^{page 914 note 1} Reports, vol. 2 (1871), p. 68, and note.Google Scholar

^{page 914 note 2} See p. 965 ff. below.

^{page 914 note 3} On this matter in full see my paper in JRAS, 1906, p. 979.Google Scholar

^{page 914 note 4} For references see JRAS, 1906, p. 883, and note 1.Google Scholar

^{page 914 note 5} For references see JRAS, 1906, p. 982, note 2.Google Scholar

^{page 914 note 6} See JRAS, 1906, pp. 982–3.Google Scholar

^{page 915 note 1} For Hiuen-tsiang's words see JRAS, 1906, p. 979.Google Scholar

^{page 915 note 2} Si-yu-ki, vol. 1, p. 151Google Scholar, note 97.

^{page 915 note 3} See my paper “The Date of the Death of Buddha” in JRAS, 1912, p. 239.Google Scholar

^{page 916 note 1} See Geiger, 's MahāvaṃsaGoogle Scholar, translation, introd., p. 46: see also ibid., p. 59 ff., regarding the confusion of Dharmāśōka and Kālāśōka.

^{page 916 note 2} See, e. g., a remark in my Indian Epigraphy, in the Imperial Gazetteer of India, vol. 2 (1908), p. 75.Google Scholar

^{page 916 note 3} See JRAS, 1912, pp. 996–1001.Google Scholar

^{page 917 note 1} On this point see fully my paper “A Point in Palæography” in JRAS, 1907, p. 1041.Google Scholar

^{page 917 note 2} See JRAS, 1907, p. 1029Google Scholar; and p. 1041 for a figuring of the word which has the h. It seemed better (and every consideration justifies the decision) to adopt the Latin h, rather than to assume a revival of the h-value of the Greek ēta, which had lost that value at some time about b.c. 350, ẉhen the ēta was cut in half vertically and developed into the rough breathing: see Taylor, , The Alphabet (1883), vol. 2, p. 86Google Scholar, quoted by me in another matter in JRAS, 1908, p. 58.Google Scholar

^{page 917 note 3} JBBRAS, vol. 22, p. 223.Google Scholar

^{page 918 note 1} While not disputing the importance of my determination of the h-value of the sign on the coins, he has severely said (p. 642) that my discovery “might deceive the very elect”. Perhaps he will expound the meaning of this cryptic utterance in his final remarks: all that seems clear just now is that those alone who believe in his teaching are to be ranked among the elect.

^{page 918 note 2} To such an extent that, as observed by a writer in JRAS, 1908, p. 551Google Scholar, “the headdress, the style of dressing the hair, the absence of moustache, and, above all, the shape of the head and features are very similar to the heads on coins of the Roman emperors of from 30 b.c. to 150 a.d., and the figures on these plates [Mr. Scott's plates 2 and 3], if examined with no previous knowledge of where they came from, might easily be mistaken for Roman coins, especially those of Alexandria bearing the Greek legends.”

^{page 918 note 3} See the sample given by me in JRAS, 1907, p. 1044.Google Scholar

^{page 918 note 4} It has been proposed to recognize a likeness of Augustus (ProfessorRapson, , Indian Coins (1897), §§ 15, 66)Google Scholar; or of Augustus or Tiberius (MrSmith, Vincent, Early History (1908), p. 238Google Scholar, and ProfessorRapson, , p. 912 above)Google Scholar; and also of Caius or Lucius, grandsons of Augustus (see JRAS, 1903, p. 30, n. 1).Google Scholar But all that can really be said is (I understand) that it is a Roman emperor's head of a style not later than about a.d. 60, but the exact original cannot be identified with any certainty.

^{page 919 note 1} These coins were struck from various dies: for figurings of the word Maasēno on two of them, see JRAS, 1907, p. 1047.Google Scholar

^{page 919 note 2} I hold that he did begin to reign in a.d. 78, and did found the Śakaera: see my remarks in JRAS, 1910, p. 820Google Scholar; 1912, p. 786; p. 992 below.

^{page 919 note 3} For these relationships see the inscriptions on the Mathurā lioncapital, edited by DrThomas, in Epi. Ind., vol. 9, A on p. 141Google Scholar, and B on p. 143: the records being in Kharōshṭhī characters, Rājūvula, mentioned in them by the contracted form of his name, is spoken of as Rajula ( = Rājla) and Śoṇḍāsa as Śuḍasa.

^{page 919 note 4} Professor Lüders’ List of the Brāhmī Inscriptions, Epi. Ind., vol. 10Google Scholar, appendix, No. 59.

^{page 921 note 1} JRAS. 1912, pp. 665 ff., 981 ff.Google Scholar

^{page 920 note 2} So also with the history of Yin-mŭh-foo. It is intimately connected with the history of the embassies to and from China, and with the question of the route followed by the traders.

^{page 922 note 1} The period of Kozoulo Kadphises' activity falls after a.d. 25. Dr. Thomas puts the date of his death c. a.d. 40. I have given reasons for dating the conquest of Kābul and the death of Kozoulo Kadphises. between a.d. 46 and 65. Mr. V. Smith puts the conquest of Kābul c. a.d. 60.

^{page 922 note 2} JHS. 1902, p. 286.Google Scholar

^{page 922 note 3} Gardner, , Coins of Greek and Scythic Kings of Bactria and India, Introd., pp. li, liii.Google Scholar

^{page 922 note 4} JRAS. 1912, p. 1014.Google Scholar

^{page 923 note 1} Letronne, , Recueil des inscr. Grecques et Lat. del'Egypte, i, pp. 99–100.Google Scholar Here are two other instances from Dittenberger, , Orientis Grœci Inscriptiones SelectœGoogle Scholar: No. 611 is from the Hauran, ὑпρ σωτηρας Αὐτοκρτορος ΤραϊανοΝρουα Σεβαστο υἱς Γερμανικο Δακικς— a strange jumble; No. 660 is from the quarries of Hammamat, near Kossair, Пοпλον Ἰουεντου Άγαθοпος пελεθερος αὐτο κα пρονοτης, etc. None of these examples is later than Trajan.

^{page 924 note 1} JRAS. 1913, p. 668.Google Scholar

^{page 924 note 2} Anyone can form an opinion by comparing the coins in the B.M., or, failing that, in Gardner's work, plates xxvi–xxviii, with facsimiles of the papyri. I find both forms of alpha, also the rounded sigma and the sigma with its upper curve prolonged, on Huvishka, 's coins, and in pl. iv, p. 42Google Scholar, of Kenyon, 's Palœography of the Greek Papyri.Google Scholar The same beta, eta, kappa, nu, and omicron may be found in both. The date of the papyrus is a.d. 15. The rounded and the wedge-shaped epsilon of Huvishka's coins reproduce the two forms in the graffito of Dōriōn, the elephant hunter, given by Weigall, , Travels in the Upper Egyptian Desert, p. 166—aGoogle Scholargraffito of presumably Ptolemaic times. For an example of the wedge-shaped epsilon on Huvishka's coins v. Ardeikhsho, Gardner, , pl. xxvii, 9.Google Scholar The rounded beauty of the early Roman hand may, I think, be recognized in Kanishka's cursive, due allowance being made for the difference of material. Like the legends of the Elamite Orodes and Phraates, Huvishka's legends are turned sometimes inwards, sometimes outwards.

