^{page 998 note 1} Both these men were contemporaries of Ātreya. Bheḍa, indeed, is said to have been one of his six disciples, and a unique manuscript of a Saṁhitā which goes by his name has survived. This work must have been available to Vāgbhaṭa I. But as no work of Kānkhāyana now survives, it is doubtful whether Vāgbhaṭa drew on an actual work of his, or merely on quotations from it surviving in other works of later date.

^{page 1000 note 1} Regarding the sources on which the Table is based, I may explain that Column I has the support of the Summary List of the chapters, at the end of the Sūtra Sthāna, in the two manuscripts, Government of India, No. 2503 (now in deposit with the Asiatic Society of Bengal), p. 695; and (partially) India Office, No. 335, fl. 123. It is taught in the commentary of Chakrapāṇidatta, at the end of the Cikitsita Sthāna, in Tübingen, No. 463, fol. 534b, and is adopted in the editions of Jīvānanda and Abinās Chandra. Column ii has the support of the Summary List, in the manuscripts, Tübingen, No. 458, fol. 177a, Tübingen, No. 459, fol. 163b, and Deccan College, No. 925, fol. 93a; also partially in India Office, No. 335, fol. 123. It has also the support of the actual order of the chapters in the Cikitsita Sthāna, in all six manuscripts available to me, viz., Tüb., 458 and 459, Ind. Off., 335 and 359, Decc., 925, and the old Nepal MS. (dated 303 N.E. = 1183 A.D.). It is adopted in the editions of Gangādhar, and of the two Sens.

^{page 1003 note 1} I adopt, for the sake of convenient reference, the divisions into paragraphs of the Jīvānanda edition of 1896, though it is by no means perfect.

^{page 1005 note 1} The facts seem to be these: Vāgbhaṭa I (in AS., vol. ii, p. 89, 11. 8–12Google Scholar) compressed in prose the substance of Charaka's versified remarks in verses 18–25 of his therapeutic chapter (p. 485), preserving a few catchwords (jitvā, mārutam, etc.). Afterwards Mādhava turned the compressed prose version once more into verse (MS., p. 261, vv. 1–4)Google Scholar, and in doing so preserved the same catchwords (mārute, vijite, etc.). Still later, Dṛiḍhabala added the prose of Vāgbhaṭa I and the verse of Mādhava to Charaka's genuine Nidāna (as §§ 20 and 21, p. 214), without apparently realising, not only that the prose and verse versions were duplicates, but that both these versions themselves were actually duplicates of Charaka's own genuine verses in his Cikiṭsita chapter.

^{page 1008 note 1} Lubricant, in the original, is sneha, or oleaginous preparation, especially medicated oil or clarified butter, to be taken internally (sneha-pāna).

^{page 1009 note 1} The scheme of Charaka is essentially one of three kinds, viz., tumours of one humour, of two humours combined, and of three humours combined. But the second kind admits of two varieties, viz., air plus bile, and air plus phlegm. Hence, in a sense, the scheme may be said to be one of four kinds. The scheme of Suśruta the younger (see p. 1022) is one of five kinds, viz., tumours of the air, bile, and phlegm humours (each singly), of the three humours combined, and of the blood. But the bile and phlegm tumours of Suśruta, though either of them seemingly of a single humour, are really identical with the air-bile and air-phlegm tumours of Charaka. Vāgbhaṭa I, in his Aṣṭāṅga Saṁgraha (vol. i, p. 288Google Scholar, 11. 8, 9), propounds a theory of eight kinds, viz., three tumours of a single humour (air, bile, phlegm), three tumours of a couple of humours (air-bile, air-phlegm, and bile-phlegm), one tumour of all three humours combined (air-bile-phlegm), and one blood-tumour. This, however, is mere scholastic trifling, and is practically admitted to be such by Vāgbhaṭa himself; for in his subsequent description of the several kinds of gulma he speaks only of the five kinds of Suśruta's scheme, but ignores entirely his own additional three (bile, phlegm, and bile-phlegm) as unrealities. The eightfold division of Vāgbhaṭa I is adopted by Vāgbhaṭa II in his Aṣṭāṅga Hṛdaya (vol. i, p. 784Google Scholar, v. 32). Mādhava, in his Nidāna (p. 172, v. 1)Google Scholar adopts the fivefold division of Suśruta; and he is followed by Dṛiḍhabala, who foists that division into Charaka's account of gulma (ante, p. 1003). Instead of §§ 1–3 of the existing redaction, the original text of Charaka may be suggested to have been something as follows:–Iha khalu trayo gulmā bhavanti | tad-yathā vāta-gulmaḥ saṁśṛiṣṭa-gulmo nicaya-gulmaḥ. ║ And combining this with Suśruta's scheme, above-mentioned, Vāgbhaṭa I writes (AS., p. 288, 11. 8b, 9): Gulmo 'ṣṭadhā pṛthag-doṣaiḥ saṁśrṣtair = nicayaṁ-gataiḥ | ārtavasya ca doṣeṇa nāriṇārṁ jāyate 'ṣṭamaḥ. ║ The scheme, found in the Bower MS., is the fivefold one of Suśruta II. Thus five gulma are mentioned in part ii, vv. 237 and 256, and the blood-tumour in part ii, v. 361. This places the date of the treatise in the Bower MS. after Suśruta II. In this connection the scheme of the Hārāta Saṁhitā is noteworthy. It includes five gulma; but the blood-tumour is not among them. It is based on the principle of locality: of these localities there are five, hṛd, kukṣi, nābhi, vasti, and madhya, and in them respectively there are five gulma, viz., yakṛt (liver), aṣṭhīlikā, granthī, caṇḍa-vṛddhi (hernia?), and plīhan (spleen). This scheme appears to include diseases to which the term gulma as used by Charaka and Suśruta does not apply at all.

