JOURNAL ASIATIC SOCIETY OF BENGAL. VOL. LIII. PART I. (History, ANTIQUITIES, &C.) (Nos. I and II, 1884: with six Plates and a Map; and witha Special Number as substitute for Nos. III and IV). EDITED BY que Puiotocicat SEcRETARY. “Jt will furish, if naturalists, chemists, antquaries, philologers, and men of science in different parts of Asia, will commit their observations to writing, and send them to the Asiatic Society at Calcutta. It will languish, if such communications shall be long intermitted ; and it will die away, if they shall entirely cease.” SIR WM. Jones: CALCUTTA: PRINTED BY J. W. THOMAS; AT THE BAPTIST MISSION PRESS, AND PUBLISHED BY THE ASIATIC socteTY 57, PARK STREET. 1884.abe JOURNAL OF THE ASIATIC SOCIETY OF BENGAL, —o Part I.—-HISTORY, LITERATURE, &e. No. II.—1884. 4 Classified and Detailed Catalogue of the Gold Coins of the Imperial Gupta Dynasty of Northern India, with an Introductory Essay.— By V. A. Surra, B. A. Dubl., B. C. S. (With 4 Plates and a Table of Weights.) Intropuctory Essay. Section I—Preliminary. The imperial Gupta dynasty is known to have consisted of a lineal succession from father to son of seven sovereigns, not including Budha Gupta, a local ruler in the country between the Jamuné and the Narmadé, nor various other princes who retained a grasp on the eastern portions of the Gupta empire, after the decadence of the imperial family. No coins are known to exist which can be referred with certainty to the founder of the dynasty, who is in the inscriptions simply named Sri Gupta.® Ghatot Kacha, Chandra Gupta J, and Samudra Gupta who * A gold coin found in Jessore was formerly attributed to Srf Gupta, (J. A. 8. B.XMI, p. 401; Pl. XII, 10), but this attribution cannot be maintained. The unique Ailver coin, belonging to Mrs. Freeling, which was at one time believed to belong to the reign of Sri Gupta, is plainly a coin of Skanda Gupta (Records of the Gupta Dynasty, yp. 49, 50). General Cunningham, nevertheless, still assigns to Sri Gupta an unpub- Tiahed coin in his cabinet. In the caso of this prince the word Sri would seem to be tm integral part of his name, for the past participle ‘Gupta’ can hardly stand done. Sri Gupta would therefore mean ‘protected by Sri’ or Lakshmi. In the tames of the succeeding princes the word ‘Sri’ is used only as the customary hono- Tif prefix, which is, in my opinion, best left untranslated. I-tsing speaks of the king who preceded his time by 500 yoars as ‘Sri Gupta,’ not simply aa ‘Gupta’ (R.A. 8. Vol. XIII, N. 8. p. 671.) @120 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No.2 were respectively the second, third, and fourth sovereigns of the line, appear to have coined in gold only, and gold pieces of all these princes are extant. It is probable that during their reigns, as in the time of their Indo-Scythian predecessors, the silver currency was supplied “by the abundant issues of the Greek princes.”* The fifth king, Chandra Gupta II, has left coins in gold, silver, and copper, as algo has his son and successor Kuméra Gupta Mahendra. Skanda Gupta, the last of his line who enjoyed imperial power,t did not, as far as is at present known, issue any copper coinage, but specimens both of his gold and silver mintages exist in considerable numbers. Budha Gupta’s money is known in silver only; and the coins of Nira Gupta and other eastern successors of the imperial dynasty occur only in gold, though the metal is often very impure. The design of my catalogue is limited to a description of the coins of the imperial Guptes, as above defined, but, inasmuch as earlier pub- lications on the subject do not diecriminate the local and imperial coin- ages, I have been compelled to notice briefly in a Supplement some of the coins of Néra Gupta and other minor kings of uncertain date and lineage. The full discussion of these later coins would require a long dissertation to itself. For various reasons I shall not attempt to discuss the silver coinages of the Gupta kings, although there is still room for a comprehensive essay on the subject.t The rare copper coins of Chandra Gupte II and Kuméra Gupta Mahendra seem to possess comparatively little historical interest, and, except as curiosities, are certainly of less importance than the gold and silver coins. I therefore pass them by for the present, without detailed notice. It is probable, as suggested by Wilson, that the vast Indo Scythian issues of copper coin supplied the Gupta era with nearly eufi- * Ariana Antiqua, p. 348, + On another occasion I may perhaps venture on a review of what is known of Gupta history and chronology, but there is reason to hope that the task may be undertaken by 8 more competent hand. For the present it will suffice to say that I consider the death of Skanda Gupte, or, at least, the break up of his empire, to have occurred in A. D. 818-319, and both the reign of Sri Gupta and the Gupta era to have begun about 160-170 A, D. I altogether dissent from the view of Prof. Oldenberg and other writers who make the Gupta dynasty begin in A, D. 818-9; and Tam equally unable to agree with Mr. Thomas in identifying the Gupta and Saka eras. {This remark must not be understood as signifying any failure to appreciate the value of Mr. Thomas’ and Goneral Cunningham's writings,1884.) _V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 121 cient currency in that metal ;* and, as in modern times, a large propor- tion of the small change required may have consisted of cowries, and of tokens issued by private persons. The gold coinage of the Gupta kings is, on many grounds, of excep- tional interest. The great variety of type is remarkable, and suggests may problems in the history of art, religion, and nations. Though tone of the types are common, others are of extreme rarity, and to be reckoned among the most desirable treasures of the Oriental numis- matist. The proper attribution of several of the types is doubtful, and mpplies a theme for abundant discussion, and for the exercise of numis- matic acumen. The execution of many of the coins is of a comparatively high order of art, and the design is not unworthy of the execution ; while, in most of the types, both design and execution have such strongly marked national characteristics that they are far more interesting than - the mere imitations of foreign work which are found in the majority of Indian coinages. Foreign ideas are clearly traceable in this series of coins, but they are, in the best types, skilfully assimilated and Hindu- ined. The Hindt character of nearly all the Gupta gold coins is a plainly * General Cunningham informs me that, so far as he knows, only one copper coin of Kaméra Gupta has yet been found. It has not been published. Sir E. C. Barley (Num. Chron. for 1882 p. 168) mentions the Gupta copper coins as being “among the rarest of all Indian coins,” and expresses a belief that they “seldom cccur except in the immediate neighbourhood of the Gupta capital, Kanauj” (sic.] Copper coins, as Prof. Gardner has observed (Catal. of Seleucid Coins, p. zxzit), we very seldom dug up far from their place of mintage, and, therefore, if Sir E. C. Barley's belief aa to the provenance of the Gupta copper pieces is correct, the com- non opinion that Kenanj was the Gupta capital would receive some support. But, the evidence, #0 far as it goos, indicates that the copper coins, like those in gold, were coined further east. Prinsop describes six specimens, and of these three were from the cabinet of Mr. Tregear, who collected at Jaunpur. The other three were respectively in the Stacy, Swiney, and Prinsep collections, and it is not eaid that any of them came from Kanauj. I have not any further information as to the find-spois of the Gupta copper coins. Sir E. C. Bayley in the passage above quoted rather emggerates the rarity of the copper issues of Chandra Gupta II. Ten specimens ue in the British Museum, and one is in the India Office collection. The cabinet of the Asiatic Society of Bengal contains “many” similar to fig. 16 in Pl. XXX of Prinsep’s Eesays, one like fig.12 of the samo plate, and one of the ‘ vase’ type as figured ial. A.8. B. XXXIV, Pl. V. figs. 20, 21. General Cunningham and Mr. Grant Prsees specimens, and Mr. Theobald has a large coin, a duplicate of Prinsep's fig. 11, the obverse of which presente the king shaded by an umbrella, Farther specimens doubtless exist in the oabineta of other collectors. See Prinsep’s Essays, Vol. I, ‘Pp. 374-875, and Pl. XXX, figs. 11-15; Ariana Antiqua, Pl. XVIII, fig. 15 (the same @ Pringep’s fig. 14); and J. A. 8. B. Vol. XXXIV (1868) p. 125, and Pl. V. figs. 20 sad 21,122 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. (No. 2, legible record of a native reaction directed by the Gupta kings against the domination of the foreign Scythian idess.* The numerous, and frequently well preserved, legends on the coins of the Gupta dynesty offer much interesting material for the study of the historian and palmographer. In addition to all the above reasons which render attractive the study of the Gupta gold coinage, another is furnished by the chaotic state of the literature on the subject and the incompleteness of the existing catalogues, which loudly call for re-arrangement and revision. The difficulty experienced by myself in studying the coins with the help of existing publications first induced me to make an attempt to summarize and systematise tho known facts. The work has grown ander my hands, and, imperfect as it is in many respects, I trust that the labour bestowed upon it may not have been altogether thrown away. “No trouble,” says Dr. Burnell, “is thrown away, which saves trouble to others,”+ and, even if I have failed to solve any of the nu merous historical and numismatic problems suggested by the study of these coins, I can scarcely have failed in smoothing the path for investi- gators more fally equipped with the needful learning and technical experience. I have been encouraged in my undertaking by the recently expressed opinion of General Cunningham that “the gold coins of the Guptes require to be carefully re-examined.” So far as my opportunities permitted I have made a careful exe mination of this series of coins, and now submit the results of the investi- gation and tho opinions I have formed to the candid criticism of all competent judges in the hope that they will supplement my facts where they are incomplete, and correct my opinions where they are erroneous. I have endeavoured to work in the spirit of the words of Saint- Hilaire :—‘ La Numismatique est patiente, et elle amasse les faite spéci- aux qui ls concernent, jusqu’d ce que 1” histoire vienne plus tard en donner la véritable clef, si jamais elle le peut.”§ Section II.