is another example in Schmidt Nachtrage. _ . „ _ „ 4) A. B. Keith — J. R. A. S. 1908, p. 827- «) Andras Kshatrapas-Catalogue of Indian Coins (Introduction P'Sn the words mentioned by Meyer (register to Kautilya) akshapatala (Finanzministerium) and akshasala (Edelmetalschmi ). dharana or purana is the name of the well-known „punchmarked coin" x). Most of these coins are copper2), some are also made of gold or silver, or with a thin layer of these metais. Shape: oblong or square, (later on also?) round 3). On one side all sorts of characters are punched, on the other side nothing at all, or at any rate only a single character. Sometimes they are concave-convex. Investigators are all convinced that they are at any rate praeAlexandrian, without however being able to determine whether they date from 600 or from 1600 bef. Christ. Investigators have wondered long and much at the meaning of the marks on the „punch-marked coins". At one time it was thought that the symbols on the frontside originated with the various moneychangers through whose hands the coin had passed, and that the marks eventually to be found on the backside, referred to the place of origin, or that they had been punched in by different guilds. Afterwards D. B. Spooner observed that it is absolutely out of the question that these marks should have been punched in arbitrarily, for they mostly occur in regular combinations 4). Consequently they cannot but have some definite signification. Theobald has closely examined a number of these symbols and came to the conclusion that they represented a. o. a man, woman, elephant, buil, dog, rhinoceros, goat, hare, peacock, turtle, snake, fish, frog, linga and yoni, solar wheel, trident, stüpa, caduceus, tree, and arrow. Of course not all interpretations have been definitely settled. Spooner found that a.o. the ') Theobald-Notes on some of the symbols of the Punch-marked coins J. A. S. Bengal LIX, 1891 p. 182 nt; E. H. C. Walsh-Indian Punch-marked Coins J. R. A. S. Centen. Suppl. 1923 p. 175 ssq.; V. A. Smith-Catalogue of the coins in the Indian Museum of Calcutta I. p. 131; Bhandarkar-Lectures on ancient Numismatics (Calcutta 1921) p. 93 ssq. 2) Rhys Davids reports, that the kahapana only consisted of copper. There are gold „punch-marked coins". indeed, but these were so flimsy, that they cannot have circulated as coins. (Buddhist India p. 100). This argument is probably unsound. Among the very ancient circulating media Bhandarkar also counts the hiranyapinda and the gopuccha (cow's tail) which leaves it an open question what may have been the meaning of this. 3) See e.g. Rapson-Ancient India p. 151. Pictures of the „punchmarked coin" may be found in several places besides in Rapson loc. cit. and Rhys Davids. 4) D. B. Spooner-A new find of Punch marked Coins-Arch. Survey of India 1905/06 p. 150 ssq. Also Walsh and Bhandarkar. 17 solar wheel, the stüpa and the tree belonged together. The symbol which Theobald took for a stüpa he described as follows: — „The stüpa is represented either by three semicircles, one of which on the other two or with an additional row below, making six semicircles in all. That these are regarded as so many cripts or relic chambers in posse, is proved by those examples in which each division or chamber is seen occupied by a reliquary, shaped like a dice box, or the small Indian drum, called „dagdagi", used by itinerant leaders of bears and monkeys. Besides the triple stüpa, this symbol has many variants"1). Spooner, however, is not quite sure of this stüpa. He inclines more to a representation of a chaitya,front view or in section. Sir Bhandarkar is not quite convinced either, because it is also mounted by a dog or a peacock. We may add that a tree is also found on it. Symbols on Punch-marked Coins (after Theobald) Assuming that these puranas bear some relation to the akshas, it follows that such a stüpa-shaped little building may very well stand for the sabha. The sabha is indeed (we shall revert to this presently) the place where dicing goes on. It is not material whether it is exactly a stüpa or a chaitya. It is little likely to be the Buddhistic, original stüpa. The front view of the sabha, the dog (of Rudra) on top, the wheel, (the solar wheel, the symbol of cosmic motion, the dharmachakra) and further the tree (the wish-tree, the trysting-tree, the potlatch-tree), the language which all these symbols speak, leaves, in our opinion nothing to be desired as to clearness2). *) Theobald — op. cit. p. 49. , . 2) We could not ascertain, alas, what Theobald is thinking ol, when he speaks of a „dice-box" in the stüpa. Foucher (The Begvnning of Buddhist Art 1917 p. 14 ssq.) calls the coins in question, „Buddhist coins". According to him they are originally commemorative medals, in memory of the Buddha. The wheel, the stüpa and the tree may remind us of Buddhism, this explanation is unsatisfactory as far as other characters are concerned. For the rest we refer to Havell's objections to a one-sided interpretation in Buddhistic sense. That the „punch-marked coin" bears some relation to the die Professor Cunningham indeed says in so many words, when he writes: „These punch-marked coins are generally known in South-India by the name of Salaka, which is usually applied to a domino, as well as to the oblong die which is used at the game of Pachisi"1). That the „punch-marked coin" functioned in the game, also others have noticed, but apparently they have paid no further attention to it. Lüders sees in the objects, used on the stüpa of Bharhut, in the gambling-scene, seven „dice or coins" 2). That we have really to do with the „punch-marked coin", we see when we lay by its side the picture on the same stüpa from Anathapindika's donation of Jetavana park, in which occasion he covered the ground with Kahapanas3). These kahapanas look exactly like the pieces in the gambling-scene4). When we take the second element of which the word karshapana consists, we find the stem pan, which means: to exchange, to sell, to play, to bet, dare, and pana derived from it, means: — „what one sells, contract, stake in gambling", and further the „coin" in question. In the dual it means the dowry of either family °). It can hardly be called going too far to recognize the potlatch-sphere in this word. One need only open Davy's book to see how closely the contract is connected with potlatch and marriage6). Let us now once more examine some of these stamps. The dog, the howling companion of Rudra, is certainly in place on a die. The expression Svaghmn (killer of dogs) for the lucky gambler, bears relation to it, in our opinion. Dicing is a ritual, ') „I infer" — Cunningham proceeds — „that this name has been given them, partly on account of their oblong form and partly on account of the marks upon them". (Coins of Ancient India p. ). 2) J. R. A. S. 1898 p. 120 nt. 3. 3) We draw attention to this characteristic potlatch-gesture. 4) Cunningham — The stüpa of Bharhut pl. XV and LXVII: Rhys Davids-Buddhist India. Fig. 11. 23. 5) Dahlmann — Das MBh. als Rechtsbuch p. 255; Meyer — Das Weib p. 47 nt. 2, translates „Kauf und Verkauf des Madchens". Is it correct when Russell (op. cit. II. p. 164, 192) derives the term dowry (pan) from Skt. parna (beteileaf)? In ancient India markets are not directly spoken of. Bouglé says: — ..On devine que dans les fêtes religieuses par exemple, si nombreuses et si fameuses, les pélerins étaient autant d'acheteurs" (Essais sur le Régime des Castes p. 240). For the connection between market and potlatch we refer to the word aoana (market). • ') La Foi Jurée. Paris 1922. in which one practically fights the powers of the underworld. One „kills" the pieces of one's adversary, as the dicing-jargon of the Red Indians often expresses it. Nala says to his brother Pushkara, when at the return match he regains all he had lost: — „It is not you who have performed the action by which I was conquered before, but Kali. And you, fooi, have no notion of it"1). Na tvaya tat kritam karma yenaham vijitah pura Kalina tat kritam karma tvamcha müdha na budhyase Another animal on the „punch-marked coin", the serpent, can also very well be accounted for. We have pointed out that serpents are possessors of wealth and of mana-objects, e.g. of jewels, which fulfil all wishes, or which bring life. The Indian serpent, the hostile relative of the Garuda, apparantly occupies a similar place in Indian societies as Dr. Locher has stated of the snake among the Kwakiutl. He describes the snake as the dual being of the potlatch and of strife. One is immediately reminded of this on seeing the picture of the serpent, with the dog one of the satellites of Rudra, on the „punch-marked coin". Indirectly the relation between this coin and the aksha is confirmed once more, because aksha itself also seems to mean „snake", and further, as if to render the simile quite complete, also „Garuda". The duality of the aksha can hardly be expressed more clearly, for the bird and the serpent are the typical representatives of the two phratries2). There exist statues of serpents, which may be explained in the spirit of our conceptions. Professor Vogel describes a group of Nagas, discovered in the Mathura district, the oldest and bestpreserved specimen of which, the Naga of Chhargaon, holds up its right hand, as if to strike, whilst the left hand holds a vase. Now there are other statues which Vogel connects with these, viz. the figure of the chakravartin. „Here the meaning of the gesture is evident. The chakravartin, as Dr. Burgess has rightly observed, causes a shower of money (square coins!) to 1) MBh. III. 78. 