Page images

and go away. The Prastoty, with the word eșă, urges the Hotr to begin the Niskevalya Sastra. The Hotṛ then utters after the last Pratihāra the Āhāva, adhvaryo śomsāvo3m.1

The Śrauta Sūtras proceed, in book xviii, to give in detail the composition of the several parts of the Mahad Uktha. These details will be found abstracted in the notes to my edition of the Aitareya Aranyaka and need not be repeated here.


The Mahāvrata section is followed by the Kauṣitaki Brāhmaṇa Upanisad, forming books iii-vi of the Aranyaka. The first book deals in a confused fashion with the fate of the soul after death, apparently attempting to reconcile the double version of the fate of the dead presented in the Bṛhadāraṇyaka and Chandogya Upaniṣads. The second is an exposition of the pre-eminence of Prāņa as the truth of the universe, but is mainly devoted to showing the practical and quasi-magical uses of the conception. The third is more philosophical, and identifies the real with the inner self, the subject. In the fourth there is a later and more elaborate version of the questioning of Ajataśatru by Bālāki found in the Bṛhadāraṇyaka Upaniṣad.

Then follows the Samhita Upanisad (books vii and viii), treating of the symbolism of the Samhita, Pada, and Krama texts of the Ṛgveda. Then an Upanisad (book ix) dealing with the rivalry of the Prāņas, one of the commonest of Upanisad topics. In the following book (x) the internal Agnihotra is minutely described as a substitute for the formal sacrifice; the next book (xi) contains a brief account of the Prāṇasamvada, presages of death, and a set of spells. In book xii


Śānkhāyana Śrauta Sūtra, xvii, 16; 17. On śomsāvo3m, cf. L'Agniṣṭoma,

p. 232.

2 Ed. and trans. by Cowell, Bibl. Ind., 1861; ed. as one of 32 Upanisads in Anandāśrama series, 1895; trans. by Max Müller, S.B.E., i (2nd ed., 1900); de Harlez, Louvain, 1887; Deussen, Sechzig Upanishad's des Veda, 1897, pp. 21-58. I follow in my version the recension adopted by Cowell, noting all variants which make the sense different. The Upanisad is analysed after Anquetil du Perron's version and a Chambers MS. by Weber, Ind. Stud., i, 392-420. As usual, Anquetil's version is now of no real service.

3 Deussen, Phil. of the Upanishads, pp. 336, 337. 4 Ibid., pp. 87, 88.


Four new definitions are added to the original

is a hymn, or rather spell, addressed to an amulet of Bilva. Book xiii reverts, in a series of quotations inaccurately attributed and cited, to the pre-eminence of the self; book xiv insists on the need of knowing the meaning of the Veda, and book xv contains the Vamśa.1

For books i and ii the translation follows the text of Dr. W. F. Friedländer (Berlin, 1900), for iii-vi that of Cowell, for vii-xv that published by myself. Occasional reference is made to two MSS., the Berlin, MS. Orient. fol. 630 (ff. 82, dated Samvat 1734 (A.D. 1677), at Rajapura, but many pages injured by water and imperfectly restored), and the Bodleian,3 MS. Sansk. e. 2. Both of these are excellent MSS., and in most places correct each other's deficiencies. I have derived much benefit from the translations of the Upanisads, especially from Cowell's version, a very remarkable piece of work for so early a date, and from Dr. Friedländer's rendering of book i. Of native commentaries I have only seen that of Sankarānanda on the Kaușitaki Upanisad, which is a fair interpretation of the text, but contains many blunders. The Upanisad is also paraphrased in part by Vidyāraṇya5 in his Sarvopaniṣadarthānubhūtiprakāśa, of which the eighth and ninth chapters, corresponding to the fifth and sixth books of the Āraṇyaka, are printed in Cowell's edition, and follow the version adopted by Sankarananda. This is probably explained by the fact that Sankarananda is described as the teacher of Madhavacārya 6

1 For further details, see my article, J.R.A.S., 1908, pp. 363-88, which deals also with the probable date of the several parts of the Aranyaka The relations of the Aranyaka to the Aitareya Aranyaka are fully discussed in my edition of the latter (pp. 30 seq., and in the explanatory notes), in which will also be found an index of the proper names of the Sankhāyana, the text of and a very full index to Adhyāyas vii-xv, and a full index to Adhyāyas i and ii. Jacob's Concordance includes the Upanisad. For the relation of the Aranyaka and Śānkhāyana Śrauta Sutra, see my note, J.R.A.S., 1907, pp. 410-12; Eggeling, S.B.E., xliv, pp. xliv seq.

2 Weber, Berlin Catal., ii, 5, 6. The MS. was very kindly lent to me by the Royal Library, through the India Office, by Prof. Pischel's suggestion. 3 Winternitz & Keith, Bodl. Catal., p. 60.

• In Cowell's edition; cf. Max Müller, S.B.E., i, p. c.

5 That is, Madhava; see Klemm, Gurupūjākaumudi, pp. 41 seq.

Hall, Bibliogr. Ind., p. 98; Max Müller, S.B.E., i, p. c; Deussen, Phil. of the Upanishads, p. 29.


and pupil of Anandātma. The version of the Upanisad given by these authors prevailed in the south, but its inferior validity is shown by the fact that Sankara3 followed the other text. Rāmatīrtha on Maitrāyaṇī Upaniṣad, iii, 2, quotes v, 8, from the ordinary version, and the upper limit of age for Sankarananda's version is uncertain.