—Types and Devices. In Mr. Thomas’ valuable catalogues|| the several types and varieties are distinguished by an arbitrary alphabetical notation, for example, In the N. W. P. Gazetteer for Basti (Vol. VI, p. 718) the rise of the Gupta dynasty is absurdly described as » triamph of Buddhism over Hinduism. Sri Gupta may have been a Buddhist possibly, but certainly his successors were all Hinds. + Quoted in Max Miller's ‘ India, What oan it Teach Us,’ p. vii. t Proc. A. 8. B. August 1882, p. 113. § Journal des Savants for 1865, p. 418. || 5.4.8. B, XXIV, pp. 487-502; and Prinsop's Essays, Vol. I, pp. 377-987.1884.) V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 128 E,B},2E. Sucha notation is confusing and gives little assistance to the memory. I have ventured on novel nomenclature which will, it is hoped, prove appropriate and convenient. The devices, both obverse and reverse, of the Gupta gold coins display a remarkable amount of variety in conception and execution, and thus afford ample facilities for chasification. The obverse devices, when regarded with reference to the most prominent or characteristic feature in each, are readily divisible into 19 classes, of which few are common to two or more reigns. The reverse devices, when classified in s similar way, fall into but 9 classes, and are far less characteristic of the several reigns. It is evident, therefore, that the classification of types should be based, as it is in Mr. Thomas's catalogues, on the obverse devices. The main types are named and classified in my catalogue as follows, the name of each type being intended to indicate the most conspicuous, or most characteristic ele-- ment in the obverse device. The definition of each type will be found in the Catalogue. Reign. Type. I. Ghatot Kachs. Solar Stendard. (Pl. IZ; 1). Il. Chandra Gupta I. King and Queen. (Pl.II;2). I. Samudra Gupta. Javelin, (Pl. IT; 3, 4, 5). Archer. (Pl. II; 6). Lyrist. (Pl. II; 7, 8). Aéwamedha. (Pl. II; 9). Tiger. (Pl. IZ; 10). Boy and Battle-axe. (Pl. II; Uy, 12). Couch. (PI. IT; 13). Archer. (Pl. II; 14: Pl. IT; 1,2, 3). Lancer. (Pl. III; 4). Horseman to Left. (not fi- gured). Lion-Trampler. Pl. III ; 5). Combatant Lion. (PI. III; 6). Retreating Lion. (Pl. III; 7). Swordsman and Umbrella. (Pl. HIT; 8). Pauper IV. Chandra Gupta II. PP pe DIM * The term ‘archer coins’ has already been used by Wilson. (Vishau Pur, 1B 480, note 70.)124 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. (No. 2, V. Kumira Gupta Mahendra. 1. Swordsman. (Pl. III; 9). 2. Archer. (Pl. III; 10, 11). 3. Horseman to Right. (Pl. HI; 12). 4, Horseman to Left. (Pl. II; 13). B. Peacock. (Pl. IV; 1,2). 6. Lion-Trampler. (not figured). 7. Combatant Lion. (Pl. IV; 3). 8. Two Queens. (not figured). VI. Skanda Gupta. 1, Archer. (Pl. IV; 4). 2. King and Queen. (Pl. IV; 5). VII. Doubtful (Chandra etc.) 1, Archer. (Pl. IV; 7-10). » ~~ (Prakéséditya.) 2, Lion and Horseman. (Pl. IV; Ul, 12). I have not found it practicable to classify the subordinate varieties on any definite principle, and they are arranged as seemed convenient in each case. Tho 9 classes of reverse devices all agree in presenting as their main clement the figure of a female, associated with emblems which prove that she is intended to represent a divine personage. These reverse devices may be classified as follows -— Device. Reign. Type. I. a Standing goddess holding —_lotus- flower and cornu- copia. Ghatot Kacha. Solar Standard, » B. Ditto, holding fillet and lotus-flower, or fillet only. Chandra Gupte IT. Swordsman & TI. Goddess standing on Umbrella. dragon, holding standard and lo- tus-flower. Samudra Gupta. Tiger. Tl. Female (? goddess) standing, with fly- whisk, i:1884,]_V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 125 IV. Goddess seated on See Javelin. holding fillet and Archer. cornucopia, or fl- { Chandra Gupta II. Couch. let and Tota | flower. : Archer. V. Goddess seated (Chandra Gupta II. ‘Archer. cross-legged on ee Gupta Mahen- open lotus-fower, Swordsman. generally holding Archer. fillet and lotus- Two Queens. flower. Skanda Gupia. Archer. King & Queen, Doubtful (ail) Archer. » (2Prakééaditya.) | Lionand Horee- man. VI. Goddess, seated on wicker stool to left; » @ holding fillet and | Samudra Gupta. Lyrist. tus, or fillet and cornucopia. » B. holding fillet and lo- sceptre, or lotus only. Chandra Gupta IT. Lancer. . a Horseman to Left. Kuméra Gupta Mahendra. Horseman to Right. n y- feeding peacock and holding lotus. . 7 ” oe Lo» » ” » » Left. VI. Goddess riding pear cock. 7 Py » Peacock. VIL Goddess standing, feeding peacock. 7 7 » Combatant Li- on,126 V.A.Smith—Gold Coins of the Imperial Gupta Dynasty. (No. 2, TX. Goddess seated on tnckotomsest | lion ; " a. holding fillet and cornucopia. Chandra Gupta I. King & Quen. » B. holding fillet ond lotus, or fillet only, or lotus only. Chandra Gupta II. —-Lion-Trampler. ” oo Combatant Li- on, Le’ ” Retreating ,, Kumére Gupta Mahondra.Lion-Trampler. Prinsep quickly perceived that the ‘Kananj series,’ as he called the Gupta gold coinage, was ® continuation, and, to some extent, an imitation of the Indo-Scythian mintages ; and the intimate relation be- tween the two series of coins is well exhibited in Plato XXXVI of Vol. V of the Journal of the Asiatic Society of Bengal (Pl. XXIX of Prinsep’s Essays ; ed. Thomas). The same relation is more amply demonstrated by the series of plates in the Ariana Antiqua, and Wilson was rightly convinced (p. 418) that “the coins of the Gupta princes succeeded immediately to those of the Mithraic princes.” The fact of such immediate succession appears to my mind indisputable, and is in itself fatal to the theories of those authors who seck to date the imperial Gupta dynasty in the fourth and fifth centuries A.D. Iam convinced that to a certain extent the Indo- Scythian and the Gupta gold coinages were actually contemporary. Tho standing king, engaged in sacrificing at a small altar, who ap- pears on the obverse of the coins of Ghatot Kacha, is almost an exact copy of the corresponding figure on many coins of Kanerki and other Indo-Scythian princes. The altar appears again in the Javelin coins of Samudra Gupta, in the f variety of the Archer type of the same prince, and in the Swords- man and Umbrella type, which I attribute to Chandra Gupta II; and it is seen for the last time in the unique Swordsman coin of Kuméra Gupta. The supposition has been hazarded that the object referred to is a vessel containing the sacred Tulst plant (Ocymum sanctum), and not an altar, but comparison with the Indo-Scythian coins proves certainly that it is the latter. Moreover, in at least one specimen in the British Museam collection, the grains of incense falling on tho fire-altar are plainly indicated.1834.) V. A. Smith—Gold Ooing of the Imperial Gupta Dynasty. 127 The coins of Ghatot Kacha possess no distinctive Hindi character- istics. The king, who sacrifices at a fire-altar, grasps peculiar rose- hesded standard, which seems obviously intended to symbolize the rayed: sun. The Sun and Fire are in mythology almost convertible terms, and I think it may fairly be assumed on the evidence of the coins, that Gha- tot Kacha (though he may have been a Hinda) was a worshipper of the solar fire, as his Indo-Scythian predecessors undoubtedly were. I am also disposed to beliove that in most of the types of the Gupta gold coins the figure of the king on the obverse is intended to represent him idealized ae 8 god, and that in the case of Ghatot Kacha, he is represented in the character of the solar god, shedding beneficent influences upon his subjects. The standing goddess on the reverse bears a lotus-flower and cornu- copia. The lotus-flower is an emblem very commonly used in Hind my- thology, but is especially appropriate to the Sun® and to Sri or Lakshmi, the goddess of good fortune.t The Sun (Sérya) may himself be regard- ed ss a form or manifestation of Vishnu the Preserver, the lord of Lakshmi. The cornucopia undoubtedly indicates Western influence, but whether the design was borrowed directly from Greek, or Roman, or Syrian coins, it is not easy to decide. Cornucopis occur on the coins of the Seleucid dynasty of Syriat, but it is perhaps most probable that the device was borrowed directly from Roman aurei. In the Gupta series the cornucopia appears for the last time in the rare coins forming Class I of the Archer type of Chandra Gupta II, which were probably strack carly in his reign. It is perhaps possible that a close comparison between the forma of the Roman and the Gupta cornucopia might help in settling the great question of the Gupta dates.§ According to the chronology which I adopt, the last appearance of the cornucopia on the Gupta coins ia to be dated abont 240 A.D. I regard the standing goddess on the reverse of the coins of Ghatot Kacha as the equivalent of the Greek and Seleucid iy, and of the Roman Fortuna, and believe her to be a copy, in part, of the Fortuna Augusti and similar figures on Roman coins, and, in part, of the elemental goddesses on the reverse of the Indo-Scythian coins. If she must be given a Hindi name, I have no doubt that she must be named Sri or Lakshm{, the consort of Vishnu the Preserver. *“'The Indian mythology connected the lotus in all manner of forms with the mn.” Thomas in Num. Chron, for 1880, p. 26 note. Cf. Burgess Arch. Rep. for W. India for 1874-5, p. 216 and PI. LXV. + Birdwood, Industrial Arts of India, Vol. I, p. 58. 1 Gardner's Catalogue of Seloncid Coins, p. 46, Pl. XIV. Prof. Gardner informs ‘se that Seleucid ooins have been found in India. § See Thomas, Early Faith of Asoka, in J. R. A. 8. Vol. IX, N. 8. pp. 212-217. R128 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No.2 Certainly she cannot be intended for Pérvati.® I believe that she also may be regarded as representing the consort of the idealized king on the obverse, for it is a commonplace of Indian panegyric to represent Laksh- mi as the king's consort.t The standing goddess, holding fillet and lotus-flower, or fillet only, who appears on the reverse of the Swordsman and Umbrella coins of Chandra Gupte II with the legend ‘ VikramAditya’, is, perhaps, as sug- gested by Wilson, primarily intended to represent Victory, who so fre- quently appears on the Graeco-Bactrian coins ;t but she may be only 8 slightly varied representation of Lakshmi, and it is also possible that, at the time the coins were struck, her effigy connoted equally the ideas of Victory and of Lakshmi or Good Fortune. Different symbolic inter pretations are not necessarily mutually exclusive. I have followed Mr. Thomas in calling the object in the right hand of this personification a fillet, and, if the figure is intended for Victory, no doubt the nomenclature is correct. But, as we shall see, similar object constantly recurs in the hand of the female deities whose effigies aro displayed on the Gupta coins, and in many cases I believe it would be more proper to follow Prinsep (Vol. I. p. 230) in calling it pésa or noose.§ For convenience I shall use throughout the term ‘fillet,’ but it should be interpreted with regard to the qualification now stated. The goddess standing on what looks like a dragon or marine mon- ster (makara or jalampa) who is shown on the unique Tiger coin of Samudra Gupta does not appear to be intended for Lakshmi. In my remarks on that coin in the catalogue I have ventured to suggest two alternative interpretations of the symbolism. The Aéwamedha coins of Samudra Gupta were undoubtedly struck to commemorate the performance of the sacrifice of the horse, with the ceremonies which expressed the performer's claim to be the supreme power in India. These pieces agree in weight with the ordinary coins of the period, but in other respects rather resemble medals, and the con- jecture is allowable that they were issued as » special type of coin for * Mr. Thomas, however, describes her as “‘a rather clogant standing figure of P&rvat{, with the oxotio cornucopia.” (Epoch of the Guptas, p. 28, from J.R. A. 3. (W. 8.) 1881. The same learned writer thinks that the solar standard of Ghatot Kaoha mey signify a claim to solar descent. + E.. Aphsar inscription of lator Guptas, 1. 1. 8, 16, seqq. (J. 4. 9. B. XXXT, Pt. I, pp. 232, 234), and inscription from Nopél (Indian Ant. for 1880, p. 165). tT Ar. Ant,, p. 418. § Kittoo also uses the term ‘noose’ in his description of the Bharsar hosrd, (J, A. 8. B. XXI, pp. 390-400).1884.) V. A. Smith—G@old Coins of the Imperial Gupta Dynasty. 129 distribution among the Brahmans engaged in the ritual of the sacrifice.