^{page 925 note 1} This author places the Yavanas in the north, the Śakas in the south, the Pahlavas in the west, and the Tushāras in the east. The location of the first three in the first century b. c. is historically certain. The Yavanas were in Kābul, as Dr. Thomas admits; the Śakas in the lower Indus Valley, the Indo-Scythia of the Periplus and Ptolemy; and the Parthians in Arachosia and the west. Dr. Thomas suggests that according to Chinese notions of orientation I should have allotted the Tusharas to Eastern Turkestan, a country with which they had never any connexion. Why I should adapt Chinese notions of orientation to the interpretation of an Indian author, or why I should commit a historical absurdity, I fail to see. An Indian author who put the Yavanas in Kābul and the Śakas in the lower Indus Valley would certainly regard the Panjāb or Hindustan as the east. Nor do I know from what source Dr. Thomas has derived his notions of Chinese orientation. Ideas regarding the cardinal points of the compass differ in different parts of China. As to Aśvaghosha, he employs a mode of speech which any inhabitant of Eastern Hindustan might use at the present day in speaking of the Panjāb.

^{page 926 note 1} Dr. Thomas has a somewhat summary way of dealing with inconvenient witnesses. If Pan-ku says that there was a gold currency of Ki-pin in the first century b.c., Dr. Thomas merely contradicts him, begging the whole question. The Buddhist monk gets even shorter shrift. When Ma-tuan-lin's account of the Śakas does not agree with his (JRAS. 1906, p. 189, n. 3)Google Scholar, DrThomas, assures us that “the whole story seems incorrect”Google Scholar to him. At this rate one can build up any historical edifice one pleases

^{page 927 note 1} Fu.

^{page 927 note 2} Dr. Thomas might study the following passage with advantage: “Comme officiers, ils ont les che-hou (jab-gou), les t'ele (tegin) qui sont toujours pris parmi les fils ou les frères cadets ou les parents du Kagan” —Chavannes, , Documents sur, les Tou-kiue Occidentaux, p. 21.Google Scholar

^{page 928 note 1} Mommsen, , Provinces of the Roman Empire, Eng. trans., vol. ii, p. 12.Google Scholar

^{page 928 note 3} JRAS. 1912, p. 1002.Google Scholar

^{page 928 note 3} Dr. Thomas repeats without a qualm the old story that the Kushans got their gold supply from Rome. I have pointed out (1) that gold came from Rome only in the shape of coined money; (2) that the finds of Roman coins in the Panjāb are remarkably few; (3) that no Roman coin has yet been found re-struck by a Kushan. Roman gold in the first century of our era went chiefly to Malabar. As to Dr. Thomas' remarks regarding certain coins with legends in a North Aryan language, I frankly confess that I do not understand them. The coins in question are of copper, and therefore only meant for local use. They have a lion or a Bactrian camel on the reverse, which points, like their legends, to an origin north of the Paropamisus, and the broken Greek would seem to indicate that they were struck by subordinates of the Kushan monarchs of Bactria. As they were struck for local use, they must have been in the popular tongue.

^{page 929 note 1} Stein, , Ruins of Desert Cathay, vol. i, p. 52.Google Scholar

^{page 929 note 2} The whole Yue-che horde was settled at the time on the north bank of the Oxus.

^{page 931 note 1} Ariana Antiqua, p. 91.Google Scholar

^{page 931 note 2} Masson's report on his excavations is given in Wilson, 's Ariana Antiqua, pp. 55 ff.Google Scholar The figures of his coin-finds at Beghrām will be found in JASB. 1836, p. 547.Google Scholar

^{page 931 note 3} Okro is the old misreading of Oēsho on the Kushan coins.

^{page 931 note 4} It is of course possible that Kanishka and Huvishka, one of whose capitals was Peshāwar, may have held both ends of the Khyber Pass, including Jalālābād. Hindu princes have done so, but I see no necessity for the supposition.

^{page 932 note 1} See farther on in this number of our Journal.

^{page 933 note 1} ASI. ii, p. 64.Google Scholar

^{page 933 note 2} Ariana Antiqua, p. 364Google Scholar

^{page 935 note 1} The so-called Roman head on certain coins of Kozoulo Kadphises, on which Messrs. Rapson and V. Smith lay great stress, is undoubtedly an imitation of an early Imperial head. A very competent expert pointed out to me that the arrangement of the hair, especially at the back of the neck, is characteristic of the Roman medals before Nero. On the other hand the fillet is looped in a non-Roman fashion, and there is no likeness to any individual emperor: so my friend said. Those who consider it a likeness cannot agree as to which of four persons it resembles. The question does not enter into my argument, as we are all agreed that Kozoulo Kadphises belongs to some date between a.d. 25 and a.d. 50 or so

^{page 936 note 1} For my paper on this find see JRAS. 1913, p. 371.Google Scholar

^{page 939 note 1} “The copper coins of Kadphises and Kanerkes are found in considerable quantities in the hands of the money-changers of most of the large towns of Hindustan” (Ariana Antiqua, p. 349).Google Scholar In a supplemental list furnished me by Mr. Allan I find that, in three cases out of seven, only Kanishka and Kozoulo Kadphises appear. Am I justified in supposing that the one came straightway after the other?

^{page 939 note 2} ASR. ii, 67, 68, noteGoogle Scholar

^{page 940 note 1} Beiträge zur Türkvölker, p. 100.Google Scholar

^{page 941 note 1} Watters, , i, 224.Google Scholar

^{page 941 note 2} Stein, , trans., bk. i, v. 172.Google Scholar

^{page 941 note 3} Nachfolger Alexanders, pp. 79, 81.Google Scholar

^{page 942 note 1} Ariana Antiqua, pp. 347 seqq.Google Scholar

^{page 942 note 2} Cunningham, ASR. ii, 162.

^{page 942 note 3} Epi. Ind., vol. xi, p. 202Google Scholar; and see JRAS. 1912, pp. 1060–2.Google Scholar

^{page 945 note 1} Imperial Gazetteer of India, ii, 115, 1908.Google Scholar

^{page 945 note 2} Ann. Rep. Arch. Surv. India, 1908–9, pp. 33–4, 50.

^{page 945 note 3} London lectures, 1913. Cf. also L’art gréco-bouddhique du Gandhāra, 1905, pp. 40–2.Google Scholar

^{page 946 note 1} Foucher, , op. cit., p. 42.Google Scholar

^{page 947 note 1} [Regarding these three dates see farther on, under Mr. Fleet's remarks on the Seleucidan era: it was proposed to refer the date of the year 179 to the Śaka era of a.d. 78, not to the Seleucidan era.—Ed.]

^{page 947 note 2} History of Indian and Eastern Architecture, p. 182.Google Scholar

^{page 947 note 3} Actes Onzième Congrès International des Orientalistes, sect, i, 246, Paris, 1897.Google Scholar

^{page 948 note 1} This work, although referring to several divine Bodhisattvas by name, is held to belong to the Hīnayāna Sarvāstivādins (Kern, H., Manual of Indian Buddhism, p. 6)Google Scholar, to whom Kaniṣka's relic-casket is dedicated and to whom his “Council” was restricted.

^{page 948 note 2} The conception of Maitreya as a flask-holder, which is a common motive, as pointed out by ProfessorGrünwedel, (Buddhist Art in India, pp. 130, 188)Google Scholar, does not appear in the Hīnayāna, but in the Mahāyāna cult.