22. , 2) The difference between aksha and aksha is apparently not great, because the Vibhïdakanut (aksha) is a die (aksha). Is the word aksha (axle) perhaps founded on the meaning serpent? Rudra is also the aksha of the chariot (MBh. XIII. 17. 121). In another place the serpent sesha is the aksha of the chariot (MBh. VII. 203. 96). Nala saw the akshas as birds — MBh. III. 61. 12). lace" is wrong. Visvarüpa is omniform, not multiform. If one thinks again of: „bearer of all clan- or club-emblems" no one but Rudra, the initiation-god, is entitled so much to wear such a visvarüpanishka. The second word to which we beg to draw attention is akshamala. We suspect the translation „rosary" to be misleading. The seed of the Elaeocarpus Ganitrus the Sivaite calls Rudraksha. A person possessing a lucky akshamala has Lakshmï (luck). Is the akshamala, ascribed to Siva, to be derived from the visvarüpanishka, which Rudra wore? *). We will now return to the die, after having stated that it is said of the aksha, that he who throws them ekarüpa, wins the game. This is also said in dicing with shells. The winner is he who throws them ekarüpa (with the same side head or tail)2). If one examines the „punch-marked coins" it strikes one at once that there has been a tendency to distinguish the two sides. The rüpas at any rate stand for the greater part on one side. Sometimesthe coins are moreover concave — convex. We suppose that a distinction into male and female is implied, which, as Culin asserts, is usually made on the akshas, two sides which are easily found on the shell as well as on salakas and the bradhnas 3). One gets the impression that also the epic akshas had only two sides. The akshas of Yudhishthira feil paranmukhah (literally, with averted face)4). It is interesting that a modern Pandit also prepared the Vibhïdakanuts for dicing by distinguishing two sides. On one side he wrote pa, i. e. Pandavas, and on the other kau, i. e. Kauravas. He next used the nut as a little top. Without saying that the Vedic Indians did likewise with the Vibhïdaka- Krickeberg has called „Renomiergeld". The coins served more as a kind of medaj1 than as objective Standard of value. Not in the first place the measurable quantity of the metal used, but the part which the object has played in the ritual, determines its value, which thesis is strengthened by the fact that the „punch-marked coins" sometimes had a coating of another metal than that of which the core consists; and that, moreover they do not possess an exact Standard-weight. *) Hastings Encycl. s. v. Rosary. Hopkins-Ep. Myth. p. 219 nt. 1. Mala or necklace of gold or silver coins or corals (Russell II p. 12). The woiks of Leumann, which Hopkins mentions, are not at our disposal. 2) Lüders — op. cit. p. 29. 3) The pasaka and the „chukree" have four sides. Perhaps this is an extension or a new form. (These square objects were, apparently, still suggestive of a classification, for cube-shaped dice are, as far as we know of still more recent date). 4) MBh. V. 2. 12. nuts, the Pandit can, in our opinion, hardly have given ear to the dictates of a whim. Dice frequently have two sides, which are to be taken in the spirit of the well-known phratry-contrast, an identical contrast as between Pandavas and Kauravas, as will be demonstrated 1). We shall return for a moment to the relation between the „punch-marked coin" and the castes and guilds. As we stated, the castes comport themselves partly like clans, from which indeed they have sprung. The coins depicted and described by Rapson as „Cast Coin" and „Guild token" are „punchmarked coins", which have still preserved their character of emblems there2). When these coins are losing their character of emblems and are desacralized, they gradually begin to resemble money as a neutral Standard of value, as we know it. Of course one can never teil from the coins found, if they had already been desacralized, and were consequently no longer used as emblems. We are convinced that the „punch-marked coins" have been akshas, and that these akshas were emblems of a club or clan. This does not prove of course that the specimens found were really used as akshas in the potlatch-ritual. The die has the doublé function of technical object to fix a place in the classification-system, and of emblem. Culin pointed out that especially the arrow can fill this doublé function. One should reckon with the possibility that the aksha = karshapana was a specialization of the significance of the aksha as an emblem, just as the die proper, with pips on it, is a specialization of the technical function of the aksha. We are not quite sure of the real use of the karshapana as a die either, even though the dicing-scene on the Stüpa of Bharhut depicts these coins apparently. If,however,one views the data collected by us in their mutual relation, we consider it proved that the old akshas were emblems with which the gambler staked his dignity, in a deeper sense, himself, so that as far as this point is concerned, the ritual significance of dicing may be considered as settled. * * * 1) Possibly the unintelligible place in the i5th book of the A. V. in which is spoken of the blue-black of Rudra's belly with which he covers the bhratrivya and the red with which he hits his adversary alludes to the two sides of a die. 2) Rapson — Ancient India pl. II; p. 151* Now in dicing one could throw various ayas. These ayas (lit. course) are krita, treta, dvapara, kali, abhibhü, aksharaja and askanda. Kali is one, dvapara is two, treta three and krita four x). These ayas are classified according to the points of the compass in the way given in the annexed sketch2). For the sake of comparison we have added two more sketches (Iland III), which are entirely different and by which is indicated who during the Purushamedha were sacrificed and to whom they ought to be sacrificed3). From these sketches one sees that these ayas only stand for places in the classification-system. This place is determined by the throwing of the dice; so far it is immaterial whether one takes aya as a definite throw, or rather as a certain side of the die. The point at issue is to fix a number, and after that the place in the classification-system. In this *) According to P. W. Kali, dvapara, treta and krita are names of dice, ha ving respectively 1, 2, 3 and 4 pips. The other terms (abhibhü, aksharaja and askanda) cannot be explained in this case. The P. W. calls them „a certain kind of die". Nflakantha, when thinking of aya, takes it as the four sides of the akshas, identical with Samkara's commentary on Chhand. Up. IV. I. 4. Lüders understands the word aya as „throw". *) In Sat. Br. V. 4. 4. 6. Kali is called abhibhü. 2) In diagrams Hand III the name of the man sacrificed is uppermost. Undermost is the name of the being to whom he is sacrificed. 3) What an important place dicing must have occupied, is seen again from the fact that the persons sacrificed had all something to do with the game of dice, though it is not quite clear what their names actually meant. Is the name dvapara perhaps connected with dvar (door) ? The main door is expected in the west, because then one stands in a favourable position when entering (with one's face to the east). Seep. 139. respect the throw is identical with the side of the die and the place in the classification. Lüders has greatly undervalued this datum. He apparently thought that this classification signified little and he is surprised that Sayana should also group the four world periods with the four ayas (kali, dvapara, treta and krita). Yet Sayana's classification in the system is very well conceivable. The indication of the world periods by names from the game of dice is a further and very strong argument in favour of the ritual significance which this game must have had. This same classification one also finds in the description of the game of dice belonging to the ritual of the Agnyadheya, according to Baudhayana. „Zuvor werden unter den Requisiten eine fette Kuh und neun und vierzig Würfel herbei geschafft und wird der Spielboden fertig gemacht. Nachdem ein Teil des Opfers an die „Vater" verrichtet worden", „setzt man sich", so heiszt es weiter, „zu vieren um die Würfel hin, n.1. der Vater mit seinen Söhnen: der Vater vorne (östl.), der alteste Sohn rechts (südl.), der zweite Sohn hinten (westl.), der jüngste Sohn nördlich. Der Vater nimmt zwölf Würfel, dadurch gewinnt er; zwölf nimmt der alteste Sohn, dadurch gewinnt er; zwölf nimmt der zweite Sohn, dadurch gewinnt er; die übrigen (dreizehn, also) schieben sie dem jüngsten Sohne zu. Wenn nun (nur) zwei Söhne da sind, so nimmt der Vater zweimal: wenn (nur) einer, so kommt die Gattin als dritte hinzu; wenn gar keine da sind, so nehmen Mann und Weib zweimal. Dieselbe Regel (des Spiels) gilt, wenn sie zu dreien und wenn sie zu zweien spielen. Jeder sagt nachdem er die Würfel genommen hat (also nicht der Jüngste, dem ja die Würfel zugeschoben werden): „gemacht" (krtam) und mit den Worten „die Kuh ist ver- spielt", stehen sie alle auf" Der ganze Vorgang wird im Apastamba Sütra durch die Worte krtam yajamano vijinati: „der Opferherr gewinnt das Krta", angedeutet"*). Howexactly the game was played technically we could not gather from the data known to us. Caland thinks differently from Lüders. Caland visualizes the game thus: — a certain number of nuts are lying on the adhidevana (spot where the dicing is going on). Now each of the players takes some (grihnati), he makes a 1) W. Caland — Ueber das rituelle Sütra des Baudhayana. Abh. Kunde d. Morgenlandes. XII/1903. p. 17. „glaha", and on this the game depends. Lüders visualizes it as follows: — one party puts up a stake and throws a number