* Teannot guess at the exact meaning of the figure of the female with the fiywhisk on the reverse, but she is certainly intended for some sacred personage. Considering the undisputed solar character of Ghatot Kacha’s coinage it may not be irrelevant to allude to the connection which existed between the ASwamedha ceremony and Solar worship.t Itisquite possible that Samudra Gupte, though a good Hindd, may have been, as many Hindis still are, specially devoted to the worship of the sun, The legends of the King and Queen coins of Chandra Gupta I leave no donbt that the effigies on the obverse are primarily intended for the wovereign and his consort, for we know from the lapidary inscriptions that the name of the latter was Kuméri Devi, and that she belonged to the Lichchhavi family. The king is figured leaning on a apear, and this device may be intended secondarily to symbolize Kuméra Deva, the god of war, and husband of the goddess Kuméri Devi. The reverse goddess wated ona couchant lion is probably Dirgé, another form of Kuméri Devi, but the cornucopia in her left arm indicates that the deity is pre- tented under her beneficent, as well as her terrific aspect. The device of the king and queen standing facing each other re- appears in the coinage of Skanda Gupta, but in » much modified, and thoroughly Hinduised form. It has not yet been met with in the issues of any of the intermediate reigns. The unique coin of Kuméra Gupta lately discovered by Mr. H. R. Carnac (Proc. A. 8. B. Nov. 1883. p. 144), Presents the king standing between two females, whom I suppose to be his queens, The Javelin type is the commonest form of Samudra Gupta’s coin- age. The device of the obverse is but a slight modification of the ordi- nary Indo-Scythian pattern, and the throned goddess on the reverse isas obviously a copy of the figure called ApSoxpo or ApSoxpo on the Indo-Scythian coins of Kanerki and his successors. Mr. Thomas argues that this throned goddess should be identified With Parvati, the consort of Siva, for five reasons, of which the follow- ing iss summary :— (1). She is identical in form with the Indo-Scythian ApSoxpo or Aploypo whose name is commonly interpreted as Arddh-ogro (Wa}WT) or “half-Siva’, i. e., Parvati. (2) Even if it be admitted that the early Guptas had Vaishnava “Inthe northern Bilsay inscription, dated in the year 96, Kuméra Gupta is ‘lggized os the “giver of millions of gold, performer of the ASwamedha” &o, (Onna, Arch. Rep. XI. 20.) t Birdwood, Industrial Arts of India, I, p. 25130 V. A. Smith—Gold Ooins of the Imperial Gupta Dynasty. [No. 2, tendencies, the adoption from the Indo-Seythians of the reverse device in question “may well have been a mere act of ‘imitation of a foreign design,’ irrespective of any aim at demonstration of creed.” Reverse devices locally vary, and are not of much significance, e. g., the Sasea- nians retained the Siva and Nandi device of Kadphises, and the Muslim Ghaznavis retained the Hindi recumbent bull on their Lahor coinage. (3.) The female seated on lion, who appears on the reverse of four types of the Gupta coins, is plainly Parvati in her form of Dirgé. (4.) On four types the same goddess appears in the form of Kauméri Devi, associated with her sacred bird the peacock ; and (5.) Skanda, the name of the last of the imperial Guptes, is an alias of Kuméra Deva, the god of war, son of the goddess Kuméri Devi.* These arguments seem to me to be of little weight. The interpre- tation of Ardokro or Ardochro as meaning ‘half Siva’ isa very forced one, and I doubt greatly if such a compound as W&TWT, or rather Witt could have in Sanskrit the meaning assigned to it. The name is never written Ardogro, whereas the title of Siva which is supposed to form an clement of the compound is Ugre, and I do not see how the ‘g’ can be converted into x or x, nor why the aspirate at the end of arddha should be lost. The supposed compound ‘ Arddhogra’ has no analogy with the genuine compound ‘ Arddhanéri’; it is one thing to speak of a creature as half-female, and quite another thing to speak of Joan as half-John.+ ‘The Indo-Scythian goddess may or may not be intended to represent Pérvati, though Ido not believe that she was, but I am convinced that her name does not mean ‘ half-Ugra,’ and that such a compound never existed. The name ApSoxpo or ApSoxpo is probably a Scythian name, and not an Indian word at all. If the throned figure is to be identified with any goddess of the modern Hindi pantheon, I consider that she should be identified, as suggested by Wilson, with Sri or Lakshmi, the benign goddess of for- tune, and not with the terrible Pérvati. The supposed Vaishnava tendencies of the early Guptes have been believed in chiefly on the testimony of the Bhitari pillar inscription, which, if correctly interpreted by Dr. Mill, proves Chandra Gupta IT and Kuméra Gupta to have been Vaishgava, and Skanda Gupta to have © J. A.8. B., XXIV (1855) pp. 489-490. + Cf. Wileon’s criticisms in Ar. Ant., pp. 361-362. In tho Pa-Shake coin in the British Museum the name of the goddess is spelled OPAOX[PO], a form which it is absurd to identify with ‘ Arddhogro” (This uniquecoin is described in Mr. ‘Thomas's Indo-Scythian Coins with Hindé Legends, p. 11.) General Cunningham con- curs with me in giving the naine of Lakshmi to the goddess, whether seated on the throne or the lotus-flower.1884.) V. A. Srnith—Gold Coins of the Imperial Gupta Dynasty. 181 beon Saiva. But the translation of the Bhitari inscription is avowedly imperfect, and, until it has been revised by a competent echolar, is of little use for historical purposes.* The interpretation of the device of the throned goddess requires no assumption as to the sectarian preferences of the early Guptas, for the attributes of the figure are manifestly those of Lakshmi rather than of Pérvati, and I venture to affirm that but for the ‘ half-Siva’’ interpre- tation of the word Ardokro, no one would ever have thought of calling the figure Pérvati. The suggestion that the figure of the Ardokro god- dess was adopted by Samudra in mere imitation of a foreign design doos not appear to be tenable. The coins with this reverse undoubtedly show evident traces of foreign influence, but they are far from being mechani- eal copies of alien designs. If Samudre Gupta’s die engraver had been & mere copyist he would naturally have copied from the coins of Samu- dm’s father and grandfather, but the reverse devices of their coins are tolally different both from the Ardokro figure and from each other. Samudra himself employed four distinct reverse devices, and evidently adopted each of them deliberately. Mr. Thomas’ remaining evidence in favour of his interpretation con- ste in proofs of the Saiva preferences of Kuméra Gupta and Skanda Gupta. But the facts that one of these princes placed on his coins effigies of Kaméri Devi and of Durgé, and that the name of the other is a syno- nym of Kuméra Deva, by no means prove that all female figures on the reverses of other Gupta coins are intended for forms of Pérvati. I have diseussed above some of the representations of standing goddesses, none of whom can with any probsbility be identified with Parvati. The pea- cock of Kumari Devi, and the lion of Dirgé are never associated with the throned Ardokro goddess. She occurs only on the Javelin and Archer coins of Samudra Gupta, and on the unique Conch coin, and the tare coins forming Class I of the Archer type of Chandra Gupte II. An emblem, which is very characteristic of the Gupte gold coins, makes its first appearance on the obverse of Samudra’s Javelin type, This is » standard bearing on the top the figure of a bird, and having a general resemblance to a Roman eagle standard. Wilson (who is followed by General Cunningham) was inclined to interpret the bird as meaning Garuda, the winged vehicle of Vishnu; but this interpretation appears to me forced and improbable. The object indicated is simply a bird, whereas the mythologists describe Garuda © For the Bhitari inscription see Prinsep's Essays, Vol. I, pp. 240, seqq. A ‘Ferieed facsimile is given in Cunningham Arch. Rep. I, pp. 97—99, and PI. XXX. A welledited translation is much wanted, and it is surprising that the want has re- Mained 10 long ansupplied.132 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. (No. 2, as @ monster, half man and half bird. I prefer Mr. Thomas’ former opinion that “the most natural and obvious interpretation is to look upon it as designed to represent the peacock, which appears with such fre- quency on the gold coins, and occupies the entire reverse field of ono type of the silver coinage.”* It is, however, quite possible that the emblem is merely a copy of the Roman eagle, and the term ‘ bird-standard,’ which involves no theory, is the safest to adopt. In his Archer type Samudra Gupta substitutes for the javelin in the king’s hand a bow, and the device thus introduced long remained the favourite obverse pattern. It is found on the coins of Chandra Gupta II, Kumére Gupta, and Skanda Gupta, and is, with few exceptions, the only design used by the rade imitators of the Gupta types, some of whose coins are noticed in the Supplement to the Catalogue. It seems impossible at present to decide whether the Archer device was an independent invention, or was borrowed from Persia or some other foreign source, and it is equally doubtful whether it has or has not any symbolic meaning. If it has, it may be regarded as another expedient for indicating the analogy between the eun that rules the heavens, and the king who rules the earth. Chandra Gupta II issued gold coins of at least eight different types, but specially favoured the Archer type, specimens of which in large numbers have been found. The Lyrist type of Samudra Gupta’s coinage, which depicts the King as a musician playing the Indian lyre, is interesting in several respects.¢ The type is rare, and the specimens known are mostly in fine condition, and, with the exception of the India Office example, are broad thin coins well struck, but in singularly low relief. The dress of the king is thoroughly Hindu, but his attitude recalls that of the king on the Indo-Scythian coins classed as ‘couch-loungers’ by Prinsep. The re- verse device is likewise in appearance completely Hindi, though ap- parently suggested by foreign models. Tt consists of a female seated sideways to the left on a wicker stool, and holding fillet and cornucopia. The attitude of the goddess, and the form of the stool on which she sits recall the device of Apollo seated on the Supados , with its cover of the dypmviv net, as seen on the Seleucid coins of Syria,t and I believe that © J. A. 8. B. XXIV, (1855) p. 494, note. In ‘Records of the Gupta Dynasty” (1876) p. 2 ‘Mr. Thomas adopts the Garnda interpretation. + Line 240f the Allahabad Pillar inscription mentions Samudra Gupta’s accom- plishmente in singing and playing. (Prinsep's Essays, pp. 233 veaq.) t E.g. the coins of Antiochus I, figured in J. A. 8. B. Vol. L. for 1881, p. 178, and Pl. XVIII, 14, 15. General Cunningham calls the scat ‘cortins, but ‘om- phalos’ is more correct,1984.) V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 183 the resemblance is not accidental ; but the closest parallel to the Gupta- device is met with in an unexpected place. The goddess on the Gupta coins is almost an exact copy of Demeter as represented on a rare coin of the island of Paros, now in the British Museum, and the resemblance is s0 close that it is scarcely possible to doubt that in some unknown way both devices must be derived from @ common source. The cornucopia in the band of the goddess of Samudra Gupta’s coins shows that sho was intended to have attributes similar to those of Demeter, and she may therefore be regarded as novel representation of the Hind Lakshmi, the counterpart of the Greek goddess. The same reverse device, but with some modifications, and asso- ciated with other