traditions far older than the formulas in which they are handed down. But we must, of course, be conservative against allowing a too great elasticity of time in this regard. Such concrete examples of political wisdom and bon mots of polity are worthless except as possible embodiments of older usage; for they include much that the Epic proper does not touch upon, and that could not have been contained in the earlier version. Political or social truths, therefore, involved in many of these sayings can be predicated positively only of the period following the composition of the original poem, and asserted as universal only when strengthened by legal evidence of greater antiquity, or by support in Epic practice. For a true interpretation of the large collection of inferable and formally stated sociological data in the Epic, it is necessary to draw first a sketch of the old and then of the new world thereby presented. It stands to reason that in general much will here be found doubtful and open to criticism; and that, in particular, a certain meagreness will characterize the first, a comparative richness and perhaps suspicious fullness of detail the second picture. But even the modern Epic, the full completed work—were we to deny to the student the chance of discriminating accurately between the bodies of material necessary to the making up of his two sketches-is not as a whole unimportant in the elucidation of the customs of India in the Middle Ages, reaching back more than two thousand years; though it may be that further study will necessitate our giving a much later date than has been assumed to much of the pseudo-Epic. Further, the impossibility of effecting a complete discrimination of old and new may make it seem to some a vain task to distinguish the factors by their age. It is true that our verdict as to which is early and which is late must in a measure be based upon purely a priori assumptions; while it should, where this is possible, certainly be dependent upon an intimate acquaintance with the literature preceding and following the Epic; for many of the threads of our poem are older than its present literary form, and have often been preserved as fragments caught in a substance foreign to them; while what influence, on the other hand, of other sects or other races has made itself felt in the re-weaving of the tale needs careful analysis, being yet far from determined. But a review of what the Epic tells us may, it seems to me, be serviceable in supplying facts that in turn may help the critique of the Epic itself, after these have been compared with results drawn from other sources. I have therefore collected the Epic data as positive aids to research, but as historical material would use them at present only tentatively.

III. SOCIAL POSITION OF THE RULING CASTE.

I have called the warrior-caste the ruling caste of India, because all power, political as well as martial, lay in the hands of the military organization.

The hold that the priest obtained upon the king has been compared with that secured by the European priest in the Middle Ages. There is a great and essential difference. The Hindu priests had no strength of combination. They formed no union of political power parallel to, and capable of opposing itself as a whole against, the sovereignty of the throne. It is true that they formed an association, that they were an exclusive and distinct class. But they formed no corporate body, and had no head. They worked as individuals. Moreover, their power possessed no financial basis such as that of the Roman church. They drew no direct and constant property-contributions from the people. They were dependent on the king. From him they obtained largesses; from him, or rich members of his caste, they obtained their wealth of cattle and later of land: wealth that did not, however, go to swell a general fund, but enriched favored individuals. They lived on charity, and stood under armed patronage. Their very exclusiveness hindered their upgrowth; for had they with the religious tools familiar to them been able to ally themselves by marriage with the nobles, had the priests' daughters (for by their law celibacy was forbidden) wedded the priests' protectors, the religious order might by such family alliance have gained a thorough control of the state. As it chanced, mighty as was the individual ministerial influence of certain priests, coercive as was the religious power they could wield, they still stood apart from the rulers, depending on those whom (it is only fair to say) they despised-a fragmentary class, that enforced respect as a whole through fear of the fate to come upon the king that denied their influence with the gods, but never a class that rose to be independent of that king in respect of support. Moreover, as matter of fact, most of the priests lived retired and quiet lives, content to beg for food, satisfied with a little rice, cows, and a hut, and without worldly ambition; regarded with love or awe by the common outside world, with honor by the nobles, and only occasionally, in the person of the king's private priest and advisers, interfering at all in the matters that concerned the warrior-caste.

This view is of course not brought forward in the law-books. Their authors were the later priests, who regarded the world as made for them alone, and looked upon the king as a steward divinely appointed to provide them with what they needed. Far otherwise appears the Aryan state in the early Epic. A freer life is found here. The king is a king, not an appendage

1

to the priest. The view of the formal law reflects the vain ideas of men conscious of mental superiority and anxious to bring the state into harmonious relations with their egotism. The Epic, an unconscious mirror, furnishes social facts as found in an age as yet comparatively independent, and portrays conditions that survived even the unscrupulous handling of the text by those opposing this independence. The priest is the standard of the world,' says the formal law; the king is the standard,' says the Epic.*

In looking at the state from a political point of view, we must, therefore, reverse the arrangement formally proclaimed by the priests themselves, and put their order below that of the military caste. And next came the 'people.'