of dice, and the opposing party at once throws down so many dice that the total becomes krita, i. e. divisible by four. Lüders claims that it is emphatically stated that the dice are thrown (nivapati, prakshipati), that they roll, roll down and fly up. Caland refers to the term glaha, which is doubtlessly derived from grihnati (to take), and to much-used terms like udbhinatti, pracchinatti and vichinoti (kritam vichinoti). With these last terms it is expressly stated that one should arrivé at the ayas by means of division, and separation. Bothagree that it all depended on quick and accurate counting. This follows in the very first place from the great number of dice used (numbers of 49, 100, 400 etc. are mentioned); and further from the fact that king Rituparna, who knows „the heart of the akshas" (akshahridayajna), explains to Nala, that he is such an excellent counter; of which, indeed, he gives astonishing proofs. However this may have been, so much is clear that the dice were cast (in great numbers) and that they were then divided conformably to the ayas by pushing one die to each of the points of the compass so long till there was none left. The term kritam vichinoti may indeed mean „to throw kritam", as Lüders will have it, but it is an elliptical expression for „to finish at the segregation, the separation of the akshas to the four points of the compass, at krita". This may be inferred, in our opinion, from expressions like: — akshan nyupya (one throws the akshas) vyühya (divides them to the four points of the compass) samühya (brings them together again) prithayitva (makes (the result) known)1). The term glaha, which causes some difficulty, is conceived by Caland as „the handful of dice, which one takes up", by Lüders as „the dice which one holds in readiness in one's hand, for the throw". The latter assumption is, however, less probable, because in that case one would not speak of a „kaliglaha", as, indeed, one cannot know if the dice one holds in one's hand, will produce kali2). The P. W. translates glaha by „dicebox". It must, however, also mean „stake". One can *) Baudh; Bharadvaja; Hiranyakesin. Caland Z.D.M.G. LXII. p. 123. Zur Exegese und Krifik der rituellen Sütras. 2) Caland — Z.D.M.G. LXII. p. I28. win by means of a glaha. In a fight the champion is called the glaha of the army. Is it possible that the matter should be explained thus : that a number of akshas were thrown down, and that those akshas which feil down on a definite side, were taken into account1). The dice which were counted, were called „glaha"; those not counted „seshana". As early as the A. V. (n, 114,5) these two words are found side by side. In the A. V. they are given as a gift in dicing, by the gods. The unintelligible use of seshana would thus be explained at the same time2). The semasiological connection between glaha, „number of akshas which count after the throw", and „stake", is easy to discern, when one remembers the fact that originally the aksha was not merely a technical instrument for the calculation of chances, but also the emblem staked 3). Glaha = champion is founded on the identity of emblem and gambler. In one's akshas one stakes oneself. We shall revert to the relation with strife, further on. For the time being, the technical side of the old game can only be guessed at. We consider it a certainty, however, that the aya refers to the course which the aksha takes through the cosmos, according to a classification, conformable to the points of the compass. Aya, in fact, also means „course"4). Sayana, indeed, defines „ayaih" with „the various names of the different courses the akshas take, namely, krita, etc." (akshagatasamkhyaviseshaih kritadisabdavachyaih), and from what he further writes (ekadayah panchasamkhyanta akshavisesha ayah), there appear to be five ayas5). The transition of meaning from aya (going to a place) into aya (throw) is not great. In the Pasakakevalï aya is used in the sense of „throw". Of a word aya (going) !) In using pasakas, the side turned towards the ground, decides (Weber-Ind. St-eifen I. p. 280). 2) Lüders — op. cit. p. 48. 3) In the potlatch the result most decidedly depends upon the stake. Lüders, who considers this out of the question, thinks too much of a Western game of chance. To the players it was a matter of course that the highest emblem should also win. (Lüders — op. cit. p. 27). 4) Among the qualities of the gambler Sakuni also mentions, that he knows the samkhya (yo vetti samkhyam). The P. W. understands this word as „Berechnung, d. h. genaue Erwagung des pro und contra . Luders understands it as „das Zahlen". The most plausible explanation is that samkhya is used here in the direct sense of „classification-system. Lüders — op. cit. p. 58. 6) Lüders — op. cit. p. 39. one can also understand, that in the Rig Veda the gods are said to go about like the ayas (aya iva pari charanti) and that the Apsaras danced about with the ayas (yayaih parinrityati) *). More important than all this is the circling motion which the akshas make in their course along the ayas. When one starts with Kali in the North and then goes round to the East via the West and the South, (this indeed is the sequence 1,2, 3, 4), one moves against the sun, so, apradakshina. It is most improbable, after all that has been said about the rotatory motion in India, that this motion should be accidental. When indeed, by the terms krita, etc., not the ayas, but the world periods are meant, one does not go to the left, but to the right. First comes the Kritayuga, then Treta, Dvapara and Kali, and after Kali, Krita follows again; so it will be seen that here also the rotatory motion is kept up, otherwise Krita could not succeed Kali again; but it will also be observed that here one does count to the right2). The motion the akshas make is the motion of death, the motion of the netherworld, a direct connection perhaps with the motion of the serpent. This peculiar trait is still present in the modern Pachisi-game, for also there one counts along the crossshaped board from left to right, against the sun, and returns to the starting-point 3). We do not know whether for this motion of the akshas the word nivritti is employed 4), but it is beyond any doubt that *) Lüders — ib. p. 50. 2) MBh. III. igo. See p. 139. • '1 Culin °P* c't* P- 852 — Contrasting aya with anaya, aya is right and anaya wrong. We add here what Weber writes about the commentary of Patanjali on Panini V. 2. 9: „Die Angabe des Bhashya f. 33a. b. aya-'nayam neya ityucyate, tatra na jnayate: kah ayah? kah anaya. ïü, ayah pradakshinam anayah prasavyam, pradakshinaprasavyagaminamparanam yasmin paraih padanam asamavepah so'ya-'nayah aya-'nayam neya aya'-nayïnah carah, und die Erklarung Kaiyata's ,z.u:.P_a r a l r iti dvitiyadyütakarasambandhibhih ?araih; padanatn sthananam grihaparaparyayanam anakramanam anadhyasanam. sasahayasya parasya parair na-'kramyate padam asahayas tu parena paraklyena badhyata iti dyütavyavaharah, lassen keinen Zweifel darüber, dass es sich hier (evens0 wie bei Bartrihari 3, 43) einfach um unser Trictrac handelt". Why Weber rejects the explanation for aya pradakshina and prasavya for anaya in a note, we fail to understand. (Ind. Stu'd. XIII p. ai2 and nt.). r *) Lüders mentions indeed the word nirvrit, applying to the akshas. He translates by: „die Würfel aus der Hand rollen lassen" op. cit. n. 6s nt. 1. for the game of dice as a whole the word pravrit is used 1). The translation „to start a game of dice" is too weak, for when we are told that Duhsasana drags Draupadï to the sabha, when she had been gambled away by Yudhishthira, it says: dyüte pravartite 2). Elsewhere we read that the Pandavas return to fight a fresh gambling-contest: punardyütapravrittaye 3). This same term pra-vrit is also employed when a fight is referred to. Nala says to his brother Pushkara: „If you refuse to dice, you must fight (Na ched vanchhasi dyütam tvam yuddhadyütam pravartatam). And elsewhere pravartata mahayuddham ). Just like aksha the arrow may also be compared to a bird (often the hamsa) with (golden) feathers. In the same manner there may be a semantic connection between sara (piece used in the game, pawn), sara (arrow) and Sarva (god, killing with arrows; Rudra-Siva). So arrow, snake and die belong together, not only metaphorically, but really in the ritual, because the potlatch is both a conflict and an exchange-tourney in which the snake occupies the central position. P.anadurodara, pranadyüta, pranadyütabhidevana are common terms for „conflict . The conflict itself was indeed more or less a ritual. The conflict is atmatyaga (surrender of self) and atmayajna (sacrifice of self i. e. to Rudra). If no strife had existed in the sacral world between Vishnu and the Asuras, there would have been no Brahmins, and consequently no dharma. Therefore conflict is the highest dharma for the Kshatriya 5). Duryodhana describes the fight in which he is going to engage with Yudhishthira thus: „I, o my father, and Karna, will consecrate ourselves and make a war-sacrifice, in which Yudhishthira is the sacrificial animal, the chariot the altar, the little sacrificial spoon the sword, the great sacrificial spoon the club, the mace the sadas, the Chaturhotra the splinterbar, the arrows the sacred grass, and glory our libation. So, by sacrificing ourselves we shall honour Yama in battle. I and Karna and^rny brother Duhsasana, we three shall slay the Pandavas in battle ). i) See e.g. MBh. II. 58. 25; 73- 20. III. 78. 6. а) MBh. V. 73. 18. 3) MBh. II. 75- 8. 4) MBh. III. 70. 8. MBh. VI. 45.1; VII. 153- 4* 5) MBh. XII. 64. ssq. б) MBh. V. 58, 12. Further in MBh. V. 141. 