^{page 948 note 3} JRAS. 1894, pp. 53Google Scholar, etc.; my article “Evolution of the Buddhist Cult”, Asiatic Quarterly Review, 01, 1912.Google Scholar

^{page 949 note 1} In 475 a.d. (Rapson, , Indian Coins, p. 20).Google Scholar

^{page 949 note 2} Smith, V. A., JASB, 1889, p. 144 seq.Google Scholar, and IA, 1889, p. 257; Bühler, , IA, 1891, p. 394.Google Scholar For a record of this and other inscribed Gandhāra sculptures see Vogel, , Ann. Rep. Arch. Surv, India, 1903–1904, pp. 244, etc.Google Scholar; also Banerji, R. D.'s “Scythian Period of Indian History” in Ind. Ant. 1903, pp. 25Google Scholar, etc., for a useful summary and revision of the records.

^{page 949 note 3} Proc. JASB. 1898, p. 60Google Scholar; Senart, , Journ. As., 9^e xiii, 526, 1899Google Scholar; Vogel, , loc. cit., p. 245, No. 8.Google Scholar

^{page 950 note 1} Vogel, , loc. cit., p. 245, No. 12.Google Scholar

^{page 950 note 2} Bühler, , IA, 1891, p. 394Google Scholar; Journ. As., 9^e, xiii, 536Google Scholar; R. D. Banerji, loc. cit.; Smith, V., Early History of India, p. 248.Google Scholar

^{page 952 note 1} Fleet, , JRAS, 1903, p. 329.Google Scholar

^{page 952 note 2} Parker, E. H., China and Religion, p. 75Google Scholar; and other authorities.

^{page 952 note 3} Watters, , Yuan Chwany, i, 124–6, 292.Google Scholar

^{page 956 note 1} In some of the cases in which the king is shown sitting or only his head and shoulders are given, his helmet is perhaps surmounted by the head of a three-pronged spear.

^{page 959 note 1} [Compare the case of rupees, etc., of 1835 and 1840 mentioned in JRAS, 1913, p. 374, n. 4.—Ed.]

^{page 959 note 2} Coins of the Sakas, p. 2Google Scholar, = Num. Chron., 1890, p. 104.Google Scholar

^{page 959 note 3} Catalogue of Greek and Scythian Coins, introd., p. 40.

^{page 965 note 1} We gather from Ptolemy that the city had this other name: Ind. Ant., vol. 13, p. 349.Google Scholar

^{page 966 note 1} See my paper “Sāgala, Śākala, the city of Milinda and Mihirakula”, in the Acts of the Fourteenth Oriental Congress, Algiers, 1905; Indian Section, p. 164.

^{page 966 note 2} As regards some of the conquests and the raids into the interior of India which are attributed to this king, see my remarks in JRAS, 1912, p. 791Google Scholar: the raids are only mentioned in some grammatical illustrations which may refer to any Greek king quite as well as to him, and in the apocryphal Yugapurāṇa chapter of the Gārgī - Saṁhitā, which only mentions the Greeks in general terms, without specifying any names. The idea of making the grammatical illustrations apply to him seems to have been started in 1861; and it was based on the discovery of one solitary coin of him at Mathurā: see Goldstücker, 's Pāṇini, pp. 229, 234.Google Scholar

^{page 966 note 3} Not as being (as the translation says, SBE, vol. 35, p. 2)Google Scholar “in the country of the Yōnakas”: the text, ed. Trenckner, , p. 1Google Scholar, runs:—Atthi Yōnakānaṁ nānā - puṭabhēdanaṁ Sāgalan = nāma nagaraṁ. On the contrary, the work places it in Jambudīpa, India: see trans., p. 6.

^{page 966 note 4} Karṇaparvan, Calcutta text, § 44, lines 2028–70. In the Sabhāparvan, § 31. 1196, the city is mentioned as a trading-centre (puṭabhēdana) of the Madras.

^{page 967 note 1} See JRAS, 1912, p. 668–9, and p. 1002–3Google Scholar and note 1.

^{page 967 note 2} Unless, of course, we assign to him (as we probably should do) the coins which give on the reverse the name Kuyula-Kaphsa, and on the obverse the name Kozola-Kadaphes, with a head which imitates a Roman emperor's head (see p. 918 above, and note 4).

^{page 967 note 3} [See, respectively, JRAS, 1903, p. 325; 1910, p. 1311.]

^{page 967 note 4} [Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften, 1912, p. 828: see also the translation of ProfessorLüders, ' paper in Ind. Ant., 1913, p. 135.]Google Scholar

^{page 967 note 5} See my remarks in JRAS, 1912, pp. 100, 107.Google Scholar

^{page 968 note 1} A very useful abstract of M. Sylvain Lévi's Notes on the Indo-Scythians may be found in the Ind. Ant., 1903, pp. 381, 417Google Scholar (see p. 419 for the result quoted), and 1904, p. 10, with a supplementary note at p. 116; and a similar abstract of Professor Franke's notes about the Sök and Kaṇishka may be found in the same journal, 1906, p. 33 (see p. 47 for the result quoted); see also his short paper in JRAS, 1907, p. 675Google Scholar, on the identity of the Sök with the Śakas.

^{page 968 note 2} Dr. Thomas has referred vaguely to one certain case and one doubtful one of dated coins. There are, I think, several coins on which it has been proposed to read dates. But none of these dates, whether certain or doubtful, helps in our present inquiry, So it is not necessary to go into them.

^{page 969 note 1} The reference is to a small silver coin bearing the legend [ΛΙ]ΑΚΟ [Κ]ΟΖΟΥΛΟ, figured by Cunningham in his Coins of the Indo-Scythians, pl. 1, fig. 9, and see p. 91, = Num. Chron., series 3, vol. 9 (1889), pl. 13, and see p. 308.Google Scholar It is plainly an imitation of a small silver coin of Eucratides; ibid., fig. 8.

^{page 969 note 2} Epi. Ind., vol. 4, p. 55.Google Scholar

^{page 969 note 3} Epi. Ind., vol. 9, p. 144, G.Google Scholar

^{page 970 note 1} Only three heads are seen; one facing to the front, the other two facing right and left sideways. Hindū art, if representing the god as four-headed, would have shown all the four heads facing to the front, as in well-known pictures of the ten-headed demon Rāvaṇa. But these coins were made by Greeks or by Hindūs trained in Greek art: and we are to understand, I think, that there would be a fourth head, left to the imagination, at the back of the one which faces to the front.

^{page 971 note 1} Catalogue of the Coins of the Greek and Soythic Kings of Bactria and India, introd., p. 59.

^{page 972 note 1} The ēta has been cut too square: the upsīlon does not hit the mark at all; a nearer approach may be seen in JRAS, 1907, p. 1045, line 9.Google Scholar

^{page 973 note 1} Gardner, , Catalogue, pl. 25, figs. 3, 4, 5.Google Scholar

^{page 973 note 2} Cunningham, , Reports, vol. 2, p. 64.Google Scholar

^{page 974 note 1} JRAS, 1909, p. 1059Google Scholar; and compare ARASI, 1908–1909 (1912), p. 33Google Scholar

^{page 975 note 1} Epi. Ind., vol. 9, p. 141.Google Scholar

^{page 975 note 2} Epi. Ind., vol. 4, p. 55.Google Scholar

^{page 976 note 1} Ind. Ant., vol. 10, p. 326.Google Scholar

^{page 976 note 2} Ind. Ant., 1908, p. 58.Google Scholar

^{page 976 note 3} Journal Asiatique, 1890, i, p. 136.Google Scholar

^{page 976 note 4} Ind. Ant., 1908, p. 31Google Scholar: I quite agree with him.

