^{page 91 note 1} I quote Dr. Fleet's translation of No. 33, pp. 147–8.

^{page 94 note 1} On such grounds one might as well question the achievements of Napoleon, or of Alexander.

^{page 96 note 1} It may be read in further detail in Beal's, Buddhist Records of the Western World (Siyuki), vol. i, pp. 167–72.Google Scholar

^{page 99 note 1} Also in the Indian Empire of the Imperial Gazetteer, vol. ii, p. 4, n. 2.Google Scholar

^{page 100 note 1} The meaning of vikrama-kāla is rediscussed by Kielhorn in GN., 1900, p. 353 ff.,Google Scholar where the passage in question, together with its translation by Cowell and Mr. Thomas, is quoted. I am disposed to accept Kielhorn's emendation vikrama-kāa-pītam akāle'pi, but the reading as it stands, vikrama-kālam akāle'pi, after all, yields precisely the same meaning: “the terrible visage (of the old warrior), brightened by a thick white mustache which hid his cheeks, seemed to pour forth his vikrama-kāla, or time of (youthful) vigour, even out of season, in the shape of the commencing autumn, white with blooming kuśa grass.” That is to say, the exhibition of youthful vigour in old age was out of season, i.e. was not what one could normally expect. However, even if vikrama-kāla, which properly means “vigour-time” be, for the purpose of this particular passage, made to mean “war-time”, the latter constructive meaning, restricted to a particular passage, could not reasonably be supposed to have resulted in the naming of a widely current era.

^{page 100 note 2} This name or title occurs in both forms, e.g. on coins of Chandragupta II; see MrSmith, , Gupta Coinage., in JRAS., 1889.Google Scholar

^{page 102 note 1} Perhaps there is an allusion to the title in the phrase vikram-āvakraya-krīta in the Udayagiri cave inscription, F.GI., p. 35.

^{page 103 note 1} This, in fact, is essentially the theory, proposed long ago (1870), though on insufficient and partly misunderstood evidence, by Fergusson, , in JRAS., N.S., iv, p. 131 ff.Google Scholar

^{page 107 note 1} Cunningham calls him Pratāpāditya I. See Dr. Stein's remark in RT., i, 8.

^{page 108 note 1} In JRAS., 1893, p. 127,Google Scholar n. 1, I am sternly admonished that “the reading Pura is certainly untenable, and must be given up”. The last reference (known to me) to the Sthiragupta myth occurs in 1902, in Jbbras., xx, 2.Google Scholar The propagators of this myth can hardly have taken the precaution to consult the original of the seal or any of its many facsimiles.

^{page 110 note 1} On = Sindhu, see DrStein's, Ancient Geography of Kashmir, pp. 91, 107Google Scholar (in JASB., lxviii, Extra No. ii, 1899Google Scholar), also RT., ii, 335.Google Scholar

^{page 110 note 2} In S.RT., i, 10Google Scholar (v. 43), Kashmir is called “the region of Kubera”.

^{page 110 note 3} Probably one of his earliest compositions; see MrChakravarti, in JRAS., 1903, p. 186,Google Scholar and 1904, p. 161; also Professor Liebich in L.AR., pp. 3–6.Google Scholar

^{page 111 note 1} Some other coincidences are also adduced by Kielhorn in EI, vi, 3. 4. But the dates of these, even if true, do not militate against my determination of Kālidāsa's date.

^{page 111 note 2} The four ideas are: (1) that the winter is agreeable to men because it confines them to warmed rooms in the company of lovely women; (2) it is a season in which one does not wish for cool places or cooling appliances. These two are common to the two poets. The other two are: (3) in the winter the fishes lie low down in the water, and (4) the water-lilies are bitten by frost. How the matter of borrowing by Indian poets strikes an Indian mind may be seen from Mr. Chakravarti's remarks in this Journal, 1904, p. 161.Google Scholar

^{page 112 note 1} Regarding Bhartṛihāri see Kielhorn's, remarks in IA., 1883, xii, 218 ff.Google Scholar No more than three grammatical works are ascribed to him: (1) the commentary on. Mahābhāṣya; (2) the Vākyapadīya, and (3) the Prakīrṇaka. These are precisely the three works which alone are ascribed to him by Itsing (TI, pp. 178–80Google Scholar), who visited India forty years after Bhartṛihāri's death (TI., p. xxv;Google ScholarL.AR., p. 7Google Scholar). There is, therefore, no good reason for attributing the Bhaṭṭikāvya to Bhartṛihāri.