12. Ahamcha tata Karnascha ranayajnam vitatya vai Yudhishthiram pasum kntva dïkshitau Bharatarshabha 13. Ratho vedï sruvah khadgo gada sruk kavacham sadah Chaturhotramcha dhurya me sara darbha havir yasah 14. Atmayajnena nripate ishtva Vaivasvatam rane Vijitya cha sameshyavo hatamitrau sriya vritau 15. Ahamcha tata Karnascha bhrata Duhsasanascha me Ete vayam hanishyamah Pandavan samare trayah. „It was even regarded not alone as a dïsgrace, but as a sin, for a warrior not to die (as we should say) in his boots" *). The fight is Kshatriya-duty. Death on the field of battle is joy. Before the battle began, a ceremony was performed. The warrior goes into battle „playing and dancing" (kridanniva nrityan). Just as the conflict was a ritual, the ritual is a combat. The soma-sacrifice is a kind of match, and perhaps the competitionelement was not lacking in the other sacrifices either2). The game of dice is, together with another form of tourney, which was already very popular in Vedic times, charioteering and fighting, mentioned as early as the Rig-Veda 3). The relation between ritual and dicing is evident in numerous details. That akshas represented more than our dice, one can infer from this exorcism of A.V. V. 31. 6: — „The sorcery, exercised in the sabha, that in the dicingplace and that with the dice, that sorcery I withhold". Yam te chakruh sabhayam yam chakrur adhidevane Aksheshu krityam yam chakruh punah prati harami tam. Lüders points out that the adhidevana, the place where one cast the dice, was dug out in the ground, exactly like the vedi at the sacrifice, therefore. And not merely this—the digging of this excavation was to be done with the sphya, the wooden sword, used in sacrificing4). To the adjurations from the A. V. (IV. 28; V. 189; VII. 50), in which the dice are addressed and are prayed for luck, according to Kaus. S. 41. 10—13, also belonged the custom to lay the dice for some time in honey 2\ ums XIII. p. 187, 193,198 nt. 2) Hillebrandt — V.M. I. p. ng, <13. 3) Lüders. op. cit. p. 19. *) Lüders. op. cit. p. 12. sour milk, in water and sand 1). So they are ritually worshipped, sacral objects. Nor is this the only connection between game of dice and strife, for the word dyüta itself has both meanings 2). „On perd son blason au potlatch; on le perd aussi a la guerre. L émblème, le totem, sont la source d'un important pouvoir dont chaque groupement s'efforce de retenir le monopole, mais dont les autres s'efforcent de se saisir. Cette propriété religieuse est celle k laquelle on tient le plus. Les plus forts cherchent k s'en accumuler le plus possible. De \k des luttes au cours desquelles le blason se détache de son clan pour se fixer, k titre temporaire ou durable, dans des clans différents, même de la phratrie k laquelle il ne ressortissait pas primitivement" 3). The dyüta displays the same trait by which the tribal ritual is characterized, viz. that of brotherly cooperation, with a hardly controlled rivalry. The game of dice, the immediate cause of the whole conflict of the MBh., is called a suhriddyütam, a gamblingcontest of friends. Nala speaks to his brother of a yuddhadyütam. Indeed, the semasiology of various words bears testimony to this. There is for instance the word bradhna, which Lüders explains as „Lederstreife" 4). According to P. W. bradhna is however „rötlichgelb oder falb; Sonne; Spitze (oder ein anderer Teil) des Pfeils". Leaving the meaning „sun" out of consideration, we should like to observe that an arrowhead or an other part of the arrow might act very well as a die and that in any ethnographical museum partly red-coloured arrows are to be found. A strip of leather may, if necessary, be used instead of an arrow(-head), but the meaning arrow(-head) is appropriate. When a parallel is, indeed, frequently drawn between the game of dice and a conflict, or between conflict and a game of dice, one should not think of a more or less succesful comparison, but of the doublé meaning of the word dyüta5). Sakuni indeed says. n Henry — La magie dans 1'Inde Antique p. 113 ssq. 2) Unwarrented is what the P. W. writes about dyüta: — „Haufig uneig.vomKampfe, dessen Ausgang wie beim Wurfelspiel ungewiss ïst. The dyüta is a real fight. 3) M. M(auss) — Ann. Sociol. XI. p. 115. 5) This simiie has been elaborated in MBh. VII. 130. 17. See also VI. 114. 44 and VIII. 87.31. „Know, that bows to me are glahas, and arrows dice, o Bharata. Know, that the heart of the dice is to me the bowstring and a chariot the asphura"1). Glahan dhanümshi me viddhi saran akshamscha Bharata Akshanam hridayam me jyam ratham viddhi mamasphuram. Sahadeva says to Sakuni: „What you take for dice, you fooi, spoiler of Gandhara's glory, are no dice, but you have made them into sharp arrows in the battle". Akshan yan manyase müdha Gandharanam yasohara Naite'ksha nisita banas tvayaite samare vritah. The Indian arrow may equally well be used as an emblem in dicing as the North-American Indian one, for the naman of the owner is given on it2). In the chapter on classification we said so much about it, that we may consider it proved that at this time the name was more than a mere demand of the Registration Service. We may further assume that it was not an infantile amusement to put one's naman on an arrow, but that the arrow was an emblem. Tertium comparationis of arrow and die is, in our opinion the snake. The dice are arrows resembling poison- ous snakes (akshah sara bhavishyanti darunasïvisho- pamah) 3). The comparison of the arrow to a snake is very common. The bow and arrow of Arjuna are properly speaking two formidable nagas, keepers of the Amrita on the mountain of Siva, Mandara. Siva teaches Arjuna the formula (mantra) which goes with these weapons4). The Pinaka, the bow or club of Rudra-Siva is a male snake with seven heads 5). The significance of the dyüta as a ritual is irrefutably proved in the representation of Siva seated on a high mountain, dicing with his spouse. Even a god like Indra finds that Siva does not care to be disturbed in this occupation, enraptured as he is by the akshas (akshair subhrisam pramattah 6). „Die Welt als Würfelspiel eines Gottes und Qiva mit seiner Gemahlin als Spieier sind öfters wiederkehrende indische Gedanken"7). After all that has been said, it will be clear that one should not ') MBh. II. 56. 3. II ^Bh. VII. 99. 8; 138. 5. Dahlmann — Das Mahabh. I. p. 188. 3) MBh. VII. 122, 7. 4) MBh. VII. 81. 6) MBh. XIII. 14. 253. s) MBh. I. 153, 9; 197. ') ]• J- Meyer — Das Weib in a. i. Epos. p. 84 nt. 2. 18 think of all this in the spirit of a game as a pastime, but a ritual, due to a god. Worldorder is maintained by means of the ritual. The gods assist in this by sacrifices. Nothing more natural but that Rudra-Siva should contribute his part by means of a heavenly potlatch-ceremony. It ought to be taken in this sense when the six ritus (periods into which the year is divided) are said to be dicing together x). One of the most interesting proofs in favour of our hypothesis is the word used for dicing itself, i. e. div. According to P. W. this word means „strahlen; schleudern; würfeln, wetten auf; spielen; scherzen; tandein; Jmnd. zum Besten haben; loben; wetten; sich freuen; trunken sein; schlafen". The root div is contained indyut (light); Dyaur (Jove);dyotate (toshine) dyüta (game of dice); further in devana (dicing); deva (god) a.s.o.„Die Grundbedeuting ist vielleicht hervorschiessen lassen, werfen, schleudern"» In our opinion it is not necessary to accept this original meaning for this semantism. At any rate nothing points to it that in the Indian language the meaning „dicing" in the sense of „celebrating the potlatch-ritual", ought to be explained from a metaphorical usage. The word means „dicing, holding the potlatch". Now knowing the potlatch-ritual to be a definite side, and the most obvious side at that, of the triberitual, later on made absolute by the club, the close relation between ritual and social world will at any rate have to be taken into account, and the acceptance of a primary natural mythological substratum is not necessary. One should undoubtedly also start from the meaning of the potlatch as a warritual to explain why daêva in the Iranian religion possesses exactly the opposite meaning of the Sanskrit deva. The word div speaks a very clear language which cannot be misunder- stood2). . In the preceding chapter we discussed the place where dicing took place. This was the sabha. This having been settled long ago, we need not dweil on it at greater length. Now that however we propose to see in the dyüta an important ritual, the question arises, if for the game of dice the sabha was also fenced off from the profane — that is in particul, the world of 2j See for the root div L. von Schröder-Arische Religion I. p. 227 sc. women. If our explanation of the dyüta is correct, we may reasonably expect a fence round the place where the ritual was performed, so, round the sabha, not to have been wanting, (no more than round the place of sacrifice, which was also fenced off with mats)x). This is indeed the case with the sabha. One reads for instance in the Sabha-parvan, that Sakuni starts the game by addressing Yudhishthira as follows: „Upastïrna sabha rajan sarve tvapi kritakshanah". What does this mean: „upastïrna sabha"? With Nïlakantha one might hold the view that the meaning is: the sabha is covered with a gambling-cloth (upastïrna: asphurakhyena akshapatanavasasa upari stïrna). The objection raised against this by Lüders, namely that there is no proof of a gambling cloth being known in epic times, should not weigh too heavily. Words like asphura and astara may rather be counted against Lüders' contention than in its favour. The translation which Lüders himself gives: „Die Spielhalle ist (mit Teppichen zum Sitzen) belegt", is not very satisfactory. It is hardly plausible that at this important moment Sakuni just before the beginning of this grand dyüta, whilst all are seated, should begin with such a silly remark 2). Nïlakantha's translation does not satisfy us either. And this not only because the expression „upastïrna sabha" may be called somewhat curious, if one means to say: „the dicing-cloth is spread", or because then the sentence runs a little less smoothly: „The dicing-cloth is spread and all are full of expectation. Let us cast the dice now and hold a dicingcompetition, Yudhishthira". Upastïrna sabha rajan sarve tvapi kritakshanah Akshan uptva devanasya samayo'stu Yudhishthira 3). The expression occurs a few more times: Vidura invites Yudhishthira to the repeated game of dice in the following words: Upastïrna sabha rajannakshan uptva Yuddhisththira Ehi Panda va divyeti pita tvaheti Bharata 4). And in another place: S. B. E. XXVI p. 128 nt. L 2) Roy translates: „O king, the assembly is full". 