the beginning of Krttiká. But, as is pointed out in the article referred to, Colebrooke might have regarded this as a mere looseness of statement, upon which no stress need be laid, when coupled with the recognition in the same treatise of Kṛttiká as first asterism—a position to which it was no longer entitled, if the vernal equinox were contained in Bharaní. This explanation is doubtless less probable than the other; yet it need not be altogether overlooked; nor does it ascribe to Colebrooke the taking of any undue liberties with his authorities.

By far the most serious error, however, underlying all these and other similar calculations, yet remains to be pointed out. They assume unquestioningly two things respecting the ancient Hindus of the period represented by the Jyotisha: 1st, that they had a precise determination of the limits of their asterisms; and 2nd, that these limits precisely coincided with those of the later system. Now we have no good reason for supposing either the one or the other of these things to be true. We must look upon the asterisms as having been, in their inception, simply a series of twenty-seven constellations bordering upon the moon's track, selected, in view of their distribution and their conspicuousness combined, to mark the successive daily steps of her progress through the heavens. In idea, they divided her path into so many equal portions; but it was in the nature of things impossible practically to realize this idea. No Hindu of the olden time, however skilful, could have determined for himself, or pointed out to others, where was the precise end of any one twenty-seventh and the beginning of the next. Indeed, the very attempt to do so belongs to an era of scientific impulses, instruments, and methods, whereof we discern no trace in India until it has felt the influence of Greece. The asterismal "portions" of the ancient system were nothing more than parts of the moon's path, as nearly equal as the unassisted eye could measure them, lying adjacent to the asterisms which marked and denominated them; he who employed them knew well enough that the "beginning of Dhanishṭhá" and the "middle of Aśleshá" were, by the theory of the method of division, opposite points in the sky; but he could not have found either point, reckoning by equal twentysevenths from a definite initial point, otherwise than in a rudely approximative way.

But again, even if we suppose the limits of the ancient system of division to have been capable of as exact definition as those of the modern, we have no right to take for granted that the two would coincide. The starting-point of the later division is determined by the position of the vernal equinox in the sixth century of our eraVOL. II. [ESSAYS I.]

9

a point which had a value at a period when the modern science was assuming its shape, but was of no particular account to the ancient science. The considerations have yet to be discovered and pointed out which should lead to the identification of any precise spot as the starting-place of the old asterismal division. And, if our view of the original character of the system, given above, be correct, such considerations will never be found; the limits of its asterisms will never admit of even a near determination. From the fact that Kṛttiká is the first asterism of its series, we can only infer that, at the time of its establishment, the vernal equinox was supposed to be in the vicinity of the Pleiades-we cannot even say whether within the limits of an arc of 13° adjacent to that constellation.

When, then, we combine the two sources of uncertainty already set forth, on the one hand, the inaccuracy of even the modern Hindu measurements, giving a range of possible error amounting to four centuries; and, on the other hand, our ignorance of the comparative situation of the ancient division-lines, and our presumption that there never were any such admitting of precise location,-it must appear evident that the attempt to point out with definiteness the points on the ecliptic to which the description of the Jyotisha shall be understood as applying is of the most hazardous character. But in estimating the value of the Jyotisha datum, we need further to take into account the difficulty of the observation it records. The place of the equinox has to be determined by indirect and roundabout methods, by observations and combinations and inferences which lie quite out of the power of men unskilled in astronomical science-which, in the hands of any people that preceded the Greeks, would not be likely to come very near the truth. We are far, as yet, from fully understanding the early astronomy of India: but nothing that we know of the dispositions and capacities of either the ancient or modern Hindus, or of the astronomical work which they have done, gives us reason to believe that they would have clearly grasped the conditions of the problem here in question, and solved it successfully. We must even regard it as for the present doubtful whether they solved it at all; whether they did not get the observation, and even derive the asterisms themselves, from some other people.

Making due allowance for this additional source of doubt, we shall see clearly that no definite date, and especially no date applicable to a determination of the chronology of ancient Hindu literature, can be extracted from the record we are discussing. It is difficult to make a valuation in figures of elements so indefinite;

but we should say that a thousand years would be a period rather too short than too long to cover all the uncertainties involved.

As is pointed out by Sir T. E. Colebrooke in the note already referred to, the author has, in other of his writings, recognized the uncertain character of the astronomical data afforded by the ancient Hindu works, and the impossibility of deriving exact dates from them. Had his design in publishing his Essays been to make anything more than a collection and unaltered reprint of them, he would probably have guarded, by a note at this point, against too strict a construction of his statement, or too confident an inference from it.

