^{page 627 note 1} Nachrichten v. d. kgl. Gesellschaft d. Wissenschaften zu Göttingen, Phil.-hist. Klasse, pp. 427–41, 1911.Google Scholar

^{page 629 note 1} The title jabgou was in later times one of those held by the head of the Western Turks. The title kujūla was, perhaps, originally Śaka.

^{page 629 note 2} Franke, O., Beiträge aus chinesischen Quellen zur Kenntnis d. Türkvölker u. Skythen Zentralasiens (Abhandlungen of the Berlin Academy, 1904), p. 66.Google Scholar

^{page 630 note 1} The Kharapallāna of the Sārnāth inscription (ed. Vogel in Epigraphia Indica, viii, 173 sqq.Google Scholar) will perhaps be identical with the ΧΑΡΟΒΑΛΑΝΟ named on two seals (Cunningham, , Coins of the Kushans, iii, 58).Google Scholar

^{page 631 note 1} JRAS. 1906, pp. 181–216, 460–4.Google Scholar

^{page 631 note 2} Op. cit., pp. 21 sqq., esp. pp. 44–5; see, however, infra, p. 639, n. 1.

^{page 631 note 3} Asiatic Quarterly, vol. xxx, pp. 109–10, 1910Google Scholar; cf. Franke, , op. cit., p. 45.Google Scholar

^{page 632 note 1} The first part of the name will be = Pahlavī kad, “chief,” or καδ(ονίας)?; the second part is the pises or pes (= Skt. peśa, Zend paesa) of Spargapises, Sapadbizes, Porpes (cf. JRAS. 1906, p. 210Google Scholar, and for the aspiration, ibid., p. 206); the same aspiration occurs in this king's Khusanasya and ΧΟΡΑΝСΥ—Could the CY be a genitive suffix = tse of Unknown Language I? A genitive value has already been suggested by Professor Rapson for the ΣΥ of the coin of Hermæus (JRAS. 1897, p. 321).

^{page 632 note 2} A decadent period is suggested by Spooner, D. B., Archœological Survey of India:, Annual Report, 1908–1909, p. 50.Google Scholar

^{page 633 note 1} For a list see Mr. Kennedy's note, JRAS. 1912, p. 686 and reff.Google Scholar

^{page 633 note 2} JRAS. 1903 (pp. 333–4)Google Scholar, 1905 (pp. 223–36, 357–8), 1906 (pp. 706–11, 979–92), 1907 (pp. 169–72, 1013–40, 1041–9), 1908 (pp. 55–62, 177–86), 1910 (pp. 818–24, 1315–17), 1913 (pp. 95–107).

^{page 634 note 1} The Wardak inscription (naming Huviṣka and the year 51) comes from a tope found in a place Kohwāt, 30 miles west of Kabul (see Mr. Pargiber's edition in Epigraphia Indica, xi, pp. 202–19).Google Scholar

^{page 635 note 1} As proposed by DrFleet, , JRAS. 1903, pp. 333–4Google Scholar. The pilgrims' route (via the Swat Valley), with its “hanging bridges” (Beal, , Buddhist Records, i, pp. xxx and xciiiGoogle Scholar), was unsuitable for, and has not to my knowledge been used by, an army.

^{page 635 note 2} Coins of the Indo-Scythians, pp. 2–3Google Scholar. For a Chinese view of the impracticability of this route see Lévi, M.'s article in Journal Asiatiqtie, ix, ix, pp. 7–8 and ref.Google Scholar

^{page 635 note 3} The title Mahārāja was at this time a high one, not assumed even by Great Satraps or King Satraps.

^{page 635 note 4} On this era and its designations see the articles by DrFleet, , Indian Antiquary, x, pp. 208–15Google Scholar, JRAS. 1910, pp. 818–24Google Scholar, and by Kielhorn, , Indian Antiquary, xxvi, pp. 146–53.Google Scholar

^{page 636 note 1} JRAS. 1912, p. 982, n. 1.Google Scholar

^{page 636 note 2} Archæological Survey of India: Annual Report, 1903–1904, pp. 244–60.Google Scholar

^{page 636 note 3} The reading saṃba(d)dhae, originally proposed by M. Senart (Journ. Asiat., sér. viii, vol. xv, pp. 116–17Google Scholar) and retained by M. Boyer (ibid., sér. x, vol. iii, p. 459), would mean rather “connected with”. The proposed reading saṃbatśarae, which seems quite certain in M. Senart's plate as compared with a rubbing kindly supplied by Professor Konow, agrees with that of the Āra inscription, edited by Mr. R. D. Banerji (Ind. Ant., 1908, pp. 58–9Google Scholar) and Professor Lüders (Sitzungsberichte of the Berlin Academy, 1912, pp. 824 sqq.). The b in place of v, which was the cause of difficulty to M. Senart and some earlier scholars, is, therefore, to be accepted. In 11. 3–4 of the inscription we ought probably to read the same word Mira(sa ?) boyanasa.