3) MBh. II. 58. 2. *) MBh. II. 76. 2. 69. Duryodhanascha Karnascha Sakunischapi Saubalah 70. Duhsasanascha kitavo hate Bhïshme kim abruvan Yachchharïrair upastïrnam naravaranavajinam 71. Sarasaktimahakhadgatomaraksham mahabhayam Pravisan kitava mandah sabham yuddhavisaradam 72. Pranadyüte pratibhaye ke'dlvyanta nararshabhah Ke jïyante jitastatra kritalakshmya nipatitah 1). From this latter passage it is obvious in our opinion, that the word upastïrna cannot refer to the sabha being covered with a dicing-cloth. There is another important passage. Kaus. S. XI describes the following „charm for obtaining villages", as Bloomfield calls it: 7. Etani gramasampadani. 8. Vikara sthünamülavatakshanani sabhanam upastaranani. 9. Graminebhyo' nnam. xo. Suram surapebhyah2). Reading the eighth line, it appears that this expression sabhanam upastaranani does not at all make us think of dicing- cloths or carpets, but of an enclosure of branches and treetrunks. According to the Petersburg Dictionary upastri is: „etwas umlegen, bedecken, umkleiden mit". So it is obvious that the expression before us ought literally to be translated: „the sabha is covered". When, with these words Sakuni begins the dyüta, this has a profound meaning. With these words he wants to convey to his adversary that the sabha it quite covered up, so that the ritual preparations for the game have been made. With these words we once more state emphatically that the dyüta is a rite in the sabha. That, in this respect, the sabha resembles the stüpa, also holds good, without any doubt, for the miracle-sabha of the Pandavas, which is mahamanisilapattabaddhaparyantavedika 3). Entirely in accordance with our hypothesis of the meaning of the sabha and the purpose of a possible fence, the Epic states emphatically that the sabha was not accessible to women. That Draupadï was dragged into the sabha, is disgraceful, for, she exclaims: „The ancient people would not have taken a lawful wife to the sabha, tradition has it". (Dharmyam striyam sabham pürve na nayantiti nah srutam 4). This idea will probably have underlain the precept that before the game a dicing-circle 1) MBh. VI. 14. 69 ssq. *) J.A.O.S. XIV. 3) MBh. II. 3. 33. 4) MBh. II. 68. 9. Vide also: Vedic Index: s.v. sabha. (dyütamandala) x) was to be made, which was not to be broken before the gamblingdebt had been paid. If a die should happen to get outside the dyütamandala, the game was hereby cut short2). Once it is agreed that dyüta was a rite, one cannot but expect the time in which the dyüta was held, to coincide with the period of the great tribal feasts. For modern times the data known to us fit in quite well with our hypothesis. Gambling goes on by preference also at the Dïvalï feast in the month of Karttika: „Gambling goes on at the time as a magical means of trying one s luck for the coming year, and winning something is lucky, so at the family game it is arranged that the head of the house shall be a winner So matters stood already in the eleventh century, at the time of Alberuni 3). We do not know ,however,if the old-fashioned dyüta too, was played in the rainy season. We may assume that the battle did take place about this time 4). We fancy at any rate that we may infer this from the astrological bad omen which, according to Vyasa preceded the great battle. The death of Bhïshma occurred in the pitriyana 5). The place where dicing went on, was, as we have said, the sabha 6). A keen gambler is given the characteristic name of sabhasthanu, a name founded perhaps on the memory of Siva Sthanu, another indication of the correctness of our hypothesis, that the central column was taken anthromorphologically, or rather theomorphologicallyj otherwise the comparison with the pole of a sabha would not have been so obvious. * * * ') In MBh. II. 79. 32. synonymous with sabha. We are reminded of the sarvatbhadramandala, in which Krishna's birthday-feast was celebrated. See p. 206. 2) Lüders — op. cit. p. 12. ■ Crooke Rel. and Folkl. of. North-India. p. 346: Culin — op. cit. p. 822. y 4^J?0nÜeir£at disaPP°inting is, that theEpic not only recommends the month of Margasirsha as suited for the marching out of armies, but also the month of Chaitra (MBh. XII. 100. 9) Hopkins-T.A.O.S. XIII. P* 91. #) MBh. V. 2. •) About the men's house of the Karo- and Toba Bataks (balei) we read in Schurtz (Altersklassen und Mannerbünde p. 261): „Westenberg sah z. B. einen Pfahl mit phallischem Schnitzwerk in der Mitte eines .Balei, an den zugleich die batakkischen Gesetze angeschrieben waren. In den Boden des Versammlungsraum s ist meist ein Schachbrett eingeschnitten, das fleissig benutzt wird (spacing by me. H.). in the South-Batak country the men's house bears the striking name of „sopo or „sapu", which resembles the Skrt. sabha. We are firmly convinced that the game of dice has something to do with Rudra-Siva. We have already stated that Si va is accustomed to gamble with his spouse. We are not surprised that it is exactly Siva who is fond of dicing (akshapriya) x). We are in a position to give a few other slight indications, e. g. that at the Rajasüya wood for the altar of Rudra had to be fetched from the house of the akshavapa (königlicher Spielhalter); that the Bower Manuscript contains a namaskara to Nandirudresvara, the Acharyas, ïsvara, Manibhadra, all Yakshas, all Devas, Siva, Shashthi, Prajapati, Rudra, Vaisravana and the Maruts2); and that in the Akshasükta (R. V. X. 34.1) the intoxication of the dice is compared to the Soma of Mount Müjavat, i. e. the mountain of Rudra 3). Kapardin is a by-name of Rudra-Siva, referring to the manner of doing one's hair. Kaparda is a little shell, used as coin and as a die. Rudra-Siva himself is not the gambling-demon himself. Rudra-Si va is especially the initiation-god. The special function of gamblingdemon was assigned to Kali. This does not mean that Rudra-Siva had consequently nothing to do with gambling. The facts teil us differently. We might say that Kali is classified with Siva, when one thinks of Vishnu or with Rudra when one thinks of Siva or with the spouse of Siva (Kali or Durga) when thinking of male versus female. In the Indian Pantheon it is at no time very easy to take the gods separately. They certainly have an individuality of their own but this only appears clearly when they are placed side !) MBh, XII. 285. 47? I-197- 15; Hopkins — Ep. Myth. p.219 nt. 7. 2) Lüders — op. cit. p. 23 nt, 4. . 3) In the preceding chapter it was pointed out hat Krishna is also to a certain extent identical with the initiation demon. Now it is remarkable that in the Rig Veda-hymns, where a Krishna is introduced, „höchst specielle Bezugnahme auf das Würfelspiel, devana" is noticable, „so dass hienach wohl etwa die Vermuthung, dass wir es hier mit dem Sohn einer devakï, Spielerin, zu thun haben erlaubt sein könnte. In a note Weber further says about Krishna's mother: — „Dass namhch Devaki eigentlich so auf zu fassen ist, also etymologisch mit deva Gott nichts zu thun hat, erscheint grammatisch unausweishch. Es ist fem. zu devaka (div.), wie nartakï zu nartaka. Irgend welches Gewicht ist daraut freilich nicht zu legen," he concludes. Ind. Stud. XIII. p. 316. We are not so sure of this however. The etymological relation with deva^is ot course firmly established. Barth translates Devaki with „la Joueuse and „la Divine" and thinks of Vivsarüpa. Barth agrees with Weber, that the name contains a pun. Barth-Oeuvres I p. 156. Hillebrandt V. M. II. p. 87; I. p. 63 nt. 1. by side in a phratry-relation, by preference. Rudra is also the god of war 1). The battlefield is his playground, he is senanï, as the Epic calls it2). Yet there is a separate wargod Karttikeya or Skanda 3) who is called a son of Rudra-Siva, though also Vishnu is described as a redoubtable warrior. Fighting indeed is a good, a pravritti-rite, a victory of light over darkness, wherefore Vishnu is the conquering god of strife. Before the epic game of dicebegins,immediatelyafterhisRajasüya Yudhishthira has a dream in which Rudra appears to him, with his face turned to the south, a bad omen of the coming downfall of the Kshatriyas 4). In the battle Rudra himself marches before Arjuna, slaying all who come in his way 5). Because Arjuna propitiated Siva in the battle, he is invincible 6). And how curious the relation between the gods is, one sees when Krishna urges Arjuna, before the battle, to pronounce a Durgastotra, after which Siva's spouse appears and proclaims Arjuna's invincibility 7). Sectarianism is absolutely wanting here. Hopkins discusses Kali in the paragraph: „The gods collectively as Fate". Kali, personified, is the evil world-period in which humanity is now, in the hopeful expectation of the dawn of the new Kritayuga. Kali and Kala, time or fate, can, no more than their names ,be kept asunder. „Kali catches careless kings . (XII,12,29).... Kali is evil fate as synonym of alaksml: „Laksmi came to the gods; alaksml to the Asuras", and with alaksmï enters Kali and destroys them, „pervaded by alaksml and smitten by Kali (3, 94, 9 f). So Kali is plain destruction: „In war there is ever kali and lives are lost" (5, 72, 49). Yet the conception is not that of a permanent being but rather of a personified destruction, liable to spring into existence on occasion: „When virtue is destroyed, Kali is born", and Kali thus appears incorporate in all destructive kings, as Dhautamülaka was the Kali of the Chinese (5, 74, 12 f.); Duryodhana was 1) Barth-Oeuvres I. p. 147. His voice is heard in the war-drum, which, "L ■ V- V- 2I- 3 is said to be visvagotrya („belonging to all families"Whitney). 