A word may be said in conclusion on the philosophic merit of the Upanisad. Its contribution to thought lies in v, 5-8, where the unity of consciousness, the interdependence of the organs of sense, the activity of sense, and the objects of sense, and their unity in consciousness are expressed with some clearness and detail. Such passages in truth represent the highest doctrine of the Upanisads; the further step which identifies this unity with the Brahman and finds the macrocosm in the microcosm is conceived rather religiously or mystically than philosophically, nor is any attempt made to prove it, while the unity of consciousness is established by tolerable arguments. No doubt the Upanisad stands on a much lower plane of thought than the Theætetus or Parmenides, or the de Animâ, and the ideas of Plato and Aristotle are infinitely more subtle and complicated; but the fact remains that the Upanisad-probably of earlier date-does deal with a philosophic problem in a philosophic spirit, however much that spirit may be confused by mythology. It should be noted that this section is of Brāhmaṇic origin, and that the speculations of Citra in iii deal with pure mythology in the doctrine of the paths after death, a fact which led Max Müller to ascribe to the Ksatriyas a special interest in this unphilosophical topic, and scarcely speaks well for their theoretical devotion to pure knowledge as against ritual.

1 Hall, p. 116; Winternitz & Keith, Bodl. Catal., p. 75.

2 In the Telugu edition, Madras, 1883, of 108 Upanisads, there is no Kausītaki (cf. Deussen, Sechzig Upanishad's, pp. 533, 534), but it is included in a collection of 129 Upanisads known to the Andhrika Brāhmaṇas, made in 1850-1, in Telingana, by Sir Walter Elliot, Eggeling, India Office Catal., p. 122, and agrees with Sankarananda's recension. Anquetil usually follows it, Cowell, p. viii.

3 Cowell, p. 5. Deussen, Phil. of the Upanishads, p. 28, gives the references to Bādarāyaṇa's Sūtra.

The text of the Aranyaka is on the whole in a satisfactory condition. The exceptions to the rule are mainly in the case of forms whose meaning is obvious, but which contradict established grammatical rules.1 It is obvious that the tradition was unable to discriminate between Vedic forms and mere textual blunders, and no reasonable scholarship will hesitate to amend all the forms given by Cowell as 'strange solecisms which sometimes half remind us of the gathās of the Lalita Vistara' from the Upanisad, though the mode of emendation may be doubtful. In other cases it is uncertain whether we have a rare usage or a text error. For example, in iv, 7, we have it said of Sarvajit Kauṣītaki, yad ahorātrābhyām pāpam akarot sam tad vṛnkte, where the imperfect 2 stands in a curious relation to the present, here probably historical.3 Or again, in iii, 4, Sankarananda reads dhunute ra, where va cannot be ignored as it is by Max Müller and Deussen. He renders it aśvā iva romāņi kampanena, and Cowell regards this as possible. Yet it is hardly possible to doubt that it is a mere transposition of dhunuvate, actually read in some MSS., and that again an error for dhunvāte, the verb being intransitive. Or again, in vi, 2; 3, we cannot accept a masculine nominative bṛhat, or in iv, 4, abhi vātāt (Berlin MS.), etc. On the other hand, genuine archaisms exist, e.g., svapnyayā in vi, 15, yajūdarah in a Re in iii, 7, and in several other cases the evidence for hyper-Sandhi is convincing; e.g. iii, 5, where udgithopa(ra)śrayaḥ must represent udgitha upao. Or again, in vii, 2; viii, 1, the genitive asya is remarkable, but certain, and may perhaps be compared with atha mahāvratasya, Aitareya Āraṇyaka, v, 1, 1, though there the following words can be


1 Viz., nişiñca in iii, 2 (sişiñca or sişikta); praiti (read pūrvā prajā for purvāḥ prajāḥ) in iv, 8; 10; samvesyan (samvisyan or samveksyan) in iv, 10; veti (vyeti) in v, 1; adūḍham (udulham or adūduhat) in v, 5, etc.

2 Cf. the curious use of ajayat in Maitrāyaṇī Samhita, iv, 3, 1, discussed by Delbrück, Festgruss an Böhtlingk, p. 25, and Synt. Forsch., ii, 89 seq. ; v, 279, 577, 586.

3 This use of the historical present is doubtful, cf. Delbrück, Synt. Forsch., ii, 90; v, 278; and it may here be the true present, Sarvajit Kausītaki being alive, which is of course quite possible; then akarot becomes even more difficult.

Cf. the gen. with vid, Synt. Forsch., v, 159; Caland, Altind. Zauberritual, p. 18, n. 2.

more easily construed with the genitive. In iv, 8, the accusative with man, the subjects being the same, is against the Vedic use 1 observed in vii, 8 seq., but occurs in a Ṛc. The secondary character of ix, 7, is shown by the genitive with brũ,2 and of viii, 11, by the use yo 'tra vicikitset saṇakāram eva brūyād ṛte ṇakāram iti, where the usual Brāhmaṇa construction requires the first person, as in i, 1; ii, 17; viii, 6, and in the parallel Aitareya version.3 So in viii, 1, and xi, 8, tsariņir and Săvitrīr present apparent feminine nominatives in s, which may be pseudo-antiquities or merely errors of the text, as is suggested by the fact that the former word is very variously read in the parallel passages and the latter occurs only in one of the MSS., the other having the correct form. In viii, 9, the MSS. agree in reading the accusative param, which is quite impossible, and so forth.


1 Synt. Forsch., v, 104, 179.

2 Ibid., p. 162.

3 Cf. Maitrayani Samhita, ii, 1, 11; yo vai kāmayetānnādaḥ syād iti.

4 Cf. sabalis in Taittirīya Samhitā, iv, 3, 11, 5, but śabalī in Kathaka Samhita, xxxix, 10; ghaṭakarkarir in Sankhayana Srauta Sutra, xvii, 3, 12; ghatarir, ibid., 15.

« PreviousContinue »