obverse devices, was adopted by Chandra Gupta 11, and Kumira Gupta. The goddess, as she appears on the Lancer and Horseman to Left coins of Chandra Gupta II, and in varieties aand 8 of the Horseman to Right type of his son, would seem to be intended to symbolize nearly the same ideas as the effigy on the Lyrist pieces of Samudra. In variety y of Kuméra’s Horseman to Right type, and in all the Horseman to Left coins of the same king, the goddess is repre- sented in the act of feeding # peacock, and may, therefore, be identified ss Kuméri Devi, to whom that bird is sacred. In the gold coinage the peacock (except, perhaps, as part of the so-called ‘ peacock standard ') appears to be peculiar to the mintages of Kuméra Gupta Mahendra. The goddess on the reverse of his Combatant Lion type stands while she feeds the sacred bird. In his Peacock type the bird is still more prominent, for on the obverse the king is feeding one peacock, and on the reverse, the goddess, presumably Kuméri Devi, Tides on another. There can be little doubt that in this type at all events the king is presented in the double character of the human king and the divine Kuméra Deva. The peacock devices of the Gupta coinage ap- peer to be Hinduized adaptations of the designs of the Roman coins which bear representations of the peacock associated with Juno, or with a deified lady of the imperial house. An exact prototype of the peacock with expanded tail, which is found on the silver Gupte coins, and on var. B of Kuméra Gupta’s gold Peacock type, may be seen on the reverse of a coin of Julia Augusta, who was a daughter of Titus and died be- twoon A. D. 81 and 90. A coin of Paulina (A. D. 217-238), whose life probably extended into the early years of the reign of Chandra Gupta II, exhibits the pes- cock in a manner strikingly similar to the device on somo of the silver * Tréeor de Numismatique, Iconographio des Empereurs Romains ; Pl. XXII, LL134 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No.2 coins of Kuméra Gupta. Another coin of Paulina’s represents her in the character of Juno riding on a peacock, but the treatment of the sub- ject differs from that used by the Gupta artists.* A standing peacock, like that on the first mentioned coin of Paulina’s, appears on a coin of Mariniana, (circa 250 A. D.)+; and a coin of Manlia Scantilla Augusta (193 A. D.) exhibits a standing figure of Juno with sceptre in left hand, and holding in her right hand a patera over » peacock standing at her feet.t In the Boy and Battle-axe type of Samudra Gupte it is interesting to observe the reminiscence of Scythian influence in the form of the battle-axe, with which the king is armed, as representing the god of Death. The place of the usnal bird on the top of the standard is taken by acrescent moon. The same crescent-tipped standard occurs on the reverse of the unique Tiger coin of the same king, on the obverse of which the king is depicted as slaying a tiger.§ The obverse device of this coin is the model of three types of Chandra Gupta II, and two of Kuméra Gupta Mahendra, in which the tiger is replaced by a lion. I believe that these devices had some sym- bolic meaning but am not able to make it out. They may have been suggested by the Greek representations of Hercules contending with » lion. In the Archer type of Chandra Gupta II we first meet with the reverse design No. V, which subsequently became common conventional pattern, and was used almost exclusively by the obscure princes who rudely imitated the Gupta coinage. The device consists of the figure of a goddess facing front, seated cross-legged on an expanded lotus-flower, and holding in her left hand a lotus flower, and in her right the ‘fillet’ or ‘noose.’ The scholars who give the name Pérvati to the Ardokro goddess, of course bestow the same name on the lotus-throned divinity, but I cannot perceive in the latter device any symbolism specially sug- gestive of the attributes of Parvati, whereas the symbolism used is thoroughly appropriate to express the ideas personified as Lakshmi. In justification of my views regarding the symbolism of the reverse devices of the Gupte coins I may appeal to the following description of tho at- tributes of Lakshmi, which is based on the best authorities :—‘ Laksh- © Ibid, ibid Pl. XLVILI, figs. 6 and 4. + Ibid. sbid,, Pl. LI, 8. Ibid, ibid., Pl. XLI, 1. § In his Records of the Gupta Dynasty (1876) p. 21, Mr. Thomas calls the en- sign a ‘Garuda standard,’ bat I nm satisfied (after examination of the coin), that the object on the top of the standard is rightly describod as a croscont in the samo author's Revised Catalogue (1858).1684.) V. A. Smith— Gold Coins of the Imperial Gupta Dynasty. 135 mi, called Sri, is Vishnu’s éakti, She is the goddess of good luck and plenty...She is worshipped by filling the corn-measure with wheat or other grain, and thereon placing flowers. She is represented as a lovely and benign woman, robed in yellow, holding a lotus in her hand, and seated on a lotus, or beside Vishou. Sometimes, as is likewise Vishnu, she is painted all yellow, and has four arms, and she holds in one of her right hands a rosary, and the pdfa or cord in one of her left. ‘This cord is seon also in the bands of Varana and S'iva, and is emblematical of the sea, which girds the earth.”# It is impossible to read this description, and not to see that it is in remarkably close accordance with the delineation both of the Ardokro goddess, and of the lotns-throned divinity. But it is quite inapplicable to Parvati as ordinarily conceived, and the symbolism of the two coin- devices in question is equally inappropriate to the stern and terrible guddess. . T have therefore no doubt that the goddess who is seated on a throne in Samudra’s coins, on a lotus flower in the coins of Chandra Gupta II and his successors, and also (in certain cases, as already specified), the divinity seated on the wicker stool, are all intended to express substan fially the samo conception, that of the benign and kindly Good Fortune, the bestower of happiness and plenty; the samo who was named rix7 and Demeter by the Greeks, and Fortuna, Cores, Abundantia, ete. by the Romans. Although I have been at so much pains to distinguish between Pir- vati and Lakshmi, I am aware that the two concepts sometimes coalesce, and become indistinguishable. Tho names and attributes of gods and goddesses, in India or elsewhere, are all nothing more than the feeble efforts of the human imagination to express by metaphor and symbol imperfectly apprehended ideas of the attributes of the unspeakable divine nature, and it is futile to attempt to draw sharp lines of demarcation bo- tween these symbolical expressions. Now one, and now another idea predominates in the symbolism, and “in any lengthened description of one Hind deity it is amost impossible to avoid mixing up its character and attributes with those of another.”+ Nevertheless, the ideas per- wonified severally as Lakshmi and Parvati are ordinarily kept quite dis- finet, and nothing but confusion of thought ean result if tho namo of Pérvati is given to a personification possessing all the attributes of Lakshmi. © Birdwood, Industrial Arte of India, Vol. I, p. 68. t Birdwood, Industrial Arts of India, Vol. I, p. 59. As‘ Anna Pirgé,’ Pirvatl is identical with Lakshmi, ib. p. 61. 8136 V. A. Smith—Gold Ooins of the Imperial Gupta Dynasty. (No. 2, The only Gupta kings who appear in the coin devices as mounted on horseback are Chandra Gupta II and his son Kuméra Gupta Mahendra. In the later coins of Prakéédditya the device consists of a horseman slaying a lion or dragon, but the execution of the design is very poor. Tho rare Lancer coins of Chandra Gupta II are designed and executed with considerable freedom and spirit. The device may be an imitation of the very similar device on certain Macedonian coins, transmitted through intermediate channels.® The rayed turban or hel- met of the king in one specimen (dr. Ant, XVIII, 17) was perhaps suggested by the rayed head of Antiochus Epiphanes.t It is noticeablo that a crescent is found in the field, either on obverse or reverse, of each of the four Lancer coins known to me. The Horseman to Left coins of Chandra Gupta II, which are also very rare, resemble generally his Lancer coins, but the horse is turned to the left, the lance is wanting, and there is no crescent in the field. Kuméra Gupta Mahendra copied both these types of his father's coinage, but with some modifications. His Horseman to Right coins correspond with his father’s Lancer coins, the lance being omitted, and his Horseman to Left coins differ from the closely similar coins belonging to his predecessor chiefly in the insertion on the reverse of the peacock, the especial emblem of Kuméra Gupta.t The fact that Chandra and Kuméra Gupta used indifferently dies in which the horseman was turned to left or right is worth noting, because @ change in the direction of an obverse head on the coinage has sometimes been regarded as an indication of a change of dynasty.§ In some specimens of the curious Lion and Horseman coins of Prakéséditya a small bird-standard is seen over the horse’s head. The meaning of the character below the horse in this type, which seems to be intended for ¥ ‘u', is not known, Tam well aware that the foregoing account of the types and devices of the Gupte gold coins is far from being complete and satisfactory, but it is the best that I can give at presont, and may prove the means of stimulating further research. ‘Tho attribution of the several disputed types is discussed in the Catalogue. © For mch Macedonian coins see Mionnet, Pl. LXX, 8, and Trésor de Numis- matique (Rois Grecs), Pl. VIII. + Catalogue of Seleucid Coins, Pls, XI and XII. 1 Of. “That King gave birth to o con, oven aa did Hara to tho rider of the pencock (scil. Kértikeya or Kuméra tho god of war). Forward in battle and re- nowned strength, this son was named Kuméra Gupta," (Aphear inscription of later Guptas, line 7: in J. A. 9. B. XXXV, Pt. I, p. 278). § Eecords of tho Gupta Dynasty, p. 61, with referonce to Toramépa's coins.1884.) _-V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 137 Section IIT. Monocrammatic EMBLEMS. The so-called monograms (with one doubtful exception) occur only on the reverse of the Gupta gold coins, and, when present, are generally placed over the right shoulder of the goddess. The forms assumed by these monogrammatic emblems on tho coins accessible to me are shown in Plate IV. The most common forms con- sist of a horizontal line, or two parallel lines, surmounted by either three or four dote or short prongs, and having a square or lozenge attached be- low by one corner. Sometimes the square or lozenge is replaced by cross, and some- times by other devices, and occasionally the prongs or dots above the hori- zontal line or lines are wanting. One form (No. 25), which I know only from 8 drawing, departs altogether from the standard pattern. Exami- nation of the plate will show the large variety of minor modifications in detail which occur. ‘What ie the origin and meaning of these mysterious marks ? To this question I can give no positive and satisfactory answer, but Iam not without hope that the distinct enunciation of it, and the syste- matic presentation of the monogrammatic emblems as they actually occur may suggest to other enquirers the correct solution of the problem. The following statement exhibits the monograms which have come under my observation, arranged according to reigns :— Ghatot Kacha... Nos. 1; 2; 4a, Chandra Gupta I » Ba; 4b; 5; Bd; 226, Samudra Gupta » 8a; 40; 60; 6b; 80 9; 11; 196; 20a; 200; 21; 22, » 8a; 8b; 4c; 7a; 7b; Ba; 8b; 10a; 10b; 10c; 12; 15; 16; 17a; 175; 18; 19a; 19b; 208; 21; 22; Chandra Gupta II... 23; 24, Kuméra Gupte Mahendra’. 