^{page 114 note 1} In T.KS., ii, p. 586,Google Scholar the Persian king is called Nirmūka. See below p. 120.

^{page 117 note 1} This tradition ignores the story of Mihirakula's liberation and instalment as king of Kashmir.

^{page 118 note 1} “Especially by astronomers,” as Alberuni says. The reference here is to Bhaṭṭotpala, who, as Bhau Dāji explains (JRAS., 1865, i, 407Google Scholar) in his commentary on the Bṛhat Saṁhitā, quotes a verse from a Tantra (Pañcasiddhāntikā ?) which refers Vārāha Mihira to 427 of the śāka-kāla (=505 A.D.), and explains that term to mean “the time when the śaka kings were conquered by Vikramāditya”.

^{page 119 note 1} Also Professor Jacobi in JGOS., xxxiv, 278 ff. and 253 ff.,Google Scholar where some other chronological speculations are given.

^{page 120 note 1} In IA., xx, 347,Google Scholar the emendation paṇa-rasa for rasa-paṇa is wrong. The latter is the correct reading, as in IA., xxi, 67,Google Scholar for rasa is 6, and paṇa is 50. On the other hand, jajjaṁ (or gajjaṁ) in IA., xx, 347,Google Scholar should be rajjaṁ

^{page 121 note 1} Chandragupta II, who calls himself Vikramāditya on his coins, is here out of the question, for he was the father of Mahendrāditya Kumāragupta II.

^{page 121 note 2} Curiously enough, Kālidāsa's description is followed by Somadeva even more closely (T.KS., i, 52, 150 ff.Google Scholar) in another digvijaya, ascribed to a king of Vatsa, a country with the capital Kauśambhī. The agreement here is so striking as to suggest Somadeva imitating Kālidāsa.

^{page 121 note 3} The reference to Cunningham, (ASR., i, 280Google Scholar) by Bhagwanlal Indraji, I think, is wrong. Cunningham says nothing about any conquest of Nepal by Harsha Vardhana. Professor S. Lévi's ingenious explanation of the term tukhāra (Turk), against Bühler's reading tuṣāra (snowy), has been adopted by Mr. Ettinghausen (Harṣa Vardhana, pp. 46, 47Google Scholar) without acknowledgment.

^{page 122 note 1} The question of the exact identity of Mo-la-po has no real bearing on that of the identity of śīlāditya. There is nothing to show, nor does Hiuen Tsang say so, that he inspected the local records himself. He only reports what the people told him that they contained.

^{page 122 note 2} The statement on this subject in S.EH., p. 306,Google Scholar is somewhat misleading. See preceding note.

^{page 126 note 1} I adopt here and elsewhere the words of Mr. Smith to mark the points of agreement.

^{page 129 note 1} In the Deo-Baraṇārk inscription Narasimha is mentioned as the ruler only of madhyadeśa, or the central portion of Northern India (JASB., lviii, 15Google Scholar).

^{page 131 note 1} This explanation was first suggested by DrFleet, in JA., xv, 187,Google Scholar n. 3. Droṇasiṁha is described as Mahārāja already in his grant of 502 A.D. (JBBBAS., xx, p. 4Google Scholar). That year falls within the period of Toramāṇa's Hunic domination, and it may be that Droṇasiṁha attempted to profit by the disturbance of that time to assume the higher title; but it received its formal sanction only in 525 A.D. from Yaśodharman as emperor. The genuineness of the grant, however, is doubted by Kielhorn in EI., viii, App. i, p. 2, n. 2.Google Scholar The treatment meted out to the captive Mihiragula is graphically described in Yaśodharman's column inscription (F.GI., p. 148, 1. 6Google Scholar). He was compelled to fall down, and touch with his head, the hair-lock of which was adorned with flowers, the feet of the victor. He was apparently forced to pay the same humble obeisance to Yaśodharman's ally, Droṇasiṁha, as implied in 1. 4 (cf. with 1. 3) of the Valabhī inscription (F.GI., p. 168Google Scholar). From this it is difficult to say how much, if any, credit is to be given to the story of Mihiragula's liberation and restoration to Kashmir, which was related to Hiuen Tsang, and reported to us by him; especially in the face of the very different story of the same events in the Kashmir Chronicle,

^{page 134 note 1} In Cat. IM., p. 121,Google Scholar n. 1, it is rejected, without assigning any reasons.