2) MBh. VII. 203. 34. 3) This word is certainly related with the month of Karttika. 4) MBh. II. 46. 6) MBh. VII. 203. ') MBh. VII. 124. 55. ') MBh. VI. 23. a part of Kali (11, 8, 30); and Subhadra „was born as Kali and for the destruction of the Vrsnis" (S x, 245» 19)- Vidula says to her pusillanimous boy: „You have been born my Kali" (5, 133, 30). Kali is the middle one of the triad „virtue, gain, and desire"; gaining is destructive strife (5, 124, 35). From destruction to bad luck is but a step; hence „they say that Kali (bad luck) is in broken vessels" (bhinnabhande kalim prahuh; all broken beds and vessels are aprasastani, inauspicious,asare cocks and dogs and trees in a house, 13, 127, 16). In 13, 23, 4, a kalipürvam is a gift of food obtained by strife, and, like anything leaped over or licked or kalahina, is impure and taken by fiends. As the sun represents time (Kala), Kali, like the sun, is called sarvamalasraya (3, 3, 20) in his part as agent of all ills. As the fiend of dice, Kali is then playing only one part of his general character" 1). We see that Kali is the middle one of the trivarga dharmaartha-kama; Kali is therefore artha (gain, profit, wealth). The one piece of this trivarga (dharma) roots in the sacral world; the other piece was based by the Indians, judges of human character as they were, on desire, love (kama). Artha is more than the commercial, real property of man. One might say: artha is the quotiënt of a man's darma and his kama. For a harmonious life the principal thing is an exact balance of these three factors. Kama, now praised more highly than artha and dharma, now again condemned as the reason of the samsara ,is the „son of dharma and faith2). On dharma, observance of divine order, rests both arha and kama 3). Artha should not be taken as the word for hoarded property in a static sense, but as the word for productive capital in a dynamic sense. In this connection one should not think of a purely economical dynamic force, but of the influence of divine order. Kali is the prestige, the capital one stakes in gambling, in the potlatch; therefore Kali may be „plain destruction , artha and the gamblingdemon at the same time; Kali is the personified risk of the potlatch and is therefore also present in war. This curious complex of motives for economie activity ï) Hopkins — Ep. Myth. p. 76. 2) Hopkins — ib. p. 164. 3) MBh. V. 115. 38. Marcel Mauss has described in his „Essai sur le don" 1). In this essay he has studied the making of presents in various primitive cultures and has come to most remarkable conclusions. In general the present appeared to be based on the dharma, as it would be called in India. The present creates a tie between giver and donee. The Maori (New Zealand) speak of an „esprit de la chose donnée qui n'est pas inerte"2). Thus also was the state of things in old India. Oldenberg says of the „stark hervorgehobene Pflicht des Gebens": „Es besteht ein förmlicher Tarif für den Lohn, den die Zauberwirking des Gebens dem Geber bringt.... Die personifizierte „Speise" sagt: — „Ich bin die Speise. Den der die Speise speist, verspeise ich" (T. B. II, 8, 8, 1): d. h. ich vernichte den, der selbst iszt, was er andern—gemeint ist, Priestern—geben sollte. Es ist bekannt, wie im Lauf der Zeit die Empfehlungen des Gebens sich zur Verherrlichung geradezu wahnwitziger Ausschweifungen einer schrankenlosen Gebelust gesteigert haben"3). Oldenberg does not exaggerate here. To the danadharma the greater part of the thirteenth book of the MBh. is devoted. The hero Karna is sung in the lamentation for the dead as one who always had the word „giving" on his lips, and who never said: „I have nothing to give" (dadanïtyeva yo 'vochanna nastïtyartihibhih) 4). There was ,however, also a dangerous side to this making of gifts: the giver had the duty to give, and the donee the duty to receive, but the donee was in duty bound to make countergifts to the giver. He was therefore bound by the present. Thus it was impossible to Salya to help the Pandavas, however gladly he would have done so, for he was bound by the presents of the Kauravas. „And", Salya adds philosophically, „man is the slave of artha, but artha is no man's slave"5). Arthasya purusho daso dasas tvartho na kasyachit Iti satyam vacho rajam baddho 'smyarthena Kauravaih. Through the present one is bound, both to give and to accept. Who receives, acknowledges his inferiority, unless he gives in ') Année Sociol. N.S. I. p. 30—x86. *) op. cit. p. 47. 3) Weltansch. der Brahm. t. p. 215. 4) MBh. VII. 94. 43. See further e.g. VII. 67 and II. 33. ') MBh. VI. 43, 77. return more than he received. So it is also with the dyüta: it is an imperative obligation for the Kshatriya, a duel of exchanging, which may not be declined at risk of loss of honour. „How could a king like myself who cares for his name and traditions withdraw after having been challenged", asks Yudishthira1). Katham vai madvidho raja svadharmam anupalayan Ahüto vinivarteta. And when the impetuous Bhima is about to punish Yudhishthira severely for the latter's continual losses, Arjuna quiets him with these words: „The king, indeed, has received a challenge from others and has diced, as a knight should. This brings us much glory"2). Ahüto hi parai raja kshatram vratam anusmaran Dïvyate parakamena tannah kïrtikaram mahat. A challenge had always to be accepted. When Drona grants a wish to Duryodhana, the latter asks for Yudhisthira. He wishes to challenge him anew and then oblige him to undergo a second period of exile. He therefore counts upon Yudhishthira's accepting a fresh challenge at any rate. Every potlatch-party obliges to a return-match. Nala compels his brother Pushkara, who has won all his possessions from him, to a return-match with the words: „When one has won another s property and has taken his wealth and kingdom, one is held to give a counterstake. That is a law of the highest significance 3). Jitva parasvam ahritya rajyam va yadiva vasu Pratipanah pradatavyah paramo dharma uchyate. Of the giving of presents in Indian culture Mauss writes: „Le don est donc k la fois ce qu'il faut faire, ce qu il faut recevoir et ce qui est cependant dangereux k prendre. C'est que la chose donnée elle-même forme un lien bilatéral et irrevocable, surtout quand c'est un don de nourriture. Le donataire dépend de la colère du donateur, et même chacun dépend de 1'autre. Aussi ne doit-on pas manger chez son ennemi. Toutessortesde précautions archaiques sont prises. Les codes et 1 épopée s'étendent, comme savent s'étendre les litérateurs hindous, sur ce thème que dons, donateurs, choses données, sont termes k *■) MBh. II. 75, 19. 2) MBh. II. 67, 9. 3) MBh. III. 78, 7. considérer relativement, avec précisions et scrupules, de fagon qu'il n'y ait aucune faute dans la fagon de donner et de recevoir. Tout est d'étiquette; ce n'est pas comme au marché oü, objectivement, pour un prix, on prend une chose. Rien n'est indifférent. Contrats, alliances, transmissions de biens, liens, créés par ces biens transmis entre personnages donnants et recevants, cette moralité économique tient compte de tout eet ensemble. La nature et 1'intention des contractants, la nature de la chose donnée sont individibles" 1). Now the most privileged object of the Indian rage for giving is the Brahmin. Kings, who give great feasts, consider it a duty of honour to treat great numbers of Brahmins to food. The „dïyatam", „give", heard from the lips of the Brahmins, is however a dangerous word for them, because with it they bind themselves to the givers. So long as they are able to make counter-presents, there is no danger for them. This was the case so long as the tribe was a close unit, in which the priestclans looked after the sacral-, and the warrior-clans after the social side of the great tribal ritual, after the manner of the organization familiar to the Omaha-Indians2). The priests and warriors are thus committed to and dependent on one another through the readiness to assist one another, required for the ritual. As soon, however, as the clan-organization begins to decay, the highest castes and clubs begin to isolate themselves. The warrior-clans produce clubs of nobles, who consider themselves the community in nuce. Then things become dangerous for the Brahmin. Will he be able to preserve his prestige of god upon earth, or will he have to become a wage-earning priest, dependent on, and bound to the goodwill and liberality of the richest, whose ritual feasts he has to grace with his venerableness, for his upkeep? Only a deep sense of its own dignity, based on the solidarity of its