84; 8b; Be; 10c; 170; 17d; 19b ; 20a; 25. Skanda Gupta n» 8a; 8b; 4cP; Ba. Doubtful... +» 8a; Be; 10a; 13; 14; 190, The following types Samudra Gupta .. Chandra Gupta I. Lancer, var. @138 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. (No.2, Chandra Gupta II Horseman to Left. Lion-Trampler, var. 8. Horseman to Left. Peacock. In tho following types the monogram is sometimes present, nd sometimes wanting :— Samudra Gupta... Chandra Gupta IT Swordsman and Umbrella. Kumira Gupta Mahendra Horseman to Right. Wo learn from the last two lists that the monogram was not indis- ponsable, and was frequently omitted, though more usually inserted. The monogrammatic devices on the Greco-Bactrian coins, with which the Gupta mint-masters must have been familiar, are real mono grams, combinations of letters, usually those of the Greek alphabet. The so-called monograms on the Gupta coins, and the similar ones on the Indo-Scythian mintages, are certainly not combinations of al- phabetical characters, and the application to them of the word monogram, which has become usual, is, strictly speaking, a misnomer. Kittoe pre- ferred to designate them by the term ‘emblem’, but that word is incon- veniently vague, and, for want of a better term, I follow the ordinary practice, and call the marks in question monograms. Fow, if any, of the forms of tho Gupta monograms are exactly the same in every detail as those met with on the Indo-Scythian coins, but the general appearance of the monograms on the two series of coins is obviously identical, and many of the Gupta forms are only trivial variations of the Indo-Scythian patterns. Consequently, whatever interpretation is given to the Gupta mono- grams must be sufficiently comprehensive to include the analogous and similar Indo-Scythian ones. It appears to be established that some of the Greco-Bactrian mono- grams are names, more or less abbrevinted, of mint-cities. General Cunningham's ingenious interpretations of @ large number of these monograme cannot be implicitly accepted, but the proposition that some of tyose which “are common to a number of different princes” express the names of the mint-cities may safely Be admitted. Others probably indicate the names of mint-masters or other functionaries.* ‘The monogrammatic emblems on tho Indo-Scythian and Gupta coins look as if intended to take the place of the Greco-Bactrian mono- grams, and the hypothesis that they bear the same meaning or meanings Kumére Gupte Mahendra Lyrist. © Coine of Alexander's Successors in the East, in Num. Chron, N.S. VIII (1868), pp. 185 99.1884] V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 189 maturally suggested itself. Wilson noticed that the three and four- pronged patterns of monogram were continued from the coins of the Indo- Scythian sovereigns Kadphises, and Kanerki on those of the Gupta kings, snd observed that “agreeably to the purport which there seems reason to assign to these monograms, the recurrence of this emblem on all these coins should denote the place of their coinage.”* But he hesitated to adopt this theory because it appeared to him that the Indo-Scythian dominions must have lain far to the northwest of the Gupta kingdom, and he suggested the alternative hypothesis that the Gupta monograms might be merely “a proof of imitation” of the Indo- Scythian coinage, and “introduced without any definite object.” Such » muggestion is, on the face of it, improbable, and it is at onco disproved by a carefal examination of the monograms. A mere copyist would have tried to copy the Indo-Scythian monograms as they stood, and, however he might have failed in the mechanical execution, the evidence of the attempt to copy would have been unmistakeable. But, as I have already remarked, and as any one can readily verify by comparing my plte of monograms with that in the Ariana Antiqua, the Gupta mono- grams, while following the Indo-Scythian in the general pattern, differ indetail, and it is incredible that the systematic variety which exists could bo the result of chance caprice. Moreover, the mechanical execu- cation of the Gupta monograms is nowise inferior to that of the Indo- Scythian, No one can study the designs of the better types of the Gupta gold coinage without seeing that the artista who cut the dies, though indebted in some respects to foreign models, yet possessed considerable originality, and knew how to assimilate and nationalize the conceptions of alien art. The hypothesis that the Gupta monograms are the work of blind and unintelligent imitators may therefore be dismissed without doubt or hesitation. The hypothesis that the monograms indicate the mint-cities is much nore plansible, but is not altogether satisfactory. The Indo-Scythian coins are found chiefly in the Panjab and neighbouring parts of Afghén- istin where Gupta coins are never found,t whereas the Gupta gold coins, 4s will be proved in a subsoqyent section, have been found for the most pert in the province of Benares and the neighbouring districts. It is extremely improbable that the Panjab Indo-Scythian and the Gupta coins should have issued, to any considerable extent, from the same mints, or should bear cognate mint-marks. Indo-Scythian coins of Kadphises and © Ar, Ant. p. 418. t In Arch. Rep. XIV. p. 65, Goneral Cunningham mentions the finding of one Gupta coin among upwards of 1,000 of other kinds at Sunit near Ludidna in the Panjéb,140 V. A. Smith—Gold Ooins of the Imperial Gupta Dynasty. [No. 2, Kanorki are, however, found in N. E. Oudh and Benares, and it is pos sible that certain of the Indo-Scythian provincial mints may have been occupied by the Gupta kings when they shook off the Indo-Scythian yoke, and that the ‘ monograms ’ on the Gupta and eastern Indo-Scythian coins may indicate mint-cities. Unfortunately no detailed catalogue of Indo-Scythian coins has yet been published, and the statistics of their provenance have not yet been analysed. The occurrence of coins together ina hoard raises presumption that they proceeded, if not from a single mint, at least from mints not very far distant from each other. Few details as to the components of the various hoards of Gupta coins are available, but when such details are Imown, we find very various monograms associated in a single hoard. Thus, the 32 described coins of the Bharsar hoard exhibit monograms as follows :—* 1 coin of Samudra Gupta; 6 of Chandra Gupta IT; 1 of Kuméra Gupta Mahendra; and 2 of Prakiééditya ; total 20. 2 of Samudra Gupta. 8 of Samadra Gupta; 6 of Skanda Gupta; total 9. 2 of Chandra Gupta II. No. 25. 2 of Kuméra Gupta Mahendra, No monogram 1 of Chandra Gupta II; 6 of Kuméra Gupta Mahendra; total 7. Grand total 32. The above considerations seem sufficient to throw doubt on the theory that the Gupta (and consequently the Indo-Scythian) mono grams are the indications of mint-citics. Nor does it seem possible that they should be the marks of mint- masters or other official persons, for the samo monogram runs through several reigns. For example, the monogram No. 3a is found on coins of Chandra Gupta I, Samudra Gupta, Chandra Gupta IT, and Skanda Gup- ta, and its uso, therefore, continued for at least a hundred years. If then these monograms are not the result of blind imitation, nor the devices of mint-cities, nor the marks of public functionaries, what are they? It seems to me most probable that (though they may be mint-marks) they are religious emblems or symbols of some sort. The description of types in the last preceding section will have left no doubt on the reader’s mind that religious symbolism and the effigies of deities ap- pear everywhere on the Gupta gold coins, as they did on their Indo-Sey- thian forerunners, and it is reasonable to suppose that the same love for religious symbolism dictated the sclection of tho so-called monograms. © For an account of this hoard, see post, Sec. V.1884.) V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 141 I cannot profess to explain the precise significance of any of the Gapta monograms, but it is possible that some Hindd scholar may be able to elucidate the subject. Mr. Thomas has called attention to the curiously close likeness between monogram No 4a, and the Egyptian symbol for the bee, which was the sign royal in the Hieratic character.* Atrident which bears a resemblance to some of the Indo-Scythian and Gupta monograms occurs, detached like them, in the field of a coin of Rhescuporis II, king of the Bosphorus (A. D. 17 to 34).¢ The standing figure of Victory, who appears on some coins of Azes holds in her right hand a four-pronged symbol which is identical with the upper part of so many of the Indo-Scythian and Gupta monograms. These instances of resemblance between the monograms in question and other symbols may be cases of casual coincidence, but I have thought it worth while to note them on the chance of their suggesting a correct solution of the problem of the origin and meaning of the so-called monograms of the Indo-Scythian and Gupta dynasties. Section IV. Weicuts. The authors of essays on Indian numismatics have in general con- tented themselves with more or less complete descriptions of the devices and legends of coins, and have paid little attention to weighments. Numismatists in Europe of late years have become alive to the importance of dry details of the weight of coins, and have spared no Pains to obtain copious lists of weights as materials for induction. A knowledge of the weight standards of ancicnt coins is indis- pensable for the attainment of accurate notions respecting the history and development of coin types, and helps to throw light on the ill-un- derstood commercial relations of the states of the ancient world. The tholar who devotes himself to the examination of the numismatic treasures of Europe cannot hope to do more than fill in the blank spaces of a sketch which has already been drawn in firm outlines by the hand of histury. The enquirer who ventures to explore the labyrinth of Indian numismatics can expect but little help from the friendly band of the historic muse, but is perhaps compensated for the difficulties which heencounters by the unfailing hope of discovery, and by the consciousness that he is tracing the plan of the foundations on which history should rest, * Records of the Gupta Dynasty, p. 21, note. t Trécor do Numismatique, Rois Grecs, Pl. XXV, 12. 1 Ar. Ant, PL VI, fige 12 ond 18,142 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. (No.2, So general has been the neglect in Indian publications of all sys- tematic study of coin weights, that I may be pardoned if I dwell for s moment on its importance, and call to witness an expert who has studied European and Oriental numismatics with equal ardour. “The history of the standards of weight on which Greek coins were strack did not, until quite recently, become a subject of serious study. wan Nothing has done more of late years to give a scientific form to Greek numismatics than the great attention given to weight standards. ‘The fact has been recognized that a coin is, after all, but a stamped piece of precious metal, and that its value was derived, when it was issued, not from the stamp, but from the metal. Distinguished scholars like Hultsch and Brandis have in consequence spent years of their lives in weighing coin after coin, recording the results, and trying thence to reach principles. The greatest of living archmologists, Professor Mommsen, has given much time to the study of the weights and developments of Greck and Roman coins, and his strength has opened a way through jungles which were before impenetrable obstacles to science.”