^{page 976 note 5} For the latter see Epi. Ind., vol. 11, p. 210.Google Scholar

^{page 977 note 1} Namely, the Mōra and Mathurā inscriptions, Nos. 14, 59, 82, in Lüders' List of the Brāhmī, Inscriptions, Epi. Ind., vol. 10, appendix.Google Scholar

^{page 977 note 2} Ind. Ant., 1908, p. 40.Google Scholar

^{page 977 note 3} Lüders' List of the Brāhmī Inscriptions, No. 78.

^{page 977 note 4} See my remarks on the Hathigumphā inscription in JRAS, 1910, p. 824Google Scholar: and compare Professor Lüders' remark on the record, No. 1345 in his List:—“There is no date in this inscription.” No evidence of a Maurya era has been adduced, except that which was mistakenly supposed to exist in this Hathigumphā record.

^{page 978 note 1} Ind. Ant., 1908, p. 51–2Google Scholar

^{page 979 note 1} So also, in later times, the Chalukya kings of Bādāmi dated some of their charters in the years of their reigns alongside of the years of the Śaka era: see, e.g. Kielhorn, 's List of the Inscriptions of Southern India, Epi. Ind., vol. 7, appendix, Nos. 3, 9, 27–30, 32–34, 49.Google Scholar

^{page 979 note 2} See my account of the Hindū eras and other reckonings in the Encyclopœdia Britannica, 11th edition, vol. 14 (1910), pp. 495, 498.Google Scholar

^{page 979 note 3} The Buddhist era had not commended itself to any general use, but was confined to esoteric Buddhist circles. As to the Jain era, it is a question whether it was devised before the fifth century a.d.

^{page 980 note 1} For the record of the year 19 see Lüders' List of the Brāhmī, Inscriptions, Epi. Ind., vol. 10, No. 918Google Scholar. It was by pure conjecture that the editor of the inscription, Dr. Bloch, referred it to “Kanishka or Huvishka”. The king's name is quite illegible. The record is certainly not one of Huvishka: the choice lies between Kanishka and Vāsishka.

^{page 980 note 2} For the record of the year 31, see Lüders' List, No. 13a.

^{page 981 note 1} Archœol. Surv. West. India, vol. 2 (1876), p. 31Google Scholar; and see his letter to the Academy, dated 16 12, 1874, reproduced there on p. 32.Google Scholar

^{page 981 note 2} E. Thomas had the year 98 for Vāsudēva; but the year 9 was the earliest date then known for Kanishka.

^{page 981 note 3} JRAS, new series, vol. 7 (1875), p. 382.Google Scholar

^{page 981 note 4} Cunningham, (1892), Coins of the Kushāns, p. 5Google Scholar, = Num. Chron., 3rd series, vol. 12, p. 44Google Scholar: and later Bühler, (1896), Indische Palaeographie, § 19, BGoogle Scholar; the text says the “fourth” century, but that is a slip for “fifth”, —the century beginning with the year 401; see the English version, Ind. Ant., 1904Google Scholar, appendix, introductory note, p. 3, note 2.

^{page 982 note 1} Bühler, (1896), Vienna Oriental Journal, vol. 10, p. 173Google Scholar: whether this suggestion, dated 20 April, 1896, is to be taken as earlier or later than the other suggestion (referred to in the preceding note) in his Indische Palaeographie, which was published in 1896, I do not know.

^{page 982 note 2} Bhandarkar, D. R., JBBRAS, vol. 20, p. 297.Google Scholar

^{page 982 note 3} V. A. Smith, JRAS, 1902, p. 175Google Scholar; 1903, p. 7: at the same time it was proposed (1903, p. 45) to explain the year 72 of Śoṇḏāsa and the year 78 of Moga as meaning the years 2972 and 2978, = b.c. 105 and 99. The idea of applying this reckoning had been suggested, in fact, by Growse, F. S. in 1877, in Ind. Ant., vol. 6, p. 218Google Scholar, but only vaguely, without an indication of any particular century of the reckoning.

^{page 982 note 4} If, by chance, anyone should ever wish to revive it in any form let him first discover (as I have suggested in JRAS, 1906, p. 981)Google Scholar a record of Kanishka with a date ranging (say) from the year 91 to 100, or a record of Vāsudēva with a date ranging (say) from the year 1 to the year 10.

^{page 983 note 1} Rājataraṁgiṇī, 4. 703.

^{page 983 note 2} For further information about this reckoning, reference may be made for the present to my account in the Encyclopœdia Britannica, 11th edition, vol. 13 (1910), p. 499Google Scholar. Properly speaking, the names Laukika and Lōkakāla belong only to the reckoning with the hundreds omitted; and they are found used, I think, only with it. The other names belong properly to the full reckoning only, but are sometimes found coupled with the abbreviated reckoning.

^{page 983 note 3} On the great rarity of such cases, compare a remark by ProfessorKielhorn, in Ind. Ant., vol. 22, p. 109bGoogle Scholar. Even the few instances which are forthcoming are not of a complete and convincing kind: thus:—

The Girnār inscriptions of a.d. 1232 cite the years 76, 77, and 79, but explain them as meaning 1276, 1277, and 1279, by mentioning first the year Vikrama-samvat, 1288: Epi. Ind., vol. 5, appendix, No. 212.Google Scholar

A manuscript of a.d. 1527 gives its date as saṁvat Āshāḍhādi 83 varshē, etc., by which it means the Vikrama year 1583 in the Āshāḍhādi variety: Ind. Ant., vol. 19, p. 360, No. 167.Google Scholar

These are the only real instances that I can quote: and perhaps the second of them was not actually intended, but (see Professor Kielhorn's remarks, loc. cit.) is due to a careless omission of the figures 15 after saṁvat, which would transfer it to the same class with the following three cases, which are only of a partial nature:—

A manuscript of a.d. 1477 gives its date as saṁvat 15 Āshāḍhādi 34 varshē, etc., meaning the Vikrama year 1534 in the Āshāḍhādi variety: Ind. Ant., vol. 19, p. 32Google Scholar, No. 47.

A manuscript of a.d. 1523 gives its date as saṁvat pañchadaśa 15 aśītau 80 pravartamānē, etc., meaning Vikrama 1580: ibid., p. 33, No. 49.

A manuscript of a.d. 1643 gives its date as saṁvat 16 Āshāḍhādi 99 varshē, etc., meaning Vikrama 1699 in the Āshāḍhādi variety: ibid., p. 171, No. 101.

^{page 984 note 1} Cunningham, , Reports, vol. 5, p. 61Google Scholar, plate 16, No. 4: and see better, for the word rajami, JASB, vol. 23 (1854), p. 705, plate.Google Scholar

^{page 984 note 2} Lüders, List of the Brāhml Inscriptions, No. 78.

^{page 984 note 3} ARASI, 1903–1904 (published in 1906), p. 259Google Scholar. On this matter compare my remarks in JRAS, 1907, p. 184.Google Scholar

^{page 984 note 4} Col. Waddell has asked (p. 950 above) that this date should be examined again. The pedestal bearing the record is in the British Museum: and I am able to say, from an excellent squeeze for which I am indebted to Sir C. Hercules Read, that the figures are unmistakably 300, 80, and 4.

^{page 985 note 1} Varsha ēkunaśitiśatimaē, or vashra ēkanavitaśatimaē.