^{page 134 note 2} Regarding the common Indian practice of shortening (made bhīmavat) compounds, see the discussion in JGOS., xxxix, 707Google Scholar; xl, 145; xliv, 481 ff.; VOJ., viii, 240 ff.Google Scholar I owe these references to the kindness of Mr. A. B. Keith. Thus, explaining the name of a medical formula, Aruṇadatta in his commentary on Aṣṭānga Hṛdaya, iv, 14,Google Scholar v. 36, says: tathā śārdūla-śabdena śārdula-vikrīḍitaṁ vṛttam upalakṣyate, yathā bhīma-śabdena bhīmasena iti, i.e. the śārdūla-vikrīḍita metre is indicated by the word śārdūla, just as bhīmasena by the word bhīma.

^{page 135 note 1} Nor Skandagupta, if he be really identical with Puragupta. As Puragupta, i.e. having lost half his dominions (ante, p. 129), Skandagupta surely would not have committed the sarcasm against himself of claiming to have subdued the whole earth.

^{page 136 note 1} The copperplate grants of Dharmāditya, which in Ind. Ant., 1892, xxi, 45,Google Scholar I suggested to be referable to Samūdragupta, I would now prefer to attribute to Yaśodharman. This attribution removes the difficulty of the Lakshmī seal (see JASB., lxiii, 166Google Scholar). There are now two of these Dharmāditya plates, which Mr. Pargiter is at present engaged in preparing for publication.

^{page 136 note 2} See an instructive observation of Dr. Fleet on the curious usages of titles, or birudas, in Ep. Ind., vi, 193.Google Scholar Thus Govinda III, Rashṭrakūṭa, was remembered only as Jagattunga.

^{page 138 note 1} See śivarāma Tripāṭhin's remarks on H.V., p. 7Google Scholar; also Professor Jacobi in VOJ., iii, 143, 144Google Scholar; also F. Hall himself in the Introduction, p. 6. The ridicule, as unmeasured as it is unmerited, poured by the latter on his dissentients, needs no notice; it only marks the era of the then prevailing rampant scepticism regarding traditions.

^{page 139 note 1} Professor S. Lévi makes it 500–50 (Bull. 1903, p. 49;Google ScholarJRAS., 1904, p. 44Google Scholar). This suits my argument even better; but I prefer to base it on the more extreme (though, indeed to my mind, more probable) date of Dr. Takakusu.

^{page 140 note 1} I owe these references to the kindness of Mr. Pargiter.

^{page 142 note 1} See F.GI., p. 154, 1. 21Google Scholar, Mālava-gaṇa-sthiti-vaśāt, i.e. since the first constitution of the Mālava tribes. Vaśa means here birth, origin; see M. Williams' Dictionary.

^{page 143 note 1} Verse 489 of the Vākyapadīya (in the Benares Sanskrit Series) runs as follows: (vyākaraṇ-āgamaḥ |) sa nīto bahu-śākhatvaṁ Candrācāryādibhih punaḥ ‖, i.e., the study of grammar was again brought to a high state of flourish by Chandrāchārya and others. Puṇyarāja explains Candrācāryādi by Candrācārya-Vasurāta-guru-prabhṛti, i.e. Chandrāchārya, Vasurāta, and other teachers. I do not think that Puṇyarāja's explanation necessarily implies that Vasurāta was a younger pupil of Chandragomin, as Professor Liebich (L.AR., p. 9Google Scholar) and Dr. Takakusu (JRAS., 1905, p. 45Google Scholar) assume: both may very well have been of the same age, and Vasurāta may have been only inspired by Chandragomin to pursue the same line of study. But there is a curious chronological difficulty in the matter. Bhartṛihari's language in verse 489 certainly seems to suggest, as Professor Liebich takes it, that he himself was a direct pupil of Vasurāta. But as, according to Itsing (T.I., p. 150Google Scholar), Bhartṛihari died about 650 A.D., and as the date of Vasurāta is about 470–530, or at latest 520–600 (calculating from Dr. Takakusu's or Professor S. Lévi's data respectively), it seems quite impossible that Bhartṛihari could have been a direct pupil of Vasurāta. It seems necessary, therefore, to assume (with DrTakakusu, , JRAS., 1905, p. 45Google Scholar) that Bhartṛihari was really a praśiṛya, or pupil of a pupil, of Vasurāta.