There were, thus, three Aryan castes in the Epic period. The ruling caste, comprising the king, his great lords and vassals, together with the knightly part of the army; the priestly caste, elevated by religious knowledge, often individually powerful as guiders of the king's will, but otherwise forming a lowly class of penitents and beggars, who, if not irritated into a wasplike wrath by unprovoked insult, remained a sedate, humble, and morally useful element in the state; lastly the third caste, called collectively the people, exalted only through their Aryan blood and their fully allowed claim to all Aryan privileges in the matter of legal rights and religious rites, but otherwise constituting a body that was looked upon with contempt by the military and extolled only by the priests. These knew the source of revenue. There was, too, another and un-Aryan caste, of which the members were, to the Aryan 'twice-born' (re-generate through holy ceremonies), merely once-born,' or 'deprived of good birth. These had, barring pretense, no spiritual or legal privileges. They possessed no property. Their

* Manu xi. 85: brāhmaṇaḥ pramāņam lokasya; Mbh. i. 82. 18: rājā pramāņam bhūtānām. The law and Epic give a well known rule to the effect that the right of way belongs to a blind man, a woman, a burden-bearer, and a king; while, if a king and priest meet, the way belongs to the priest (Vas. xiii. 59; Ap. ii. 5. 11.6; M. ii. 138-9; Mbh. iii. 133. 1; xiii. 104. 25, etc.); but the same Epic recounts a scene where a king meeting a priest calls out to him get out of my way' (apagaccha patho 'smākam), and when the priest repeats the law eternal,' as just quoted, it is without effect, and the king even smites him with his whip. The king is cursed by the priest and becomes a demon, but this 'law' and anecdote of Vasishtha may serve as an example to illustrate the gradual increase in priestly power, and the means by which it was obtained.

Compare v. 132.30: Let the priest beg; the knight defend the people; the people-caste make money; the slave be a server'; or xii. 91.4: work is for the slave; agriculture, for the people-caste.'

Na hi svam asti çūdrasya etc., xii. 60. 37: compare the whole section; with 39, Pāijavana, cf. M. vii. 41. The slave might clean up after the offerings were made (çūdrāir nirmārjanaṁ kāryam, xii. 294. 12–15 ;

wives were so in name.

Their lives depended on their owner's pleasure. They were born to servitude,' for they came from the foot of God." They were in fact the remnant of a displaced native population, marked by race-characteristics and stigmatized by their conqueror's pride as a people apart, worthy only of contempt and slavery. It would at times appear as if the slave were the especial servant of the people-caste, although bound also to serve the other high castes (1.100.11; but compare xii. 72. 4-8, 6-M. i. 99). The color-distinction between the castes the slave being black; the people, yellow; the warrior, red; the priest, white (xii. 188.5; in ii. 36. 24 kālapūga means time, not color)-may possibly indicate a real difference of hue (compare Muir, S.T. i. 140; Zimmer, A. L. p. 113, with literature). The people in general, both those of the country (jānapada) and those of the town (paura), are divided into the 'people at large and the common people' (mahājana, pṛthagjana; cf.xii. 321. 143; xiv. 90. 14) as distinguished en masse from the aristocrats, and so may include the people-caste proper and its leavings-that is, those following or adopting occupations too low to be recognized as fit for Aryans (under which name the three upper castes or the people-caste alone are meant). The statement that there is no distinction of castes' is meant proleptically, and implies merely that a man changes his caste in future births on earth in accordance with his acts in this; so that e. g. a priest might become yellow and be born as a member of the people-caste,' if he failed to act as a priest should (maitrayanagatah; xii. 188. 10 ff.; cf. ib. 239.13; 279.5).† I prefer to keep the significant, if not absolutely correct, translations of the caste names rather than their native forms: thus, priest' for brahmana; warrior' for kṣatriya (man of the ruling order'); 'man of the people' for vaiçya (literally inhabitant,' or, in the Vedic sense of vic, a clansman' in general; but later confined to members of such families of the Aryan folk as had not renounced farming his glory is dākṣya (activity), to obey the three castes his duty (21); cf. 295. 1 ff.: a priest who becomes çūdradharman is degraded; the slave becomes a herdsman, however, or a tradesman, if he cannot support himself by service (3,4); 297. 27 ff.=M. x. 126 ff. cf. xiii. 165. 10. The pseudo-Epic also declares the slave to be the next removed from a beast (xiii. 118. 24), but in its religious portions admits that a slave may, like the other castes, get a reward for fasting (upoșița, xiii. 132. 14); just as it says that one of the slave-caste may be made a student (çrārayec caturo varṇān kṛtvā brāhmaṇam agrataḥ, xii. 328. 49) though the parikşeta rule precedes (47). The caste will be later treated in detail.