^{page 985 note 2} Buddhistische Kunst in Indien (1893)Google Scholar, revised translation by Burgess, , Buddhist Art in India (1901), p. 5.Google Scholar

^{page 986 note 1} See Wroth, 's Catalogue of the Coins of Parthia, p. 153Google Scholar ff. for dates from a.d. 41–2 onwards, and pp. 99–149 for dates from b.c. 38–37 to a.d. 26–7: see also introd., p. 65.

^{page 986 note 2} The expression is vasha ēkunachaduśatimaē, “in the four-hundred-less-by-oneth year”; using a well-established Hindū method of denoting ‘three hundred and ninety-nine.’

^{page 986 note 3} Op. cit. (note 2 on p. 985 above), p. 84: he has said so specifically in respect of the Loriyān Tangai and Hatnagar dates: the Skārah Ḍherī image was not known when he was writing.

^{page 986 note 4} See p. 981 above.

^{page 986 note 5} Except perhaps in one very questionable instance: see Wroth, , Catalogue of the Coins of Parthia, p. 21, note 3.Google Scholar

^{page 986 note 6} Smith, G., Assyrian Discoveries (1875), p. 389.Google Scholar

^{page 987 note 1} Akbarnāma, trans. Beveridge, , vol. 1, p. 21–2.Google Scholar

^{page 987 note 2} See JRAS, 1910, 819.Google Scholar

^{page 988 note 1} Kielhom's List Of the Inscriptions of Northern. India, Epi. Ind., vol. 5Google Scholar, appendix, Nos. 14, 352.

^{page 988 note 2} Ibid., No. 362. The records Nos. 354 to 361, with dates ranging from a.d. 1018–9 to 1135–6, belong properly to Southern India; as also do various other records now standing in the Northern List.

^{page 988 note 3} Ibid., No. 351. Its date is there given as Śaka 726, = a.d. 804–5; but it is known now that the real year is 1126, = a.d. 1204–5; see ARASI, 1905–1906 (published in 1909), p. 19–20.Google Scholar

^{page 989 note 1} Albērūnī's India, trans. Sachau, , vol. 2, p. 6Google Scholar. Karūr seems to be the ‘Kahror, Karrur’, of maps, about twenty miles north-east of Bahāwalpur.

^{page 990 note 1} The various terms used by Varāhamihira to denote the era of a.d. 78 are (1) Śakēndra-kāla, here (8, 20) and in Pañchasiddhāntikā, 12. 2; (2) Śakabhūpa-kāla in Bṛihat-Saṁhitā, 8. 21; and (3) Śaka-kāla in Bṛihat-Saṁhitā, 13. 3, and Pañchasiddhāntikā, 1. 8: there is nothing in these terms, or apparently anywhere in his writings, to lead us to think that he himself (he died in a.d. 587) thought that the era began with a destruction of the Śakas.

^{page 990 note 2} Bhaṭṭōtpala's words are:—Śakā nāma Mlēchchha-jātayō rājānas=tē yasmin = kālē Vikramādityadēvēna vyāpāditāḥ sa kālō lōkē Śaka iti prasiddhaḥ | tasmāch = Chhakēndra-kālāt Śakanṛipa-badhād = ārabhya, etc. Under 13. 3, he explained Śaka-kāla by Śakanṛipa-kāla.

^{page 990 note 3} Brahma-Siddhānta, ed. Dvivedi, Sudhakara, p. 8Google Scholar, verses 8, 9. The commentator, Pṛithūdaka (who seems to date from about the same time with Bhaṭṭōtpala), explains Śak-āntē by Śaka-kālāt-prāk, “before the Śaka time or era”, and Śakanṛip-āntē by Kalp-ādēr = ārabhya Śaka - kāl - āvadhau, “from the beginning of the Kalpa up to the Śaka time or era”: he does not add any remarks. Brahmagupta's exponent, Bhāskarāchārya (wrote a.d. 1150), divided the compound into Śakanṛipasy-āntē, “at the end of the Śaka king”; Siddhānta-Śirōmaṇi, ed. Sastri, Bapu Deva, p. 12, verse 28.Google Scholar

In another place in his Siddhānta (p. 410, verse I), in laying down the epoch Śaka 550 expired, Brahmagupta used the expression Śāka, ‘of or relating to the Śaka’. The only other place in this work in which he has mentioned the era by name seems to be the verse in which he tells us that he was writing, at the age of thirty, when there had elapsed 550 years Śaka-nṛipāṇām, “of the Śaka kings”; op. cit., p. 407, verse 7. There appears to be some doubt about the reading here. The editor has quoted this verse with the reading Śaka-nṛipālāt, “from or since the Śaka king”, in his Gaṇakataraṁgiṇī or “Lives of Hindū Astronomers”, p. 18; and again even in the introduction, p. 1, to his edition of the work itself. And Bhau Daji also had the reading nṛipālāt from manuscripts; JBBRAS, vol. 6 (1861), p. 27Google Scholar, note †; and JRAS, new series, vol. 1 (1865), p. 410Google Scholar. Weber seems to have suggested that nṛipālāt is a mistake for nṛip-āntāt; Sansk. Lit., p. 259Google Scholar, note 287. So also might be nṛipāṇām. But, also, neither of them conflicts necessarily with the expressions Śak-āntē and Śakanṛip-āntē.

^{page 991 note 1} Albērūni's India, trans. Sachau, , vol. 2, p. 7.Google Scholar

^{page 991 note 2} Kielhorn, List of the Inscriptions of Southern India, Epi. Ind., vol. 7Google Scholar, appendix, No. 10.

^{page 992 note 1} Kielhorn, List of the Inscriptions of Southern India, Epi. Ind., vol. 7Google Scholar, appendix, No. 3:—Śakanṛipati-rājyābhishēka-saṁvatsarēshv = atikrāntēshu pañchasu śatēshu.

^{page 992 note 2} The Hindūs themselves, of later times, seem to have been much exercised in their minds over the idea that an era known as Śakanṛipakāla, “the time of the Śaka king or kings”, should commemorate a destruction of the Śakas. Bhāskarāchārya, the exponent of Brahmagupta, giving the same rule, says in his Siddhāntaśirōmaṇi, 1. 28 (written a.d. 1150), that a certain number of years of the Kali had elapsed Śakanṛipasy = āntē, “at the end of the Śaka king”. The Vāsanāvārttikā commentary (a.d. 1621) analyzes Śakanṛipa as Śaka-nṛi + pa from pā, pāti, ‘he protects or governs’, and says:—“The Śakas were certain men whom Vikramāditya protected; whence he was called Śakanṛipa, ‘protector or lord of Śaka men’; the name being given to him just as that of mṛigapati, ‘lord or protector of animals’, is applied to the lion, though he takes the life of animals; according to the saying of Bhaṭṭōtpala [quoted on p. 990 above, note 2]. It is well established among people, both high and low that the end of Vikrama was the beginning of Śālivāhana” [the later reputed founder of the Śaka era].

Another commentary, the Marīchi (a.d. 1635), analyzing in the same way, but taking pa from pā, pibati, ‘he drinks’, in the sense of mārayati, ‘he destroys’, says:—“Vikramāditya slew some barbarian men named Śakas, and so was called Śakanṛipa, ‘drinker up (slayer) of Śaka men’: the meaning of ‘at the end of him’ is ‘at the beginning of the era of Śālivāhana’. In the summing-up verse about the six era-kings in the Kali age, beginning ‘Yudhishṭhira, Vikrama and śalivāhana’ [quoted by me from almanacs in JRAS, 1911, p. 694Google Scholar: it is also found in the Jyōtirvidābharaṇa, ibid., p. 697], it is said that the era of Śālivāhana began directly after the end of the era of Vikrama.”