own caste, can save the Brahminorder assuch. No wonder that the caste of Brahmins was much less fond of Buddhism, which did not consider caste-membership of paramount importance, than the Kshatriyas, especially the kings, who in their wealth possessed the actual power, and 1) op. cit. p. 151. s) See p. 115. who could by the caste-dignity of others only be hampered in their aspiring to hegemony1). In old India the kings very nearly attained success in this ambition. At the Rajasüya, the royal consecration, the king is seated on a throne, above the vedi2). Part of the ceremony was a game of dice between a Brahmin, a Rajanya, a Vaisya and a Südra; or, according to another report, between the king and a Brahmin, a Kshatriya and a Vaisya 3). The king won. Then he obtained five akshas ,by which he became lord of the five quarters of the heavens. He is now addressed with the title of address of God and of Brahmin, and receives the sacrificial sword 4). In this ceremony, which, according to Oldenberg, in its full extent goes only back to earlier Vedic times, the position of the Brahmin has become precarious. Once the king is consecrated lord of the five quarters of heaven, the priest calls him Brahmin, indeed, by which the same divine authority is due to the king as to the Brahmin, but nevertheless the design is very transparent, namely ,to indicate that in the potlatch the king is to be considered the superior of all castes 5). This is the meaning of the presentation of the five potlatch-emblems. According to Sat. Br. XIV. 4. 2. 23—27 there was at first only Brahman. But this, by itself, did not develop. Therefore it produced Kshatra. Therefore there is nothing higher than Kshatra; therefore the Brahmin occupies a lower position at the Rajasüya than the Kshatriya. But still the Brahman is the origin (yoni) of the 1) Masson-Oursel — L'Inde antique. p. 108. 2) The hole in the ground, acting as altar. According to Johannson, representation of the Terra Mater and of the yoni of the earth. Ueber die a. i. Göttin Dhisana u. Verwandtes p. 31, 55- The female vedi embraces the male Agni. „Thereby a copulation productive of offspring is obtained" (S.B. I. 2. 5- *5) S. B. E. XII. p. 63. 3) Caland — Z. D. M. G. LXII. p. 125- . , . 4) Lüders quite rightly denies that a game of dice with five dice is referred to here. „Meines Erachtens ist die Uebe<-gabe der fünf Wurfel, wie auch der Begleitspruch deutlich verrat, lediglich eine symbolische Handlung; die Fünfzahl ist mit Rücksicht auf die fünf Himmelsgegenden, die der König beherrschen soll, gewahlt. Sie kehrt auch sonst m diesem Zusammenhang wieder; so redet zum Beispiel der König den Priester fünfmal „o Brahman" an, was von jenem fünfmal erwiedert wird". Lüders —op. cit. p. 53. See also Keith J. R.A. S. 1908 p. 825. 6) Is there any connection between the King being beaten by tne Brahmin and the fact that the Brahmin is adandya? Weber suspects something of the kind. In our opinion it should not be considered as „a beating off of evil". (Oldenberg — R. d. V. p. 490). Kshatra, and though a king may climb ever so high, in the end he turns again to the Brahman (upanisrayati)1). This development forces the Brahmin to be careful in accepting presents. Mauss says about the Brahmin doctrine on the acceptance of presents: „Toute cette theorie est même assez comique. Cette caste entière, qui vit de dons, prétend les refuser. Puis elle transige et accepte ceux qui ont été offerts spontanément. Puis elle dresse de longues listes des gens de qui, des circonstances oü, et des choses qu'on peut accepter, jusqu'è admettre tous en cas de famine, k condition, il est vrai, de légères expiations. C'est que le lien que le don établit entre le donateur et le donataire est trop fort pour les deux. Comme dans tous les systèmes que nous avons étudiés précédemment, et même encore plus, 1'un est trop lié a 1'autre. Le donataire se met dans la dépendance du donateur. C'est pourquoi le brahmane ne doit pas accepter et encore moins solliciter du roi. Divinite parmi les divinités, il est supérieur au roi et dérogerait s'il faisait autre chose que prendre. Et d'autre part du cöté du roi, la fa^on de donnerimporte autant que ce qu'ildonne"2). In battle the part the Brahmins took was apparently to render the action of the opposing party ineffectual by fitting ceremonies and well-adapted formulae. The Kshatriyas are heroes with their arms, the Brahmins with their words (Bahubhih kshatriyah sura vagbhih sura dvijatayah)3). In the convention which the Kauravas and the Pandavas entered into, to stipulate the fighting rules (dharman samsthapayamasur yuddhanam), it was, a. o. also determined, that words were to be fought with words (vacha yuddhe pravrittanam vacha va pratiyodhanam)4). This will probably refer to the Brahmins, for elsewhere it says: „Purohitas, consummate exorcists and Brahmarshis, rich in knowledge, killed the enemy with words, by reciting whispered prayers, exorcisms and by powerful herbs and pronounced benedictions everywhere (for the benefit of their party)"5). ') Weber — Ind. Stud. X. p. 9. 2) loc. cit. p. 150. 3) MBh. VII. 158. 23. 4) MBh. VI. 1. 2. 8. 5) MBh. VI. 22. 7. Purohitah satruvadham vadanto Brahmarshisiddhah srutavanta enam Japyaischa mantraischa mahaushadhïbhih Samantatah svastyayanam bruvantah. In this cooperation of the Brahmins in battle there was of course nothing incongruous, because the battle itself was more or less of a ceremony; which does not mean that no real fight ing was done *). The effect of exorcisms etc. in primitive civilizations should not be underrated, for it has been put on record that they actually sometimes produced the effect aimed at. It is not always clear in the Epic, whether the poet means to say that a Brahmin fought with natural or supernatural weapons. The ritual character of the strife was so clear to him that he did not very clearly distinguish the action of the Brahmin with his ceremonies from the actual force of arms. When fferce fighting is referred to, it is not nearly always excluded that one must merely think of ceremonial fights with potlatch- character. In Lüders' study we find the following curious passage: ,,Ich bin überzeugt, dass auch an vielen andern Stellen des Rgveda, wo die früheren Uebersetzer an Krieg und Schlachten denken, in Wahrheit von weniger blutigen Kampfen die Rede ist, von Rennen und Würfelspiel und Wettgesangen"2). A Brahmin was forbidden to fight in battle with the arms of a Kshatriya 3). Who did this might be killed. (Yo hi brahmanyam utsrijya kshatradharmarato dvijah / Sa vadhyah kshatralokasya). Therefore the Brahmin Drona could not be defeated, for one of i) Already in the Rig Veda the priest goes into battle, side by side with the royal general. Brihaspati fights side by side with Indra; „die in der Ordnung der Dinge wurzelnde Zauberkraft des Wortes wirkt wie ein Pfeil in Brhaspatis Hand Man vergleiche Stellen wie Atharva Veda V. 18, 8. 9, wo die geistliche Zaubermacht des Brahmanen nut der Ausrüsting des Bogenschützen verglichen wird: des Brahmanen Zunge ist eine Sehne, seine Stimme der Hals einer Pfeilspitze; er führt scharfe Pfeile; sein Schusz ist nicht vergeblich; mit der Macht seiner Askese und seinem Grimm stürmt er dem Feind nach und trifft ïhn aus der Ferne. Oldenberg — R. d. V. p. 66. z) Lüders — op. cit. p, 47 nt. 3. Riehtlv Hopkins observes that this does not mean that all Brahmins adhered to it. Hopkins - J. A. O. S. XIII. . 183.'One should however not forget, that the poet described the fight of the Brahmins with supernatural weapons in the same way as that of the Kshatriyas with natural ones. the fighting rules was: „kings fight with kings (rajna raja yoddhavyah):). The gods themselves had to interfere to point out Drona's unseemly behaviour to him. Also Bhïma reproaches him sharply for making use of Brahmastra against the Kshatriyas from lust of gain, and for having thus violated the ahimsadoctrine of the Brahmins. Drona then lays down his arms, yet more dejected now by an ambiguous remark of the Pandavas, from which he could gather that his son had been killed. The Pandavas do not stop at deceit; Drona is killed. This Brahminmurder, which Arjuna had in vain tried to stop, causes great consternation, and a rout. Apparently the Brahmin did not always strictly adhere to the rules of his caste. Another proof of this is the story of the Brahmin Asvatthaman. Asvatthaman, firmly resolved to avenge the death of his father, acknowledges that every caste has been created for the occupation suitable to it. He considers, however, that he has become a Kshatriya, and resolves to stick to this in order to be able to avenge his father Drona's death. He then journeys in the night to the sleeping camp of the Pandavas. At the entrance he meets with a huge creature, which appears to be invulnerable to all his weapons. This creature, which fills all the sky, is clothed in tigerskins, with bracelets made up of snakes, with terrible jaws and a thousand eyes, and is no one but Krishna. Asvatthaman descends from his chariot and praises the supreme god as Rudra-Siva and offers himself as a sacrifice (atmayajna). The time of the Panchalas has come. Krishna's protection can no longer avail them. Rudra enters into Asvatthaman 2). The nocturnal slaughter begins. In the long run things are getting rather dangerous for the Brahmins taking part in battle. Not so much because the kings are becoming more sceptical to the power of their ceremonies, as because of the visible wealth of the king. At the Rajasüya, as we have already stated, he must acknowledge the king to come first in the potlatch, though he retained the right to address the king in this position of upper-lord as Brahmin. But it becomes precarious when the king, the „kshatriya ') MBh. VII. 160. 