*It is an old myth that the people-caste came from the loins or thighs of God (Brahma or Manu), while the warrior-caste came from his arms, the priestly, from his head (mouth), and the slave-caste from his foot: RV. x. 90. 12; law, Vās. iv. 2; B. i. 10. 18.6; Epic, Mbh. viii. 32. 43 ff. ; xii. 60. 28 ff.; 319.90; R. iii. 20. 30 ff. (here from Manu).

+ Compare Muir, Sanskrit Texts, i. 138 ff. for this religious view and some quotations on caste-color.

Compare Latin flamen; see von Bradke, Beiträge zur Kenntniss der vorhistorischen Entwickelung unseres Sprachstammes, p. 13.

and trading to live an altogether military or priestly life); 'slave' for gudra. Out of these elements was made the theoretical state of the Hindus. Yet earlier only the three upper castes were recognized as tout le monde. Practically, there came in addition an early recognized and large number of 'mixed castes,' feigned all to have been either degraded branches of the pure castes or their illegitimate descendants. (Compare e. g. xii. 297. 8, 9.) In reality these were outcasts, made so either formally or naturally (through the increasing tendency of the lower orders to take up means of livelihood not sanctioned by the pure), or, far more, remnants of un-Aryan tribes gradually assimilating to the body politic of their conquerors, but not received into the social or religious body. These were increased, again, by ever growing immigration from strictly foreign nations of warriors and traders, practically esteemed but otherwise despised.†

But from people-caste, slave-caste, pariahs, or foreigners the *Compare etavad vā idaṁ sarvaṁ yāvad brahma kṣatraṁ viț, Çat. Br. (ii. 1.4. 12) iv. 2. 2. 14; Weber, Ind. Studien, x. 18.

+ Compare xii. 207. 41-45, for the disesteem felt for foreigners; xiv. 73. 25, for the distinction between Kirātas, Yavanas, with other barbarians, and Aryans; or vi. 9. 61, etc.; and xiii. 48 and 49 for the whole account of the mixed castes. The recoil from strangers is not much greater, however, than that of high caste from low caste in the developed state: 'equals with equals: let priests talk with priests; warriors, with warriors; men of the people, with men of the people; slaves, with slaves,' xii. 210. 11-12. Yet in a moral sense even the mixed castes are better than the pure if the latter do not do as they ought, iii. 180. 25-36. The high estimation in which certain foreigners, such as the Yavanas, were held is wholly due to the military point of view (see below). Aside from this accident they were all despicable, no matter whence they came, although the Madrakas were chief in positive unholiness and in the anti-Aryan character of their civilization. The Northerners in general were as bad as the rest. The sole exception is that of the storied Northern Kurus. These were looked upon as occupying an earthly Hyperborean paradise. The flank of Mount Meru is in one place given as their home (daksinena tu nilasya meroḥ pārçve tatho 'ttare, vi. 7. 2), but they are regarded as real people of the North in most cases, with laws more conservative than the Southerners (as in i. 122. 7), and blessed with a typical felicity (as in the first passage, and R. ii. 103. 26): so blessed in fact that while some happy heroes go at death to Indra's heaven, others go to the Northern Kurus (xv. 33.13 ff.), as a sort of earthly heaven. A gate sixteen yojanas wide used to be blocked up and prevent their passage; but they can now pass through (iii. 231.98). It is evident that they were half-real, half-mythical to the Epic poets. Other foreigners are simply barbarians, on a par with the Dasyus, which latter race may be blessed only by changing its way of life and becoming an adherent of Aryan rules (xii. 135. 22, etc.). Rather noteworthy is the tutoyer (tvaṁkāra) of a seer by a slave in iii. 137. 5-6, and by a Candala (tam uvāca sa canḍālo maharse crnu me vacaḥ, xii. 141.55. Compare 85, bhavan, and conversely, Viçvamitra to the Crapaca bhavan in 66, tvam in 80) though the custom is to exchange thou' with 'sir' without ceremony. But the rule is that, though persons younger or of the same age may be addressed with thou,' neither thou' nor the real name should be said to those that are better (older). Compare xii. 193. 25 (traṁkāraṁ nāmadheyam ca jyeṣṭhānām arivarjayet, etc.)