^{page 993 note 1} Nahapāna's capital, Minnagara, seems to be the modern Dōhad in the Pañch Mahāls; see JRAS, 1912, p. 788Google Scholar. In addition to his co-regent or viceroy Chashṭana at Ujjain, he had a similar representative named Bhūmaka in Kāṭhiāwār.

^{page 993 note 2} JRAS, 1910, p. 818; 1912, p. 785.Google Scholar

^{page 993 note 3} Nahapāna's son-in-law Ushavadāta, son of Dīnīka, was a Śaka: see Lüders, , List of the Brāhmī Inscriptions, Epi. Ind., vol. 10, appendix, No. 1135Google Scholar. So the Kshaharātas and the Śakas were at least on very close terms.

^{page 993 note 4} Peterson, , Third Report on Sanskrit MSS., p. 32Google Scholar, and extracts, p. 26.

^{page 994 note 1} See JRAS, 1905, pp. 646, 647Google Scholar; and p. 649 for the forms Sakka, Śakka.

^{page 994 note 2} Sh. B. Dikshit said in his Bhāratīya-Jyōtiḥśāstra or “History of Indian Astronomy”, p. 371Google Scholar, that the Vikrama era is found used along with the Śaka era in “one or two, or a few, astronomical books (ēk dōn jyōtisha-granthāṁt).” He did not name these works: but the context makes it plain that he meant some rather late writings: and there is of course a difference between taking a Vikrama year as an epoch or basis of calculation, and, on the other hand, simply citing such a year along with the Saka year, as is done habitually in almanacs. In Sewell, and Dikshit, 's Indian Calendar, p. 40Google Scholar, note 2, we are told that:—“The Vikrama era is never used by Indian astronomers.” So, also, Dikshit said in his book mentioned above, p. 372, that the Kaliyuga and Śaka eras are used in works dealing with astronomical calculations (jyōtishagaṇita-grantha), but the Vikrama and “other” eras are not so used. Here, again, we have a statement in the present tense: but it cannot be doubted that he meant to tell us what has always been the practice. For my own part, not in any direction have I found the slightest indication that the Vikrama era may be classed as an astronomical reckoning. It is an instructive point, that no instance is forthcoming of the epoch of a Karaṇa, or practical work on the astronomy and calendar, being laid down as the year so-and-so of the Vikrama era: it is always a Śaka date that is found used in such works.

^{page 995 note 1} Kielhorn, List of the Inscriptions of Northern India, No. 1. It is to be remembered that Professor Kielhorn was dealing only with the inscriptions from about a.d. 400 onwards, tie did not touch the Kushan dates at all, except in a remark in Ind. Ant., vol. 26, p. 153Google Scholar, where he drew attention to the point that they always denote the word ‘year’ by saṁvatsara or its abbreviations, whereas the records and coins of the Western Kshatrapas (dated in the Śaka era) always use the word varsha.

^{page 996 note 1} This is a new record, quite recently brought to notice by MrBhandarkar, D. R. in Ind. Ant., 1913, p. 161Google Scholar. That the perpetuation of the reckoning of b.c. 58 was due to its being preserved and handed on by the Mālavas, who were a leading people among the subjects of the founder of it, and that it thus acquired its earliest known appellation, has been pointed out by me on several occasions: see, e.g., JRAS, 1905, p. 233Google Scholar; 1907, p. 171.

In his remarks on this new record, in commenting on a term, kṛita, used in it and elsewhere, the meaning of which is not yet apparent, MrBhandarkar, has said (loc. cit., p. 163)Google Scholar:—“It is not safe just at present to make an assertion on this point, but it appears to me that what is now known as the Vikrama era was invented by the people or astronomers for the purpose of reckoning years and was consequently originally known as Kṛita, which means ‘made’”. Before advancing such a proposition, he should have looked into the facts: as I have had occasion to remark above, it may be safely said that this era has never been used as an astronomical reckoning, and there is no good reason for thinking that it is an artificial reckoning, invented for such or any other special purposes.

Dr. Thomas, of course, did not know this record of a.d. 405 when he was writing. The record which he should have cited as giving the earliest known instance of a name for the era, is the Mandasōr inscription of “the year 529 expired”, in a.d. 473, which presents also an earlier date, the year 493 expired, in a.d. 437, and uses with that earlier date the expression Mālavānām gaṇa-sthityā; Northern List, No. 3. A similar expression, Mālava-gaṇa-sthiti-vaśāt, is found in the Mandasor inscription of the year 589 expired, = a. d. 532–3; Northern List, No. 4. I rendered these expressions, for reasons which I gave, as denoting a reckoning dating from “the tribal constitution of the Mālavas”; Gupta Inscriptions, pp. 87, 158Google Scholar. ProfessorKielhorn, said in Ind. Ant., vol. 19 (1890), p. 56Google Scholar, that gaṇa-sthiti should be taken as equivalent to gaṇanā; “by or according to the reckoning of the Mālavas”. I have never felt quite sure as to the appropriateness of this correction. Still, I accepted it, and acted in accordance with it. It is now upset, of course, by the new inscription which is mentioned above. But my translation of gaṇa-sthiti by “tribal constitution” is not restored: as indicated by Mr. Bhandarkar, we must translate the two expressions in which it occurs by some such words as “according to the usage of the Mālava tribe”.

^{page 997 note 1} As shown by the Kaṇheri record of the year 245; Kielhorn's Northern List, No. 393

^{page 998 note 1} The same term is implied in the Bijayagaḍh inscription of the year 428 (expired), in a.d. 372 (Kielhorn's Northern List, No. 1), and the Gaṅgdhār inscription of the year (?) 480 expired, in a.d. 423 (ibid., No. 2), which otherwise do not present any name for the era: in these two records we have Kṛitēshu in apposition with varsha-śatēshu and vatsarēshu; and (by the way) in the second of them we have also another puzzling term, which may be meant for, or may actually be, saumyēshu.

It is hardly possible that Kṛita and Kṛita-saṁjñita can be intended (as Mr. Bhandarker has thought) to discriminate one or the other of the Kārttikādi and Chaitrādi varieties of the era which now exist: for one reason, it is questionable whether the Chaitrādi year was in use at all in that early time, and much more so whether it could have been used then with the era of b.c. 58. It seems just possible that an explanation may be found in a statement of Hiuen-tsiang, who says that, after the death of Kanishka, the sovereignty of Kashmīr was seized by a race named Ki-li-to, the members of which suppressed the Buddhist religion and held the country until the 600th year after the death of Buddha (a.d. 118), when they were overthrown and Buddhism was restored by a king of Himatala, of the Tukhāra country: Beal, , Si-yu-ki, vol. 1, p. 156Google Scholar; Waiters, , On Yuan Chwang, vol. 1, p. 278Google Scholar. But I can only offer this as a mere conjecture: and the name Ki-li-to is explained by the Chinese as meaning ‘bought’ (op. cit., pp. 150; 265), which points to krīta, instead of kṛita.

^{page 999 note 1} Regarding this king, see p. 1002 below.

^{page 999 note 2} See p. 984 above

^{page 1001 note 1} Epi. Ind., vol. 4, p. 55.Google Scholar

^{page 1001 note 2} See the text and translation given by me in JRAS, 1907, p. 1014.Google Scholar

^{page 1001 note 3} Epi. Ind., vol. 9, p. 144Google Scholar, G. There were evidently two Patikas: the second of them, mentioned as simply Patika in the Taxila record, was the son of Liaka-Kusulaka; the other, the Great Satrap Kusulaa Patika of the Mathurā inscription G, seems to have been either the father or a brother of Liaka.