38. 21 Rudra is the actual agent in the battle, just like Kali in the dicingconfhct. presently leave, only to return when the societies as such shall have decayed. Another point which, again indicated by Hillebrandt, leads to the same results, is the significance of liquor, suia x). Soma and sura are related like light and darkness. We found sura in the company of the army, in the neighbourhood of the celestial Vratya Rudra. That the army and the sabha with its dicing, are part of the society, has been explained in the foregoing passages2). In the passage of the Ath. Veda to which we refer, also sura is mentioned in their company. This is moreover not the only time that sura is mentioned in the same breath with the dice. The bridegroom, for instance, begs for his bride the charm which lies in sura, in the dice and in the cows 3). If in this connection, one is not to think of a more sacred tertium comparationis, this Indian bridegroom cannot be suspected of too great courteousness. In Rig V. VII. 86. 6 and VIII. 2. 12. man confesses his sin, which is infatuation, sura, passion, dicing and being blind. The same relation also between dicing and strong liquors, in A. V. VI. 70. 1. Sura and soma are both used with the sacrifice, though a certain difference in appreciation of these two is discernible. The R. V. already hesitates between approval and disapproval of this drink. Gradually dislike of sura becomes more clearly discernible. With the spread of Brahminism the dislike of this drink keeps pace. „Es scheint danach alsob die Verachting und allmahliche Verdrangung der Sura ganz und gar mit der Ausbreitung des Brahmanismus zusammenhinge". This very clearly appears from the appreciation of the Sautramanï, a sacrifice to Indra, who has indulged in a little too much soma. In this ceremony thesouthern point of the compass dominates. The sura must be drunk by a Kshatriya or a Vaisya. „Man hat sich von brahmanischer Seite bemüht sie (die Sautramanï) als ein Somaopfer zu erklaren; aber ohne Erfolg". Hillebrandt sees in this difference of appreciation a proof of two influxes into India, both by Aryans, viz. firstly that of the sura-people; secondly that of the soma-people, who ousted the sura-cult at last. In our opinion it is more probable that in this 1) Hillebrandt — V. M. I. p. 244 seq. 2) See p. 237. 3) A. V. XIV. 1, 35. „allmahliches Sinken des Ansehens der Sura", „das Getrank unbrahmanisch lebender Ariër", the growing dislike of the sabha and lts association is discernible among the Brahmins. Knowing that sura is often mentioned in the same breath with dice and in the passage of the A. V. together with the sabha, the samiti, and the army near the celestial Vratya, we maysafely conclude that the phenomenon observed and described by Hillebrandt, expresses the dislike of the Brahmins, of the potlatch-ritual. This may also be the reason why the texts waver on the subject of the rite of lighting the sabha-fire; and also the reason of the, in itself, quite incomprehensible opinion of some texts, according to which the initiation could be held at any time, except in the rainy season which was relegated to the lower classes; and this in the face of the fact that there are unmistakable indications of the existence of the initiation in this very rainy-season, which is indeed the period „par excellence for the study of the Vedas. This is the reason also why dicing, although in the ceremonial of the Dïvalï-feast it has preserved a little of the nature of the potlatch, up to this day, is often severely condemned in the texts. Thus the Epic contradicts itself, and often verbally disapproves of what it actually propagates. It is also clearly discernible from this that the Epic has especially been elaborated by Brahmin scholars. Indian societies have not given rise to a more absolute kingship. The power, which has successfully been able to maintain itself against the association, was that of the caste, and especially that of the Brahmin-caste. In other societies with clubs one sometimes observes kingship develop itself from the club and its rank-indication. We should like to enter a little further here into the speculations we gave in the third chapter,ontherelationof Vishnuism, Sivaism and Brahminism. In this chapter we stated our objectloos to the conception of Sivaism and Vishnuism as sects, which were hindumzed in a reviving Brahminism. We proposed to conceive Krishna and Rudra as the principal characters of the rainy-season-rïtual, which (in the period of the Brahmanas, in which, indeed, the Vratyastomas are found) centred round the clubs. As we already explained, clubdom in India could not develop fully, being hampered by the tie of castedom. The caste, especially that of the Brahmins, has been able to maintain its position. It is they who have managed the religious, literary property of the Indian people; they have raised the entire tradition to the same level, by which in many respects the sharp traits have been lost. We especially refer to the MBh. Brahmin-caste and Kshatriya-club, both sprung from the same tribal organization, lived at first on the same myths and rites. They are indeed opposed to each other, within the compass of the same culture. The club usurps the power and the wealth, and in accordance with its character, ever aspires for the occupation of the first place. Thus it exhausts its own strength, for in the long run one club takes the front rank. The head of this club is for ever trying to subdue the remaining clubs. Castedom, however, extended over the entire Indian population. The curious part of the organization is not, that the highest class has organized itself in exclusive groups; for this is also found elsewhere. But that the whole of society split into separate groups which, the highest as well as the lowest, consider themselves to make up the whole of society. Caste is not limited to a few higher classes, but has its roots in the entire national existence. All circles, also the lowest, are organized in castes x), and possess, in their corporative formation, a powerful, if often passive resisting-power against the intervention o high-handed kings. , . , . At one time warriors and priests co-operated in the tribal ritual. Later on the Brahmins, „die Wissenden, die Besitzer der mystischen Kraft" are opposed to the kings, „die machtigen und zahlungsfahigen Auftraggeber", als „nvalisierende BeOn the occasion of weddings and suchlike the caste still always alors oblisés de déployer des gens appartenant meme aux castes les plus S' ütfestsipauvre caste, nous 1'avons vu, qut ne garde son >1 T p 205, OTQ 008 Horse'-shoe-shaped Opening, 2x3 SCQ« Hypergamy, 60 seq., 75, 83, 85. Initiation, 178 seq., 212, 231, 241 seq., 291, 3ii» 325/ 333 seq. Inversion, 7, ix, 325> 331- Jatakas, 15, 39» 43» 84, 91. Jati, 40, 43. Jnana, 119, 120, 153» J54» Iö5» 336 seq. Kali, 260, 279 seq., 315» 326. Kamsa, 104,161,177» 182,183,188. Karman, 109, 119, 154/ 33° seq. Kauravas, 7, 10, 295 seq. Kayon, see Horshe-shoe-shaped Opening. Kinship Exogamy, 44, 46, 59» b0> 89. Kinship Terminology, 45, 4°/ 47» 57, 294» 295* Krishna, 5, 6, 10, 15, 19, 20, 22, 27, 104, 160 seq., 215» 23X» 279» 299» 300, 313, 334 seq. Krishnajanmashtami, 191, 19°/ 200, 210. „ Kshatriya, 4, 5/ 85, 19I/ 27, 283 seq., 336, 339; and Society, 292, 328. Kula, 49, 85. Language, 28 seq., 100 seq. Magie, 148 seq. Mana, see Brahman. Manas, 120, 125 seq. Manthana, 125, 14° seq., 210. Marriage, 53 seq., 71, 73/ 75» °3» °5> 95- Matriarchy, 51 seq., 73, 17°/ see Gotra. Men's House, 203, 226, 234, 235» 241 n. 1, 274 seq., 307, 313, 323» 324. Myth, 5, 98 seq., 331; see Epic, Rite. Nala, 260, 326. Naman, 118, 262, 273. Nivritti, 128, 129, 139/ x45> r95 n. 1, 269. Panchalas, xi, 15/ 298, 310. Pandavas, 6, 7, 9, 15» "2, 295 seq. Patriarchy, 51 seq., 95 n. 1, 170. Phratry, see Division into Two; Rivalry of, 65, 144» x6g, x8o, 229, 245, 301 seq., 3x6, 322, 326. Pillar, 206 seq., 216 seq., 231, 315- Pot, 208 seq. Potlatch, 244 seq., 254, 281 seq., 303, 307, 311 seq. Pralaya, 127 seq., 159- Pravritti, 127, 128, 129» 138/ 145/ 195 n. 1, 270. Punch-marked Coin, 228, 256 seq. Race, 32, 90 seq.; see Caste, Varna. Rainy Season, 121, 189 seq., 219» 220, 229/ 230, 277/ 289 n. 1. Reincarnation, 129 seq. Rite, 104, 107 seq., 143, 244, 247/ 270, 277, 337/ see Myth, Bran- Pntiflirt. man, Rudra, 216 seq., 237 seq., 263, 278/ 279/ 313' Rüpa, 117 seq., 261 seq. Sabha, 233 seq., 258, 27I/ 274 seq., 289, 315/ 316, 323Sakuni, 298, 3°4/ 3*3/ 315 seq. Samhara, 128, 129, 221. Samkhya, 122 seq., 153/ 33° seq. Sapinda, 44, 67, 68 seq., 96, 97. Serpent, 143, 196/ 229 seq., 245/ 247, 250, 256, 260, 261, 267, 273/ 306. Óiva, 129, 157/ 158, 187/ 216 seq., 223/ 229/ 273/ 278. Sivaism, 7, 100, 109, 156, 167, 217, 291, 293. Society, 238,241, 244 seq., 283 seq., 292, 307, 308, 333, 345éramana, 154* Srij, srishti, 128, 129. Status Indication, 41, 47, 241 seq., 283, 307. Sthanu, 216; sabhasthanu, 277; see Siva. Stüpa, 215, 226 seq., 258, 270. Südra, 85, 89 seq., 241. Supreme Being, 107 seq., 107, 175/ 186, 223; see Brahma. Sura, 237, 290, 291Svayamvara, 33. Transmigration, 129 seq. Tree, 141, 245, 258. Upanishads, 154, 155, 217, 340. Vach, 120. Varna, 39,40,43,92,93> 94, "6,262. Vedas, 4, 39, 44, 83, 84, 87, 89, 91, 98, 99» 101, 106, 116, 149, 154, 155, 217, 271. Vishnu,i27,128,145,157 seq., 190, 223, 229. Vishnuism, 7, 100, 109, 156, 167, 217, 191, 293. Visvarüpa, 262. Vratya, 237 seq., 291, 293, 308, 309, 310. Wayang, 212, 346 seq.; see Kayon. Yoga, 337, 339, 341. Yudhishthira, 13, 16, 304, 315, 324 seq.