* It cannot be expected that Anglo-Indian amateur numismatiste should devote years of their lives to weighing coins, but, even with such limited opportunities as circumstances permit, they may collect a goodly mass of the necessary details, and do something to give to Indian archeology that scientific form which it frequently lacks. Tho weights of all coins mentioned or described in my catalogue are there noted, so far as they could be ascertained, and the results are ex- hibited in the Table of Weights, which deals with 177 coins. Examination of the devices has already proved that the Gupta gold coinage immediately succeeded that of the Indo-Scythian princes, and this conclusion is confirmed by the study of the coin weights. Few details as to the weight of the Indo-Scythian coins are available, but, according to Mr. Thomas, the coins of the Kadphises group average 122'4 grains, while those of the Kanerki series are somewhat lighter, but often weigh 122 grains. Some Indo-Scythian pieces weigh as high as 125 grains.t The source from which the Indo-Scythians derived the supply of gold for their extensive mintages is not known with certainty, but is conjectured, and with much probability, to have beon the constant stream of Roman aurei which in those times poured into India in exchange for her silk and other commodities. © Types of Greek Coins by Percy Gardner, 1888, p. 62. + Early Faith of Atoka (J. R. A. 8. IV N. 8. p. 223). Ibis possiblo, and even very probable that tho Indo-Soythian and Gupta Dynasties and coinages to a certain eatent existed contemporancously in difforent parts of the N. W. P. and tho Panjab.1884] V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 143 The existence of this eastward drain of gold is fully proved by the testimony of Pliny, as well as of other witnesses. The words of Pliny are so vivid and explicit as to be worth quoting afresh. “ Minimique empatatione millies centena millia sestertifim annis omnibus India et Seres peninsulague [scil. Arabia] imperio nostro adimunt, Tanta nobis delicise et feminae constant.”® Again he observes that the trade with India was worth taking some trouble to maintain. “Digna res, nullo anno imperii nostri minus H. S. quingenties exhauriente India, et merces renittente, quse apud nos centuplicato veneant.”+ The aureus was adopted first by Julius Cesar as a regular element of the Roman currency, and his standard is said to have been 125-66 gins but his coins generally range between 120 and 125 grains. It would therefore appear that the Indo-Scythian gold coinage is based on that of Julius Cesar, and not on the Macedonian stater, or Persian daric, of which the standard was 1344 grains, or two Attic drachmae. This fact helps in some measure to settle the vexed question of the date of the Indo-Seythian kings, and consequently of their Gupta successors. The weight of the Roman aureus after the death of Julius Cmsar gradually declined, and in the reign of Nero is stated to have averaged 11539 grains.t The average weight of 4 coins of Ghatot Kacha is 114-95, and the heaviest coin weighs 118. The average weight of the aurei of Augustus in the British Museum is 121-26, and it would therefore at first sight spear as if tho coins of Ghatot Kacha were based on the Roman coinage intermediate between Augustus and Nero. But a fine coin of Chandra Gupta I, son and’ successor of Ghatot Kacha, which is in the British Museum, weighs 123'8, and this fact indicates that Chandra Gupta’s coinage was adapted to a standard of about 125 grains, and renders it probable, though not certain, that Ghatot Kacha followed the same stand- ad, I assign the coins of the King and Queen type alone to Chandra Gupta I, and the weight of 4 of these averages 117°57. The light weight of the majority of the coins of Ghatot Kacha and his son appears to be due to wear and tear. The details for the weights of the six types of Samudra Gupta’s © Pliny, Hist. Nat. XII, 41. + Pliny, Hist. Nat, VI, 26. } The average weights 12566 and 116'39 for Julius Cesar and Nero respectively ae those stated by Letronne, as quoted in Smith’s Dict. of Antiq. and in Thomas’ Yarly Faith of Asoka, uf supra. Mr. Gardner informs me that the aurei of Julius Cear average 120 to 125, and those of Nero 112 to 114. I adopt Gen. Cunning. ham's estimate of the weight of the daric; Mr. Heard makes it 130 grains. TU4 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No.2 coinage will be seen on reference to the table. The heaviest coin of his reign is one of the Boy and Battle-axe type, which weighs 123-4, and the next heaviest is a Lyrist coin weighing 122 grains. The 5 specimens of the Lyrist type weighed are all in good condition, and yet exhibit » remarkable variation in weight from 11] to 122 grains, of which I cannot offer any explanation. The B variety of Samudm's Archer type is remarkable for ite light weight, the highest weight being 114 grains. The Aéwamedha coins average 116'18, and do not exceed 117-7, but all specimens weighed are more or less worn. The mean of the weights of the heaviest coins, one of each type, is 118'87, for the reign of Samudra Gupta. With tho exception, perhaps, of the f variety of the Archer type, I do not believe that the weight standard was intentionally lowered during this reign. The coins of Chandra Gupta II are somewhat heavier, but for the most part follow the same standard as those of his predecessors. The Wheel coins (Archer type, class II 8) form a remarkable exception, the highest weight (two specimens) being 1325, and the average weight of 8 coins being 129°77, which figures agree substantially with those for the reign of Skanda Gupta, It would seem as if these Wheel coins were struck on the daric or Macedonian stater standard of 134-4 grains. I can offer no explanation of this fact, but I believe that it is an indication of some im- portant historical event. These Wheel coins of Chandra Gupta’s and the coins of Skanda cannot be intended as equivalent for Roman aurei, for the heaviest known aureus is one of Pompey, weighing 1282. Itis possible that the immediate model of the coins in question was found in the issnes of the Seleucid kings of Syria, which frequently weigh 130-132 grains, and are sometimes found in India. A few coins of the Archer type, class II a (the commonest variety) and of the same type and class var. y, exceed 125 grains, the heaviest specimen weighing 127-6, but the average for the type (excluding the Wheel variety) is about 123 grains, and I believe, therefore, that the coins were intended to follow the old Roman and Indo-Scythian stand- ard of about 125 grains. The mean weight for the reign, calculated as in the case of Samudra Gupta, and excluding the Wheel variety, is 121-61. In the reign of Kumdra Gupta Mahendra the weight standard was certainly to some extent raised, the mean weight for the reign, (cal- culated in the same manner as above) being 126-0 grains. The heaviest coin of the reign is one of the Peacock type, weighing 1286, and very few specimens of any type weigh less than 123 grains. The standard would therefore seem to have been the anciont Lydian standard of 1301884.] V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 145 grains. Why Kuméra Gupta should have reverted to this standard for his coinage is at present an unsolved problem. Skanda Gupta’s coinage ovcars in two types only, the Archer and the King and Queen, Tho heaviest Archer coin weighs 132°5, and the average weight of 9 coins of this type is 129-21, The King and Queen type is known from two specimens only, and but one of these has been weighed ; its weight is 1288, These can, therefore, be no doubt, that, as has already been observed, the coinage of Skanda Gupta conforms to the same standard as the Wheel variety of the Archer type of Chandra Gupte II. The investigation has thus established the remarkable fact that the undisputed coins of the imperial Gupta Dynasty were struck according to at least three distinct standards of weight, of approximately 125, 130, and 134-5 grains respectively. When we turn to tho later coins included in the Supplement to my Cata- logue another and more striking change in the weight standard presenta iteelf. These coins are all, except the Prakdgdditya coins, of the Archer type, with reverse device of a goddess (Lakshmi probably) seated on a Iotus-flower. The execution is rude, and the metal sometimes debased. Of the coins bearing the name of Chandra, the weights of three are known, the average being 145°66, and the highest 148. The correspond- ing figures for 4 coins inscribed with the name Kuméra, or its first syllable, are 146-3 and 148-7. The only gold coin of Skanda Gupta Kraméditya which has been tested, weighs 141-4. The coins of Néra Gupta Baldditya average 145°66, with a maximum of 148'7, and the Lion and Horseman coins of Prakdééditya show an average of 145°6 and a maximum of 146:2. These figures demonstrate that all these coins were struck according to one standard, and that quite different from any of the standards adopted for the undisputed mintages of the imperial Gupta sovereigns. What was this standard? It seems to me that it was the ancient Hindu weight and coin, the suvarga, or golden Kérsha of 80 ratis. General Cunningham finds it “for all practical purposes extremely convenient and sufficiently accurate to assume the value of the rati at 1-75 English grain, which is the value that has already been adopted by Mr. Thomas on the evidence of the coins themselves.” If this value for the rati be accepted the weight of the suvarra must be fixed at 140 grains, and the coins now under consideration, whatever they may be, cannot be intended for suvarnas. General Cunningham observes that “‘no one to my knowledge has sen a surarga,” and in the sense that no one has yet discovered an an- cient Hindu pre-Alexandrine coin of that denomination, the observation is accurate; but I venture to submit that the coing of Néra Gupta and146 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No.2, his compeers must be considered as revivals of the ancient suearna, and that this conclusion is fully warranted by General Cunningham's own researches. He has devoted much time and labour to the task of ss certaining the value of the raft, by weighing the raft seeds (Abrus pre- catorius) and the other kinds of seeds metrically associated with the ratéin the Hindu books. ‘The mean of four values of the ratt deduced from actual weighments of the seed of the Abrus is 18143, General Cun- ningham himself, with the most elaborate precaution, weighed “one thousand sound and tolerably even-sized seeds”, with the result that the average weight was 1-823 grain, and Mr. Laidlay’s weighments on his behalf gave practically the same result, 1-825. Weighments of rice and other seods alleged in the Hindu books to have definite numerical ratios to the weight of the Abrus seed gave re sults varying from 1-791 to 1-825, with a mean of 18044. By taking the mean of the two average weights above noted (18143 + 1:8044 +2) says General Cunningham, “we obtain 1-8093 as the true value of the actual rati.”