It remains only to apologize for the extreme length to which this note has been allowed to extend itself-a length excusable only by the interest of the subject discussed, and the intricacy of the conditions involved. Some of the points here touched upon are more fully treated in the paper by Prof. Whitney in vol. i. of the Journ. Roy. As. Soc., already more than once referred to. For the asterismal system and its relations, see also the Essay "On the Indian and Arabian Divisions of the Zodiac," in the next volume: and, among later discussions, especially Biot's Recherches sur l'Astronomie Indienne et l'Astronomie Chinoise (Paris, 1862; made up of articles originally published in the Journal des Savants for 1839-40, 18591861); Prof. Whitney's notes to the eighth chapter of the translation of the Súrya Siddhánta, published by the American Oriental Society (their Journal, vol. vi., 1860), and his articles in a later volume of the Society's Journal (vol. viii., 1866); and Weber's articles on the Vedic notices of the nakshatras, in the Berlin Academy's Transactions for 1860–61.

P. 99, I. 25. The Ráma-Tápaníya Upanishad is published, translated, and commented by Weber, in Trans. Berlin Acad. for 1864, p. 271 sq. The Gopála-Tápaníya is published, with Viśveśvara's commentary, in the Bibl. Indica (Calcutta, 1870).

P. 99, 1. 29.

p. 286.

For the summary referred to, see Weber, as above,

P. 100, 1. 4. Weber (as above, p. 271) calls the tápaniya Upanishads "the latest runners which have attached themselves as parasites to the branches of the Veda-tree." But he pronounces the Nrsinha-tápaniya, the oldest of them, to be at least as early as the seventh century of our era. As to the age of the worship of Ráma, see the same authority, at p. 275 sq.

P. 100, 1. 18. For the character of the religion represented in the earliest parts of the Vedas, see previous notes: thus, for their monotheism, the note to p. 24, 1. 25; for the worship of stars and planets, the note to p. 25, marg. note. Of the grand triad of

gods of the later religion, Brahman is no divinity in the Veda; Vishnu is not at all prominent; the name Siva is not known, and Rudra, the nearest correspondent of the modern Siva, is of quite a different character from the latter. See Muir's Original Sanskrit Texts, vols. iv. v., and the same scholar's articles on the Vedic Theogony and Mythology, in Journ. Roy. As. Soc. vols. i. ii.

P. 101, 1. 13. Respecting the Atmabodha or Atmaprabodha Upanishad, see Weber's Ind. Stud. vol. ii. p. 8 sq.

P. 101, note. The Sundarí Upanishad is identified by Weber with the Tripurá; see his Ráma-tápaníya Upanishad, as above, p.

272.

P. 102, end. As was altogether natural in the case of one who was the first investigator in so wide and difficult a field, and hence compelled to rely in part upon Hindu commentators and Hindu assistants, Colebrooke failed to classify distinctly in his own mind the enormous mass of works included under the name Veda, and to apprehend the relative value of the different parts of it; he therefore did not fully appreciate the transcendent importance of the original hymn-collections (especially the Rig-Veda), the nuclei about which the whole literature had grown up, as containing the germs of the whole after-development of Indian religion and polity, which are unintelligible without their aid-as, moreover, illustrating a phase of human history of a rare and most instructive primitiveness, and as casting light upon the mythologies of other races, proved by the evidence of language to belong to the same great family. Had it been otherwise, he would have been careful not to put into his closing paragraph words which sound so much like a discouragement to other scholars from following where he had led the way; he would rather have earnestly commended to them the diligent study of works which constitute, in their bearing upon the history both of India and of Europe, the most interesting and repaying portion of all the Hindu literature.

133

III.

ON THE DUTIES OF A FAITHFUL HINDU

WIDOW.1

[From the Asiatic Researches, vol. iv. pp. 209-219.
Calcutta, 1795. 4to.]

[114] WHILE the light which the labours of the Asiatic Society have thrown on the sciences and religion of the Hindus has drawn the attention of the literary world to that subject, the hint thrown out by the President for rejecting the authority of every publication preceding the translation of the Gitá does not appear to have made sufficient impression. Several late compilations in Europe betray great want of judgment in the selection of authorities; and their motley dress of true and false colours tends to perpetuate error; for this reason it seems necessary on every topic to revert to original authorities, for the purpose of cancelling error or verifying facts already published; and this object will no way be more readily attained than by the communication of detached essays on each topic, as it may present itself to the Orientalist in the progress of his researches.

From this or any other motive for indulgence, should the following authorities from Sanskrit books be thought worthy of a place in the next volume of the Society's Transactions, I shall be rewarded for the pains taken in collecting them.

1 [On this paper, cf. Prof. Wilson's Essay on the supposed Vaidik authority for the burning of Hindu widows, and Raja Rádhákánt Deb's remarks, originally published in the Journ. R.A.S., vols. xvi., xvii.; and republished in Wilson's works, vol. ii. pp. 270-309; Prof. Roth, Zeitschr. d. D. M. G., vol. viii.; Prof. Max Müller, ibid. vol. ix. Cf. also Dr. Hall, Journ. R.A.S., 1867, and Prof. Müller, Chips, vol. ii. (2nd ed.), p. 35.]