^{page 1011 note 1} It may be noted that the formula in verses 65, 66 (in CS., p. 489), is found also in Suśruta, verses 26, 27 (in SS., p. 805). It is one of the few verbal coincidences between the textbooks of Suśruta and Charaka, and may be due to copying either from one another or from a common source —a point still waiting for exact investigation. Verses 108a and 132a are quoted verbatim in AS., vol. ii, p. 95Google Scholar, 11. 18, 19, by Vāgbhata I, who explicitly indicates them as a quotation by the prefixed phrase, bhavati c = ātra.-Some verses in the Jīv. ed. of 1896 are altogether spurious, being due neither to Charaka nor Dṛiḍhabala. To these belong v. 64a, on p. 488, which cannot be genuine for several reasons: (1) the ṣaṭpala formula, for which the verse refers the reader to the rāja-yakṣman chapter, as a fact occurs in the gulma chapter itself, at p. 495, in verses 143, 144, though with the name pañcakola; (2) the ṣaṭpala of the rāja-yakṣman chapter occurs on p. 531, where, however, there is nothing to identify it as the ṣaṭpala; (3) the rāja-yakṣman chapter, coming after the gulma chapter, the reference on p. 488 would be a reference forward to p. 531, instead of backward; (4) the verse is not found in the old Nepal MS. (fl. 243b), nor in any of the other MSS. accessible to me (Ind. Off. 335, fl. 279b; Ind. Off. 359, fl. 21a; Tüb. 450, fl. 424a; Tüb. 459, fl. 53a; Decc. 925, fl. 226a), nor in any edition, except Jīv., 1896, and Avinās Chandra. The interpolation is clearly based on a remark in AS., vol. ii, p. 89Google Scholar, 1. 14, where the reader is referred to the rāja-yakṣman chapter for the ṣaṭpala formula. There the remark is justified, for in AS. the rāja-yakṣman chapter comes before the gulma chapter, and the reference, therefore, is backward, from p. 89 to p. 38, where the ṣaṭpala formula is given with that very name. Moreover, AS. does not give the pañcakola formula in its chapter on gulma. In fact, verse 64a is a very stupid and apparently modern interpolation. Also verses 82–6, on p. 490, are in all probability not genuine. They are, it is true, found in some MSS. (e.g., Ind. Off. 359, fl. 224a; Tüb. 458, fl. 425a), and are admitted in the editions of Gangādhar, the two Sen, and Abinās Chandra. But they are omitted in some of the oldest and best MSS. (e.g., old Nepal MS., fl. 244a; Ind. Off. 335, fl. 280b; Tüb. 459, fl. 54a; Decc. 925, fl. 226b; also in Jīv., lst ed., 1877, p. 515)Google Scholar, as well as by Chakrapāṇidatta, who, in his Cikitsā Saṁgraha (p. 339)Google Scholar, quotes the whole passage, verses 81–91, but omits verses 82–6. His commentator, Śiva Dāsa, however, refers to them, so that we may conclude that he had them in his text of Charaka, while they were wanting in the text used by Chakrapāṇidatta. Seeing that they are substantially identical with verses 75–80, the balance of probability is for their being spurious, though a comparatively early interpolation.

^{page 1013 note 1} It may be added that Vāgbhaṭa II, in his Aṣṭāṅga Hṛdaya (ch. xiv, vv. 19, 122–9)Google Scholar, again quotes Dṛiḥhabala's verses in a slightly modified form. The prose statement of Vāgbhaṭa I appears to be based in part on Suśruta's verses (SS., Utt. Sth. xlii, vv. 119, 120, p. 805).Google Scholar

^{page 1019 note 1} There can be no doubt that the Nidāna Sthāna, equally with the Cikitsita Sthāna, was left incomplete by Charaka; but whatever chapters they contained would be expected to have run in the same order.

^{page 1019 note 2} The Kalpa Sthāna of Vāgbhaṭa I corresponds to the Siddhi Sthāna of Dṛiḥhabala. In the Aṣṭānga Hṛdaya of Vāgbhaṭa II, it is called Kalpa Sthāna, or Kalpa-Siddhi Sthāna (AH. i, 1, v. 43; iv, 14, v. 99b, in vol. i, p. 30; ii, p. 265), the latter term witnessing to Vagbhata II's acquaintance with Dṛiḥhabala's edition of the Caraka Saṁhitā.

^{page 1021 note 1} This is practically the same as the modern division of the abdomen, as shown, e.g., in the diagram on p. 733 of DrGerrish's, Textbook of Anatomy (2nd ed.).Google Scholar

^{page 1022 note 1} The existing text is hṛdi vastau pānvayor = nābhyāṁ vā sa (gulmaḥ) sūlam= apajanayati. The original text probably was jaṭhare pakvāśaye ūrdhva-nābhyāṁ vā, etc.

^{page 1022 note 2} The equalization of the number is probably only due to the Indian scholastic love of symmetry.

^{page 1025 note 1} And, I may add, partly quoted by the later Nidāna of Mādhava (AS., vol. i, p. 296, 11. 8, 9 = MN., p. 174, v. 4).Google Scholar

^{page 1027 note 1} For another similar reference see ante, p. 1014.Google Scholar

^{page 1028 note 1} Brackets indicate modified passages.

^{page 1028 note 2} Verse 64a is certainly, and verses 82–6 possibly, spurious.

^{page 1028 note 3} Verses 98, 99, 128, 157, 159, are interpolated by Dṛiḥhabala; also possibly verses 133b–136.

^{page 1028 note 4} These four are connecting verses (see p. 1011), two of which, 108a and 132a, are quoted by Vāgbhaṭa I at the end of his chapter xvi.