^{page 1002 note 1} Regarding this record, see fully my remarks in JRAS, 1905, p. 229Google Scholar; 1906, p. 706; 1907, p. 1039.

^{page 1003 note 1} For a clear and compendious account of this tradition, see MrPhilipps, W. R.' article on “The Connection of St. Thomas the Apostle with India” in Ind. Ant., 1903, pp. 1, 145.Google Scholar

^{page 1004 note 1} JRAS, 1913, p. 95Google Scholar. Dr. Barnett at any rate, has understood me correctly (p. 943 above), without having to ask me as to my meaning.

^{page 1004 note 2} As regards Sung-yun I may refer to my remarks in JRAS, 1906, p. 989Google Scholar. For the statement of the Sūtra, quoted by Dr. Thomas from M. Sylvain Lévi's account of that work, reference may also be made to the English abstract of Lévi, M.'s account in Ind. Ant., 1903, p. 382.Google Scholar

^{page 1004 note 3} Rājataraṁgiṇī, , 1. 172.Google Scholar

^{page 1005 note 1} See my remarks in full on this point in JRAS, 1906, p. 990.Google Scholar

^{page 1005 note 2} See JRAS, 1912, p. 239.Google Scholar

^{page 1006 note 1} Lüders, List of the Brāhmī Inscriptions, Nos. 925, 927, 56.

^{page 1006 note 2} Rājataraṁgiṇī, 1. 168:—Hushka-Jushka-Kanishk-ākhyās = trayas = tatr = aiva pārthivāḥ. Kalhaṇa has mentioned them in this order, not by way of fixing the succession, but because it was convenient to use a compound word, and necessities of metre and rhythm (to say nothing of a rule of grammar) required him to place the longest base, Kanishka, last. Under the influence of his first mention of them, he has spoken in 8. 3412 of trayō=tha Hushk-ādyāḥ; “and then the three, Hushka and the others.”

^{page 1006 note 3} See Epi. Ind., vol. 8, p. 173Google Scholar. I do not raise any objection to Dr. Vogel's view that the image and the umbrella and its post came from Mathurā: why not?: there is, I believe, no stone, at any rate of a suitable kind, anywhere near Benares; and there was water-carriage all the way from Mathurā to that place. As regards the other point, he would probably have hit the mark if he had said that the Bhikshu Bala, who gave the image and the umbrella and its post, was a pilgrim from Mathurā: we find the same person making gifts at Sahēṭh-Mahēṭh in the year 19: see Lüders' List, No. 918.

^{page 1007 note 1} Lüders' List, No. 918. It was by pure conjecture that the editor of this record, Dr. Bloch, referred it to the time of “Kanishka or Huvishka”. The king's name and titles are quite illegible. The record is certainly not one of Huvishka: the choice lies between Kanishka and Vāsishka.

^{page 1007 note 2} Epi. Ind., vol. 11, p. 210.Google Scholar

^{page 1008 note 1} I freely give up my suggestion, made in JRAS, 1903, p. 334Google Scholar, that Kanishka reached India from Khotan through Kashmīr. [I did not mention “the pilgrims' route (via the Swat Valley)”, to which Dr. Thomas has referred; p. 635, note 1] The idea of a route through Kashmīr seems to have been suggested first by Gardner, who conjectured (Catalogue, introd., p. 40) that the people over whom Mauēs ruled “had entered India not through the Kabul Valley, but through Kashmir or Nepal.” Cunningham negatived that (Coins of the Sakas, p. 2Google Scholar, = Num. Chron., 3rd series, vol. 10, 1890, p. 104)Google Scholar; but not in any unpleasant terms.

^{page 1008 note 2} See JRAS, 1912, p. 669.Google Scholar

^{page 1008 note 3} Lüders' List, No. 59.

^{page 1008 note 4} Lüders' List, Nos. 14, 82.

^{page 1008 note 5} This is the ‘Rañjubula’ of some writers. See my remarks on the name in JRAS, 1907, p. 1024Google Scholar, note 2, and p. 1027, note 4. If the Kharōshṭhī legends on his coins are to be taken as giving an anusvāra in the first syllable, then the name will be Rāñjūvula.

^{page 1009 note 1} Epi. Ind., vol. 9, p. 141.Google Scholar

^{page 1009 note 2} See, e.g., JRAS, 1907, p. 1027.Google Scholar

^{page 1009 note 3} Lüders' List, Nos. 60, 76.

^{page 1009 note 4} Compare, e.g., the case of the well-known Ushavadāta, who had this Hindū name (Ṛishabhadatta) though he was a Śaka, son of Dīnīka and son-in-law of Nahapāna: it would seem from this that Ushavadāta's mother was a Hindū: and Nahapāna himself seems to have had a Hindū wife, since the name of his daughter, Ushavadāta's wife, was Dakhamitrā (Dakshamitrā); Lüders' List, Nos. 1132, 1134. We have another instance of Hinduization in the case of the Śaka Agnivarman; Lüders' List, No. 1137

^{page 1010 note 1} See p. 1000 above.

^{page 1010 note 2} See p. 1002 above.

^{page 1010 note 3} Cunningham, , Reports, vol. 5, p. 61Google Scholar, and plate 16: for the word rajami at the end of line 1, see better the plate at JASB, vol. 23 (1854), p. 705.

^{page 1011 note 1} See, as regards the Brāhmī, Bühler, , Epigraphia Indica, ii, pp. 195–6Google Scholar; Vogel, ibid., viii, p. 175; Lüders, ibid., ix, p. 247; and especially the valuable paper of the Boyer, Abbé in the Journal Asiatique, sér. ix, vol. xv, pp. 565 sqq., where details are given.Google Scholar

^{page 1012 note 1} Dr. Fleet's method of meeting this difficulty (p. 1005 infra and ref.) requites that a certain Arhat, in speaking of a council, of which he had himself been a member, as having taken place “in recent times”, is referring to an event to be dated at least a century earlier.

^{page 1013 note 1} As indicated above (p. 643), I think this H to be not Roman, but Aramaic (it is no accident that it is found only on Śaka-Pahlava coins); nor do I agree that the Roman P p appears on Nahapāna's issues.

^{page 1014 note 1} The mistakes which Mr. Kennedy cites (pp. 922–3 and n. 1) from Egyptian and other sources are no evidence of a decaying knowledge of Greek. They are popular or half-educated Greek, similar to the use of ‘you and I” after verbs and prepositions in bad English. Grammatical errors are not rare even in official inscriptions of Attica.

^{page 1015 note 1} The statement (p. 923) that Kaniṣka ‘begins with uncials”and changes to the cursive is, of course, quite unfounded; in the use of a slightly varying alphabet he simply follows the practice of his predecessors.

^{page 1015 note 2} The fact that Kujula Kadphises uses P and not Þ for sh makes no difference, and probably points to the origin of the latter.

^{page 1015 note 3} Similarly Mr. Tarn assumes (Journal of Hellenic Studies, vol. xxii, p. 286)Google Scholar that “the changes in the Greek letter-forms corresponded to those in Greek letter-forms elsewhere”. The changes after Kaniṣka are, however, from this point of view of insignificant magnitude.