* This expression is not scientifically accurate, because & mere arithmetical average of results obtained from experiments conduct- ed in different ways, and with various degrees of precaution, is not on- titled to be called a true value. It seems to me that if witnesses are to be weighed and not counted the nearest possible approximation to the ‘trae value’ is to be found in the result 1-823 obtained by General Cunningham from the truly scientifio experiment made by himself which he describes, confirmed as it is by the almost identical result, 1'825, obtained by Mr. Laidlay. General Cunningham, therefore, on his own showing, is not justified in assuming 1-75 grain as the value of the rat; and in 1865 he accepted the value 1-823 grain for the rati.t Mr. Thomas arrives at the seductive figure 1-75 by o different method. He shows, for instance, that the Hindu silver coin known as purina should contain 32 ratis, and that purdna pieces actually in existence weigh as high as 55 grains, and then, so far as I understand him, jumps to the conclusion that the full weight of the purda was 56 grains. But I cannot see anything in his arguments inconsistent with * For Gen. Cunningham's experiments and opinions see his paper ‘On the Mone- tary System of the Greeks in Bactriéna, Ariéna, and India,’ in Num. Chron. Vol. XIII, N. 8, (1873) pp. 187-219, especially pp. 198-7. Mr. Thomas has explained his views in his essays on Ancient Indian Weights (Num. Chron IV, N. 8. (1864) pp. 40-58 and 114-132, especially p. 132.) 'Theso essays have been republished with additions in the International Numismata Orientalia, + “The old Indian pana or copper coin of 145'883 grains.” (Coins of the Nine Nagas etc., in J. A. 8. B. Vol. XXXIV, 1865, p. 120.) The para of coppor corresponded in weight with the suvarse of gold.1884.]_V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 147 the assumption that the full weight of the purdaa was 57, or 58, or 59 grains, and must confess to remaining unconvinced by his reasoning, which seems to make insufficient allowance for loss of weight by wear. I believe General Cunningham’s 1-823 grain to be the nearest possible spproach to the true value of the rat, but, for convenience, would adopt Mr. Laidlay’s value 1-825, which only differs from the other by s}ath ofsgrain, Tho scale of Hindu gold coins and weights, will then stand 18 follows j— Brats =1mésha = 9125 grains. 80 ratte = 16 mdshas = 1 suvarpa = 146-000, The silver puréna will thus be equivalent to 58-4 grains, a result ap- parently quite consistent with the weights of existing specimens when allowance is made for wear. These results are, I submit, much nearer to the trath than the figures 8°75 and 140 and 56 respectively, as adopted by General Cunningham in his later publications and by Mr. Thomas, and they happen to be very nearly as convenient for purposes of calen- lation. I would urge, however, that mere convenience of calculation does not justify any appreciable modification of the results arrived at by scientific investigation, and that our business is to get at the trath so far as possible, and to make our arithmetic conform. Tried by this test our coins obviously appear to be intended for suvargas. To make the point clear I repeat tho weights :-— Suvarna = 80 ratte @ 1825 gre. = 146-00 grains. Av. wt. of Chandra barbarous coins 14566, » »Koméra y » 14630, » » Skands ” ” 14140, » » Néra ” ” 14566, » » Prokighditya,, =, = 14560, It is true that some specimens weigh as much as 148-7, and that base metal coin of the Kuméra type weighs 150°3, but, considering the rade execution of these coins, and the inferiority of tho metal in many instances, I do not think that this excess of weight invalidates the reference of these coins to the suvarsa standard. Whether I am right or wrong on this point, the discussion at least proves that an investiga tion in detail of the weights of the coins of the Gupta period ia not with- out interest, and may lead to conclusions of some importance. It is to be regretted that the materials for the discussion are at present comparatively scanty, and I hope that collectors of Indian coins may be induced to pay more attention to the weights of their coins than has hitherto been customary.148 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. (No.3, Section V. Finp-spors. The information concerning the localities in which the Gupta gold coins have been exhumed or otherwise obtained is not 80 copious as could be desired, and most coin collectors seem to take little interest in as- cortaining either the spot where their specimens were found, or the details of the contents of each trove. Yet these points eminently deserve attention. Greek coins usually indicate on their face the locality of the mint where they were struck, but the Gupta and other Indian coins ordinarily have no indication of the sort, and, in the absence of trust worthy written history, the records of the find-spots of coins are almost our only clue to the position of the ancient Hindu mints. The Guptas, and other dynasties of pre-Muhammadan India, which modern archsological research has reseued from the utter oblivion of centuries, are still for the most part the merest shadows, endowed with names certainly, but without any definite local habitation, and often as unfixed in time as in place. The fabric, weight, style, devices, and legends of coins help us to fix the chronological position of these dynasties, whose names dance before the eyes of the student in a most perplexing maze. The recorded find- spots of coins, and detailed account of the contents of individual troves should be studied with care equal to that bestowed on the more attractive parts of numismatic science in order to throw light on the position of the old mint-cities, and on the local limits of the dominion of these long- forgotten sovereigns. Full details of the contents of hoards of coins when skilfully used, can be forced to yield to the historian many valu- able hints. These few observations will, I trust, be deemed sufficient justification for the elaboration with which I have worked out this part of my subject, so far as the meagre materials available would permit. I hope that collectors will be good enough to impart to the Society additional facts to complete the imperfect information at my command, and to correct any erroneous inferences which may be based upon insufficient premises. Professor Wilson, with his usual caution, declined to commit him- self to any definite opinion as to tho seat of the dominion of the Gupta kings, or the position of their mint-cities, and contented himself with the remark that “all that can be affirmed of them (scil. Gupta gold coins) with any degree of certainty is that they are coins of the west and north- western provinces of Hindustén.” Ho also pointed out that these coins are not found in the Panjéb or Afghénistan.* * Ariana Ant. p. 417,1684.] V. A. Smith—@old Coine of the Imperial Gupta Dynasty. 149 Prinsep treated this topic with greater explicitness, but, as will be shown presently, with less accuracy. “Kanauj,” he says, “has been fixed on as the locale of the present class of gold coins, for the obvious reason that they are most frequently found in its rains, not that any his- tory ascribes them to this town.”® In another passage he appeals again to the “ frequency of his coins discovered at Kanauj” as a reason for fixing Samadra Gupta’s capital at that place.t In a subsequent essay Prinsep to some extent correcta his former attribution of the majority of the coins to Kananj, and observes, “Since my former paper on the Gupta coins of Kananj appeared, very important acquisitions have been made to our knowledge of this before unknown dynasty, through the medium of coins and inscriptions; for both of which we are almost entirely beholden to the researches of Lieut. Cunningham and Mr. Tregear in the neighbourhood of Benares.” After discussing the passage in the Vishnu Puréna, which defines the territory of the Guptas of Magadhe as extending “along the Ganges to Preydga” (Allshabid), he romarks that “the sites, whence these coins have been most frequently obtained, certainly agree with this de- scription.” A few pages later Prinsop states that the Gupta gold coins are “ dis- covered in greatest quantity at Kanauj, Jaunpur, Gayé, and even occa- sionally in Bengal.”§ Abstaining for the moment from any comment on the statements above quoted, I shall proceed to state all tho facts which I have been able to ascertain respecting the find-spota of the Gupta gold coins; first enumerating the hoards known to me, and then giving statistics of indi- ¥idual coins, including some which formed parts of certain of the hoards mentioned. 172 so-called “gold darics” were found near Benares in the time of Warren Hastings, who sent them home to the Court of Directors, himself “as making the most munificent present to his masters that he might ever have it in his power to send them....The story is that they were sent to the melting pot. At all events they had disappeared when Hastings returned to England.”|| It is almost in- credible that these 172 pieces should have been Persian darics. Tho © Becays I, 284. + ibid, 239. t id pp. 865-6, Mr. Trogear collected at Jaunpur 40 miles from Benares; Lt. {now Genl.) Cunningham wos then at Bonares. § iid p. 375. {| Genl, Cunningham on the Oxus Treasure Trove in J. A. 8. B. for 1881, p. 1845 and ‘ India, What can it Teach us,’ by Max Miillor, p. 8,150 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No. 2, latter are oxtremely rare, only about 40, I believe, being known to exist, and the neighbourhood of Benares is a place extremely unlikely in which to find a large hoard of them. I consider it highly probable that the trove consisted of Gupta gold coins of the prevailing Archer type, which might in those days be easily confounded with the Persian régorat. Another great golden treasure was found during the reign of Warren ‘Hastings in the year 1783, at Kéli Ghit, ten miles above Calcutta, on the east bank of the Hiigli. The hoard comprised over 200 coins, many of which were sent home by the Governor-General and were distri- buted among the cabinets of the British Museum, East India Company, and other public institutions, where some, at any rate, of the speci- mens are still to be seen. The coins of this hoard are described by Wilson as being “ of rude execution and debased metal,” and it is doubt- fal if any of them are authentic issnes of the imperial Gupta dynasty, though “agreeing in general design with the Archer type of those issues. A fow specimens from this hoard, which I designate by the name of Kalighét, are noticed in the Supplement to my Catalogue.* In 1838 Mr. Tregoar dug up some specimens of the Gupta gold coin- age in somo ruins, known as Jaichandra’s Mahal, near Jaunpur. The exact number of the coins so found is not stated, but it does not appear to have been large. Most, if not all, of these coins were subsoquently published by Prinsep, and are included in my Catalogue.t The important trove, which is referred to in the Catalogue as tho Bharear hoard, was found near Benares in 1851 and is described by Major Kittoo as follows :— “These coins, which are all gold, of different weight and quality, wore of a trove of ninety in number, that is, such number were delivered into the treasury. They were found, with about 70 more, by some vil- lagers, buried in a copper vessel, in a mound on which stands the village of Bharsar, in pargana Bharwal, and Théna Chandaulf, about twelve miles from Benares, betweon the Ganges and Karamnésa. Bharsar is the site of one of tho many ancient cities, the names of which are lost....... “Of the number [scil recovered] 71 were coins of Chandra Gupta, 69 being of one type of his coinage [scil. evidently, Archer type]. Of these, four were retained of the most perfect, and the remainder wore sold by auction ; they were all more or less defective, and but few of them had even a portion of the legend round the rim perfect, but the * Marsdon Num. Or., II, 726; Ariana Ant. pp. 416-17 and Plate XVIII, figs. 21 c0qq. ‘The barbarous coins figured by Marsden were from this hoard. (Prinsep's Essays, I, 230.) + J. 4.8. B. IIT, (1881), 619.1984.) _V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 151 tame [in Gupta characters in text] beneath the left arm of the figure vas distinct in all of them.”