^{page 636 note 4} This caution is suggested by the fact that at the present day about twenty eras are employed in India. The Parthian era (beginning 249 B.C.) was not very widely separated from the Seleucid (312 B.C.).

^{page 637 note 1} M. Lévi in his important articles (Journal Asiatique, IX, viii, pp. 444–84Google Scholar; ix, pp. 1–42; x, pp. 526–31) had previously discussed this and other points (esp. ix, pp. 14–26).

^{page 638 note 1} See Wroth, , Catalogue of the Coins of Parthia, pp. 241 sqq.Google Scholar

^{page 638 note 2} Huviṣka has hΡΑιΛΟ = ΗΡΑΚΛΗС.

^{page 639 note 1} If Professor Konow's theory is (contrary to expectation) not maintained, we shall perhaps be forced to regard NANO as a genitive suffix, perhaps akin to Turkish niṅ! In that case the Kaniṣka group will after all be Turkish: cf. M. Lévi's nate in the Journal Asiatique, ix, ix, pp. 10–11.Google Scholar

^{page 639 note 2} I incline to the view of Cunningham (Coins of Ancient India, pp. 31–7Google Scholar) that the Kharoṣṭhī was originally a Gandhāra (and Arachosia)

^{page 640 note 1} In many, or most cases, the normal forms also are found on coins of the kings who employ the innovations. The dates are the initial dates of the Parthian kings first employing the signs in question. Apology is due for the imperfections of the table, partly due to uncertainty or divergencies in the readings of the numismatists, who often find a difficulty in distinguishing, for instance, and . Concerning see p. 642 and n. 3. For and (and also for some of the other forms) see the tables of alphabets in SirThompson, E. M.'s Introduction to Greek and Latin Palœography (Oxford, 1912)Google Scholar. alphabet, in which case its name may be derived from Arachosia (Zend Hara(h)uvatī, Haraxvaitī), But see the articles of Professor Sylvain Lévi (Bulletin de l'École Française d'Extrême Orient, ii, pp. 246–53Google Scholar; iv, pp. 543–79), with whom we should certainly agree that the Sanskrit kharoṣṭrī involves a “popular etymology”.

^{page 641 note 1} The Stein documents seem to contain some phrases, etc., in a Central Asian language written in this alphabet.

^{page 641 note 2} The Greeks having been in Bactria at least 165 years (330–165) before the first barbarian irruption.

^{page 642 note 1} Of course, I do not dispute the importance of the determination of the h-value of the sign.

^{page 642 note 2} See the Ind. Ant. 1884, p. 58Google Scholar (Burgess), and as regards Cunningham his Coins of the Indo-Scyths, i, p. 5 and reff.Google Scholar

^{page 642 note 3} The tsade of the Nabathean Aramaic is exactly like the Kushan sign (see Encyclopœdia of Islam, i, pi. i).

^{page 642 note 4} The theory is that adumbrated by Cunningham, , Coins of the Indo-Scythians, iii, p. 16.Google Scholar

^{page 643 note 1} See Roberts, Introduction to Greek Epigraphy, p. 10Google Scholar, n. 1, and Cauer, , Delectus Inscriptionum Grœcarum, No. 480 n.Google Scholar

^{page 643 note 2} Pahlavī; see Euting, 's Drei Tafeln des Pehlevi- u. Zend-Alphabets (Strassburg, 1878)Google Scholar and Prinsep, 's Essays on Indian Antiquities (ed. Thomas), vol. ii, pis. xi–xib.Google Scholar

^{page 643 note 3} Mr. Kennedy had made his theory known to me before this publication.

^{page 643 note 4} Ed. Shama Sastri (Mysore, 1909), p. 81, 11. 3–4.

^{page 644 note 1} The weight of the coins, even if determined by an arithmetical ratio, may have been approximately based upon that of the Roman aureus.