^{page 1016 note 1} Note also that a somewhat similar Þ, which certainly is a rho (see above, p. 630 n.), is read as sh by Cunningham in the word XAÞOBAΛANO

^{page 1016 note 2} We should, however, say sixty years, since Vāsudeva adds no new name, except his own.

^{page 1018 note 1} Mr. Tarn, however, writes “Kanishka's die-sinkers, if they possessed Greek in any sense as a living tongue”.

^{page 1018 note 2} See my reff., supra, p. 648, n. 1.

^{page 1019 note 1} Coins of the Indo-Scythians, i, p. 22.Google Scholar

^{page 1019 note 2} Mr. Tarn, however, accepted the re-striking (op. cit., p. 276, n. 42).

^{page 1020 note 1} See the list given above, p. 645 n.

^{page 1020 note 2} See his paper in the Göttingische Gelehrte Nachrichten, 1910, pp. 427–41, especially p. 438Google Scholar

^{page 1021 note 1} Mr. Kennedy's total of 28 for Kaniṣka omits 68 found in the year 1834 alone (Masson, , J.R.A.S.B., 1834, p. 163).Google Scholar

^{page 1022 note 1} Archæological Survey Reports, vol. ii, p. 160. He says “Sassano-Arabian”.Google Scholar

^{page 1022 note 2} Cunningham, , op. cit., pp. 161 seqq.Google Scholar

^{page 1023 note 1} As now does Dr. Fleet himself (p. 1008 n.).

^{page 1025 note 1} See Marquart, , Ērānshahr, p. 203, n. 3.Google Scholar

^{page 1026 note 1} The conquest of Kabul, from the Parthians was pointed out in this Journal for 1906 (pp. 193–4)Google Scholar, and was supported by a classical reference.

^{page 1026 note 2} Journal Asiatique, sér. ix, vol. xv, p. 579, 1900.Google Scholar

^{page 1027 note 1} But of Vāsudeva Mr. Kennedy says that “his rule extended apparently over the Eastern Panjâb, and no farther” [west] (1912, p. 673).Google Scholar

^{page 1027 note 2} JRAS, 1912, p. 673.Google Scholar

^{page 1028 note 1} See DrVogel, 's remarks in Epigraphia Indica, viii, p. 175.Google Scholar

^{page 1028 note 2} As regards the passes, see my reply to Mr. Kennedy.

^{page 1029 note 1} Mr. Dames in discussing this type (pp. 955–6) neglects the fact that Śiva appears on Indian coins as early as b.c. 100 (see Cunningham, , Coins of Ancient India, pp. 66 seqq.Google Scholar; Rapson, , Indian Coins, § 43)Google Scholar. It seems to me that Kadphises II combined a Śiva taken from the Gondophernes line with the bull taken from the Śakas, etc., and that Vāsudeva in adopting the type added the “Okro” (or however it is to be read).

^{page 1030 note 1} Mr. Kennedy also has referred (p. 927, n. 2) to this error, which does not appear in the printed paper, “chief” being substituted (p. 629)Google Scholar. The recollection to which the oversight was due is mentioned in a note.

^{page 1031 note 1} As a matter of fact, Vergil does not refer to trade at all, whether via India or via Parthia, or by any other route; he Writes merely the line—

“velleraque ut foliis depectant tenuia Seres.”

^{page 1032 note 1} Indian Antiquary, 1903, pp. 349–50Google Scholar; Kavīndravacansaamuccaya (Bibl. Indica, 1912), pp. 25–9.Google Scholar

^{page 1032 note 2} Journal Asiatique, sér. ix, vol. viii, p. 449.Google Scholar

^{page 1033 note 1} Journal Asiatique, sét. ix, vol. viii, p. 450, n.Google Scholar

^{page 1034 note 1} (1) As to the early history of Sāgala (pp. 965–6), it is not necessary to dwell upon its long-established identity with the Śākala of Sanskrit books. (2) As to Kujula Kadphises taking Hermæus “under his protection” (pp. 966–7), let us substitute “being in alliance” with Hermæus, although in the circumstances of the case the former phrase is probably a more correct statement. (3) In regard to the Chinese evidence (pp. 967–8), I meant to indicate a general consensus on the main points; and I should certainly see “no objection to noting” the divergent views of Professor Sylvain Lévi and Professor Franke in certain respects, since, in fact, I had mentioned both these scholars (p. 637 and note). (4) In mentioning, but without adopting, the views of Bühler and Mr. Vincent Smith (pp. 980–4), who certainly deserve mention, concerning omitted hundreds, I do not feel that I went too far. (5) The Parthian era of b.c. 248 (p. 986) was not cited by me as supplying a solution of Indian dates, but as an example of an era contemporaneously current in the same sphere with a second (namely the Seleucid era of b.c. 312) of not very widely divergent epoch. (6) Concerning Dr. Vogel's application of the dates 318 and 384 to the Seleucid era (pp. 984–6), I am still of opinion that the view is in agreement with pal;æographic facts (and also, since some inscriptions have Seleucid months, I see no great objection to the supposition of Seleucid years), while as regards the date 399 read by Dr. Fleet in the Skara Dheri inscription, an inscription with which I have some familiarity, I see what there is in favour of Dr. Fleet's reading, though the matter is not decisively settled, and, if settled in favour of the year 399, it would not be counter to Dr. Vogel's view. (7) The derivation of the sh sign Þ from the sign for rho P is supported, as Professor Rapson has pointed out to me, by the fact that the P sign itself is used with the value of sh not only in the legends of Kujula. Kadphises, but also in those of the Śaka-Pahlava king Spalirises (see my table, p. 640): further, there are good phonetic reasons for a connexion between r and a sibilant, as may be seen even in English (e.g. lose and lorn), and, as Professor Rapson mentions, in the name of the composer Dvořak. (8) (p. 989) My view as regards the initiation of the Śaka era is hardly open to misconception, since I adduced years ago in this Journal (1909, pp. 465–6) the decisive evidence which proves that the dates of Kaniṣka are regnal, and I have asserted above (p. 633) that the satraps of Western India adopted this era (naturally, under compulsion) as that of their suzerain. It commemorates their overthrow inferentially, as commemorating the beginning of another power, that of the Kushans (for Wima Kadphises, the real conqueror, established no era). Though I did not mention the popular story concerning the origination of the Gupta era (I had no reason for so doing), Dr. Fleet, who saw the paper in its first proof, can hardly be unaware that it was in my mind. (9) And lastly, I did not by using the phrase “might deceive the very elect” intend severity (p. 918, n. 1), but rather a sense of the extreme plausibility of the theory: a reference to the use of this proverbial expression and to its source will show that it does not imply any identification of the “elect” or an inclusion of the user of it among their number.

^{page 1036 note 1} Epigraphia Indica, ii, pp. 195 sqq.Google Scholar

^{page 1037 note 1} Cf. MrKennedy, 's use of the phrase, p. 926, n. 1.Google Scholar

^{page 1040 note 1} One list assigns to the Kaniṣka group 1,540 coins, including 585 of Vāsudeva. Most of these must have come from Begram.

^{page 1042 note 1} New evidence has not been long in coming. In a note at the end of my paper (p. 649) reference was made to the argument from archæological stratification, especially in connexion with the excavations at Taxilā. By the courtesy of Mr. Marshall I have now received a print of a lecture delivered by him at Simla on September 4. Mr. Marshall finds the Śaka-Pahlava strata above the Greek and below the Kushan. As regards dates, he assigns Maues to about b.c. 50, in which case this king will hardly be other than Moga: Kaniṣka he would place in the second century a.d.