^{page 29 note 1} Reprinted as a separate volume, Calcutta, 1882. An English translation of Lassen's full notice of Ctesias is given in both places.

^{page 29 note 2} Dr. Jacoby kindly read through this paper in draft, and his criticisms have, I hope, freed it from statements to which classical scholars would take objection. I am also indebted to him for most of the references to classical literature. I have also greatly profited by discussion with Dr. Maas.

^{page 29 note 3} The tree here, whose fruit hangs in bunches like grapes and looks like the Pontic nut, may well be, as suggested to me by Professor Champion of the Imperial Forestry Institute, the Soap Nut, Sapindus Mukorossi. That in McCrindle's i, 21, must be the ḍhāk or palās, Butea Frondosa, and the tree in i, 22, whose fruit is dried and exported for sale, is probably the ber plum, Zizyphu Jujuba, or less probably the apricot, which may not have been known then in the Himalayas and which as a Caucasian tree should have been familiar to the Persians.

^{page 30 note 1} Photios puts it, more accurately perhaps except for the omission of the important epithet “great” of the trees, “in the mountains there are trees overhanging the river.”

^{page 30 note 2} γλνκύ δύ in Photios, but γλυκύ only in Psellos.

^{page 30 note 3} Pauly-Wissowa, , Beal-Encyclopädie, ix, 329–330Google Scholar, s. Hypobarus. In this interesting article Kiessling takes the view, which I share, that Ctesias has mixed up several different trees.

^{page 30 note 4} For a parallel with Ctesias's wording note Watt's quotation, iii, 443, from Flückiger and Hanbury, describing the shīrkhisht of the bazaars as “in irregular roundish tears, from about ¼ inch up to ¾ inch in greatest length”.

^{page 32 note 1} Psellos's ê may be due to the confusion of i and ê, a widespread phenomenon; but this depends on whether the MSS. of Psellos show other examples of the same confusion.

^{page 32 note 2} Or possibly sitachôra, but I should expect Indian ā before r and l to be transliterated by o rather than by ô.

^{page 33 note 1} Glotta (1936), 24, 220.

^{page 34 note 1} JRAS., 1941, 216.

^{page 34 note 2} Kṣīrataraṅgiṇī (ed. , Liebich), 135Google Scholar.

^{page 34 note 3} I have been unable to find any authority for this in Sanskrit literature. The statements seem to go back to Lassen's, description, Indische Altertumskunde, i, 49Google Scholar (63 in second ed.), taken from modem sources, of the three streams in the Himalayas whose confluence forms the Ganges. The most holy of them, and the one usually known as the Ganges, is the Bhāgīrathī, the other two being the Jāhnavī and the Alakānanda.

^{page 35 note 1} Cf. the Tibetan translation of Bhagīratha, skal-ldan śiṅ-rta.