^{page 645 note 1} Cunningham, , Archæological Survey Report, xiv, 116Google Scholar, cf. 136. Coins of Zoilus, Gondophares, Kaniṣka, and Huviṣka are numerous at Pathankot.

Cunningham, , Archæological Survey Report, xx, 37Google Scholar. Wima Kadphises, Kaniṣka, Huviṣka, and Vāsudeva are found at Mathurā.

Cunningham, , Archæological Survey Report, xxii, 108Google Scholar. Immense numbers of coins of Kadphises, Huviṣka, and Kaniṣka found annually in Khaira Dih.

Cunningham, , Archæological Survey Report, xiv, 48Google Scholar. “Pot full of large copper coins found at Ransi a few years ago” (i.e. before 1878); seem to have been about 500 Wima Kadphises and Kaniṣka.

Cunningham, , Archæological Survey Report, xii, 206Google Scholar. Large Æ of Wima Kadphises and Kaniṣka are found around Bua Dih.

Cunningham, , Archæological Survey Report, xii, 43Google Scholar. Find of about twenty copper coins, Wima Kadphises, Kaniṣka, Huviṣka, Vāsudeva, in Indore.

Cunningham, , Archæological Survey Report, xi, 97Google Scholar. Gold coins of Kaniṣka and Huviṣka found with gold of Domitian, Trajan, and Sabina (Ahin Posh hoard).

Cunningham, , Archæological Survey Report, xi, 25Google Scholar. Coins of Wima Kadphises, Kaniṣka, Huviṣka, Vāsudeva, also later ones, numerous around Sankisa.

Cunningham, , Archæological Survey Report, v, 176Google Scholar. Two of Wima Kadphises, one of Kaniṣka, and one of Vāsudeva at Kanhiara.

Cunningham, , Archæological Survey Report, ii, 162Google Scholar. Mānikyāla tope No. 2, 8 Æ of Kujūla, Wima Kadphises, and Kaniṣka, 4 N Kaniṣka, 7 . Roman denarii (latest of M. Antony, 43 B.C. ). Proc. A.S.B. 1895, p. 82. Find of 382 copper of Kadphises II, with forty copper Kaniṣka, on the Kalka-Kasaulī Road in Patiala.

JASB. 1881, p. 184Google Scholar. Find of c. 1, 000 coins of Kadphises, Kaniṣka, Huviṣka, and Vāsudeva at Peshawar.

^{page 646 note 1} SeeMaharājakanikalekha, vv. 1, 6,75, 80 (Ind. Ant. 1903, pp. 345 sqq.).Google Scholar

^{page 646 note 2} Bruchstücke Buddhistischer Dramen (Königlich Turfan-Expedition: kleinere Sanskrit-Texte, Heft i), p. 11.Google Scholar

^{page 646 note 3} See his articles, T'oung Pao, 1904 (pp. 10 and 12)Google Scholar, and JRAS. 1905, pp. 52–3.Google Scholar

^{page 648 note 1} See Beal, , Buddhist Records of the Western World, i, pp. 13Google Scholar sqq., and cf. Lévi, M., Journal Asiatique, ix, ix, 24 n.Google Scholar

^{page 648 note 2} Mahārājakanikalekha, v. 47Google Scholar (Ind. Ant. 1903, p. 356).Google Scholar

^{page 649 note 1} Hiuen Thsang apud Beal, , Buddhist Records of the Western World, i, pp. 99, 151Google Scholar. Elsewhere Hiuen Thsang himself gives 500 (see Watters, , On Yuan Chwang, i, p. 224Google Scholar). Sung-yun says 300 years (Beal, , op. cit., p. ciiiGoogle Scholar). The Samyukta-ratna-piṭaka (apud Lévi, Journ. Asiat., sér. ix, vol. viii, p. 463Google Scholar) says 700.

^{page 649 note 2} Ibid., p. 150.

^{page 649 note 3} In the immediate future, and especially in connexion with Taxilā, the argument from archæological stratification is likely to be very important. At present we have the two definite points where the Kaniṣka group is certainly the later, namely, (1) Kabul, where Kadphises I is contemporary with the last Greek ruler and therefore prior to the rule of Huviṣka (v. supra, p. 635), and (2) Mathurā, where the Śakas employ a Kharoṣṭhī earlier than that of the Kushan inscriptions. And these two points are not far from the extreme west and east, respectively, of the empire of Kaniṣka.