^{page 157 note 1} He considers these paintings as “representing phases of Indian life from 1800 to 2000 years ago” (p. 72). Their real date is, however, as we shall presently see, from six to eight centuries more modern, which makes all the difference as to the persons represented. In my History of Indian Architecture, p. 158, I stated that Cave No. I. belonged to the first half of the seventh century; but the Babu is not in the habit of deferring to the opinions of others in matters of this sort, and is consequently liable to make mistakes he might otherwise avoid.

^{page 159 note 1} J.R.A.S. Vol. IV. N.S. pp. 83, 86.Google Scholar

^{page 160 note 1} See Burgess, Reports, vol. iii. pl. xl. p. 60.Google Scholar

^{page 160 note 2} see Burgess, Reports, vol. i. pl. xxiv. to xxx. p. 24.Google Scholar

^{page 161 note 1} From the peculiarity of the plan of cave No. III. at Aurungabad, with its three chapels in antis, which is not found anywhere at Ajanta, and looks like an innovation, I should be inclined to believe it more modern than No. I. of the latter series.

^{page 162 note 1} See Lithograph, pl. v. fig. i. accompanying Babu Rajendra's paper, J.A.S.B. for 1878.

^{page 162 note 2} I have been informed that, in the Salle d'Apollon in the Louvre at Paris, there is an onyx cup, on which is engraved the portrait of a king, acknowledged to be that of Chosroes II., seated in the same attitude, and holding a cup in the right hand, as represented in this picture. I have not myself seen the cup, and have been unable to procure or gain access to any engraving of it; and am consequently unable to say how far it can be quoted as confirming the views expressed in this paper.

^{page 163 note 1} Antiqua, Ariana, p. 400, pl. xvii.Google ScholarPrinsep, Thomas's, vol. i. p. 402et seqq., Vol. ii. p. 114.Google Scholar

^{page 163 note 2} When I was writing my History of Indian Architecture, I saw Mr. Griffiths' paintings at South Kensington; but, as there were then no labels upon them, I assumed, erroneously, that they were copied from the roof of cave No. XVII., which is considerably earlier than No. 1., and consequently jumped to the conclusion—in a note, p. 157—that the king must be Baharam Gaur, whose visit to India and marriage with an Indian Princess is celebrated in Indian and Persian romances.

^{page 163 note 3} Texier, Armenie et la Perse, vol. ii. pl. 133.Google ScholarKer Porter, vol. i. pl. 19.Google Scholar

^{page 163 note 4} Flandin et Coste, Perse Ancienne, vol. i. p. 9.Google ScholarKer Porter, vol. ii. pl. 62.Google Scholar

^{page 164 note 1} Ker Porter, vol. ii. pl. 64.Google Scholar Whenever they represent the same person or thing, his delineations are to be preferred to either those of Texier or of Flandin and Coste.

^{page 164 note 2} In the bas-reliefs at Shapur, representing the defeat of Valerian, the King Shapur holds his son Hormuzd by the right hand, and he is represented with a head-dress nearly similar to that worn by the King in this picture. Unfortunately Flandin and Coste (Voyage en Perse, vol. i. pl. 49) mistake him for a Roman captive and put manacles on his feet, but both Texier (vol. ii. pl. 146) and Ker Porter represent him without these adornments, and with the crimped streamers of the Sassanians.Google Scholar

^{page 165 note 1} Tabari, , translated by Zotenberg, , vol. ii. p. 331.Google Scholar All the incidents accompanying the dethronement and death of Khosru II. will be found narrated in Zotenberg's translation from the Persian recension of Tabari, , vol. ii. pp. 328,Google Scholaret seqq., though in less detail than in the Arabic version, and without the king's name.

^{page 166 note 1} The name of the King in the codex is By another author, Dinawari, it is written , and with the diacritical points QRMISA. The diacritical points, however, when used in writing foreign names, have no authority. We may, therefore, assume PhRMIShA is intended, or, as the Arabs use Ph for the Persian P, and I for the Persian Ê, we must write his name PRMÊShA. At the same time as R and L are written with the same sign in Pehlvi, so is R to be taken as a false mode of expressing L. As M may be substituted for K (Q) in the Arabic, or in the Pehlvi, it follows that the name may be correctly represented in Pehlvi by and read as Pulikêša. (Nöldeke.)

^{page 167 note 1} Malcolm, , History of Persia, vol. ii. p. 602.Google Scholar

^{page 167 note 2} Burgess, vol. iii. of Archæological Report of “Western India, p. 132. Inscription translated by J. F. Fleet, B.C.S.

^{page 168 note 1} Mémoires des Contrées Occidentales, vol. ii. p. 151.Google Scholar

^{page 168 note 2} When Dr. Bird visited Ajanta, he is said to have cut out some twenty or thirty of the most interesting faces from these frescoes, with the intention of presenting them to the Museum at Bombay. But, as might be expected, they were so carelessly packed, that all except one or two were reduced to powder before reaching their destination. Whether this was one of these, or whether it was destroyed by some other British barbarian, cannot now be ascertained.

^{page 170 note 1} It is rather a curious coincidence that six years ago I should have been asked by Dr. Tristram to restore the design of the façade of a palace of this same king, some fragments of which he had found to the eastward of the Dead Sea, during a journey which resulted in the publication of his book entitled The Land of Moab (Murray, 1873). To it the reader is referred for some further particulars regarding Ferhad and Shirîn, though full details are only to he found iu works especially devoted to Persian history.