* Thirty-two coins were retained and described. This hoard being the oaly large one concerning which details at all copious have been recorded, I think it is desirable to give an analysis of the portion preserved. Tho thirty-two coins retained comprised the following types and varieties :— Samadra Gupta—Javelin type, var. (1) ... aL ” » » oo 4) ow 1 5 » Archer ,, ,, 8, 2 and a duplicate. 3 ” » Lyrist ,, —————- §- ——____. 1 Chandra Gupta II—Archer ,, class IJ, o; 2 and a dupl. 3 ” » ow nm ow BS Hepes 2 » woo oo eo VEL 2 ” no» won» 81 1 ” Horseman to Left—; 2 Kumara Gupta Mahendra — Archer type, var. a; 2 i. . — Horseman to Right ,, var.a; land ap 2 oe » — » — » Left » ——— wn 2 ” » ” — Peacock » var. B; — 1 7 fi . — Combatant Lion ,, 1 Skanda Gupta, — Archer » 8.and 3 dupl. 6 Prakiséditya, — Lion and Horseman ,, 2 Total... 32 The contents of the hoard scem to indicate that it was buried not very long after the close of the reign of Skanda Gupta, that is to say, (sccording to the chronology which I adopt), not later than about 400 A.D.; and we thus learn that at that time the mound of Bharsar was an inhabited town. The Lyrist coin of Samudra is noted as being in fine condition, and some of the Skanda pieces were likewise well preserved ; coin of Kuméra’s is described as much worn, a circumstance which renders it probable that the hoard was deposited at some considerable in- terval from the time of Kuméra’s reign. The association in a single bard of coins belonging to so many reigns, types, and varieties is re- markable, and shows that these various issues were all current together in the province of Benares. © Memo. by Major M. Kittoe, Archmological Enquirer, on some Ancient Gold Coins found near Benares in 1851, and submitted by the Government of India for the inspection of the members of the Asiatic Socioty ; with the Memo. on the same ty Mr. B.C. Bayley. (J. A. 8. B. XXI, pp. 399-400, and Pl. XII, figs 1-9. The plate ‘as miserably executed by a native). The coins from this hoard are not included in Mr. Thomas’ catalogues. v152 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No. 2, The same volame of the Journal of the Asiatic Society of Bengal which contains the description of the Bharsar treasure trove supplics a notice of some coins found at Muhammadpnr, near the Arankhéli River, in the Jessore District of Lower Bengal. They are described as being “all of the Gupta kings of Kananj (sic.] and comprise specimens of tho silver coinage of Chandra Gupta, Kumara Gupta, and Skanda Gupta The metal of these coins is very impure.” ‘The hoard included one gold coin weighing 85 grains, which, at that time, Baba Rajendralél Mitra believed to be a coin of Sri Gupta, but it is evidently of a date much subsequent to Skanda Gupta.* Mr. Thomas alludes, with tantalizing brevity, to “a batch of twenty gold coins found at Gopilpur on ‘the Ghigra River, on the site of the old village fort, ten miles west of Barhal’ in July 1854.” Barhal is in the district of Gorakhpur. Seven of these coins “from the mints of Chandra Gupta IL” were submitted to Government, and included a specimen of his Archer type, Class I, 8. No particulars are recorded concerning the other constituents of the hoard.+ General Cunningham informs me that a great hoard was found at Allahabad some twenty years ago consisting of about 200 of the gold Peacock coins of Kumara Gupta. General Cunningham saw a large number of specimens and ob- tained possession of four, two of which, namely, one of each variety, are still in his cabinet. About seven years ago twenty or thirty Gupta gold coins were found at Jhisi opposite Allahabad, comprising two specimens of Kuméra Gupta’s ‘Archer type, Class I var. a; and eight specimens of the same king's Peacock type in both varieties. Recently, thirteen gold coins of the Guptas were found near Higli, the hoard being composed as follows :— Samudra Gupta — Javelin type . 1 Chandra Gupta II —Archer ,, Class II 5 Kumira Gupta Mahendra — Archer ,, ————— . 3 ” ” » — Horseman to Right ,, 2 » ” a ”» » Left 7 . : » — Lion-Trampler ,, . 1 Total... 13§ # Note on Threo Ancient Coins found at Muhammadpur in the Jessore District, by Babu Réjendralal Mitra (J. A. 8. B. XXI p. 401; Pl. XII, figs. 10-12). One of these coins belongs to Saséngka (A. D. 600). Cunningham, Arch. Rep. III, p. 138.) + 5.4.8. B. XXIV (1855), p. 499. t From information kindly communicated by Sir F. C. Bayley. § I am indebted to Dr. Hoornle for my knowledgo of this hoard, which I designate by the name of Higli.1884.) V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 138 Of the eight Gupta gold coins in the possession of Sir E. C. Bayley, three, as above mentioned, were found at or near Allahabad ; the rest are believed to have been found either at Kananuj or Allahabad. The gold coins of the Guptas in the cabinet of Mr. A. Grant were all obtained in Oudh, and mostly near Faizébad (Ayodhya), except one, which was bought in Bombay. Mr. J. Hooper’s specimens were also obtained in Eastern Oudh. Mr. H. Rivett-Carnac has 13 gold Gupta coins, 6 of these were obtained in the province of Benares, 2 were dug up near Allahabad, 1 was procured at Lucknow, 2 at Cawnpore, and 2 at Mathura. Col. Tod's coins, including the four gold Guptas figured in Trans. R.A.S. Vol. I. Pl. XII, 4th series, were all obtained at Agra, Mathura, Ujjain, or Ajmir, but more precise information respecting them is wanting. The above notes comprise all the facts which I havo been able to collect respecting hoards of Gupta gold coins, and the origin of the collections in various cabinets.* I shall now proceed to analyse the available statistics respecting the find-spots of individual coins, including some which were comprised in certain of the hoards already mentioned. Prinsep was more careful than many other antiquarinns have been to tecord the source from which he obtained his coins. Thirty-seven Gupta, gold coins are described in his essays, and the following table of the find- spots of these coins has been compiled from his notes :— From Kananj ... . “ . 3 » Jaunpur 3 : 4 P (cabinet of Tregear, who collected at Jaunpur) 11 » Benares ... i i i a ” oe 1 » Gayé 4 » — Mirzépur 1 » Ghézipur 1 » not stated 12 Total... 37 In the following general statement, compiled from all the notes of pice in my catalogue, the coins described by Prinsep are included. * There is nothing to show the provenance of any of the coins in the India Office collection. ‘The find-spots of a very few coins in the British Museum collection are recorded, but there is reason to doubt the accuracy of some of the notes. Mr. ‘Theobald does not know where any of his Gupta coins were found, except that one ‘ras bought in Benares, and one in Matharé. Information is likewise wanting con cerning the provenance of nearly all tho coins in the cabinet of the Asiatic Society of Bengal.154 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. (No. 2, The coins of the Bharsar hoard are placed under the head of Benares and those found at Jhisi under the head of Allahabad. Tod’s coins are described as obtained in N. W. India. For facility of verification the figures are given for each reign. It is necessary to observe that the notes which form the basis of my tables are often vaguely expressed, and that in some instances it is impossible to say whether the coin was exhumed or only bought at the locality named. Tape or Fixp-srors. Total Reign. Find-spots. No. of coins. for reign. Ghatot Kacha. Jeunpur ? 1 Chandra Gupta I. a ? : Ghézipur » Benares Samudra Gupte. Kanauj ” P ” Sabéranpur ” Mathura » Oudh » Jaunpur ” Benares » Gayé » Patna ? ” Higli Chandra Gupta II. Kananj Bulandshahr Cawnpore N. W. India Oudh Ghézipur Jounpur 8 Mirzapur Benares 1ge Gorakhpur District ™ Hugli 5 M4 © The Barhal hoard comprised about 20 Gupte coins, of which 7 belonged to the reign of Chandra Gupta II; it is not known to what reign the remaining coins belonged, and I have therefore excluded them from the tablo. In the case of the Bharsar hoard I have only taken credit for the few coins described in detail, bat the hoard comprised about 160 Gupta coins, of which 71 belonged to the reign of Chandra Gupta I. a ae See eee eee Hone Ree eee1864] V.A. Smith—Gold Coins of the Imperial Gupta Dynasty. 155 Totat Reign. Find-spot. No. of coins. for reign. Kuméra Gupta Mahendra.N. W. India Allahabad (Jhisi) L Oudh Jeunpur Benares Gays Mahanada Midnapur Hagli Skanda Gupta. Kanauj Oudh Jounpur ? Benares Ghézipur Mahanada Total for all reigns ... 39 Pee DOR ORE 6 ae 118 The following figures give the meagre information available con- ceming the find-spots of the coins mentioned in the Supplement to the Catalogue -— Chandra Oudh 1 1 Kuméra Kalighat* 1 1 Skanda Gay$ 1 1 Nira Oudh 1 ” Kélighét* 1 2 Prakiséditya Kananj 1 Benares 2 3 Total 8 Grand totel 118 + 8 fe we 126 The next following statement exhibits a result of the investigation which may surprise some of my readers. L Coins obtained at Kananj (including 1 doubtful ease). Samndra Gupta. eee Chandra Gupta II — 2 © The Kalighét hoard comprised over 200 coins, but its detailed composition is not known.136 V. A. Smith—Gold Coing of the Imperial Gupta Dynasty. (No. 2, SeodsGute 7 CO Prakiséditya on . 1 Total 6 TL: Coins obtained west and north-west of: Kanes, Samudra Gupta. 1 2 Chandra GuptaTl we Kuméra Gupta Mabondra =... 2 Total’ 10 IIL. Coins obtained east of Kananj. GistKeche Chandra Gupta I oe Samadra Gupta : 20 Chandra Gupta IT... ... 36 Kuméra Gupta Mahendra 38 Skanda Gupta... 5 Doub FG ott 100 Grand total 126 The above total refers only to the coins described in detail, but, when the entire contents of the Barhal, Allahabad and Bharsar hoards are included, the total of indisputably Gupta coins found east of Kananj will be about 430, If the entire Kélighit hoard of rude coins of Gupta type be included, the total must be raised to about 690, and, if it be admitted that Warren Hastings’ 172 “gold darics” from Benares were Gupta coins, the grand total would amount to about 860. In any case the figure for Kanauj stands at 6. The fact is therefore established with mathematical certainty that Kanauj supplies only an infinitesimal pro- portion of the Gupta gold coins, the great bulk of which have been obtained far to the east of that city. Section VI. The Gupta Mint-Cities and Capital. In Southern India the ancient native governments permitted subor- dinate rulers and even private persons to coin in all metals, including gold, with little restriction, but there is no indication of similar laxity having at any time prevailed in Hindustan.* In Northern India successive dynasties followed, so far as is known, the practice of the Persian kings and Roman emperors, and jealously retained in their own hands the right of coining gold. If this bo admitted, it follows that the ancient gold coinages of Northern India will have been struck at or near the seats of government of the sovereigns who issued them. If, therefore, we can discover the position of the mint-cities where the gold pieces of the Gupta kings were ‘Thomas, Anciont Indian Woighta (Intern. Num. Or.), p. 57.1884.) V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. 157 struck, we may feel confident that we have also discovered the site of the capital, or capitals, of those princes. Prinsep designated the Gupta gold coinage by the name of the ‘Kanauj series.’ Almost without exception later writers on Indian archeology have followed him, and it seems to me, have rather blindly followed him, in assuming the existence of a special connection between the Gupta dynasty and Kanauj. Not to mention other and less authoritative writers, Mr. Burgess speaks of “ the Guptas of Kanauj,”* and Sir K. C. Bayley does not hesitate to afirm explicitly that Kanauj was the Gupta capitalt. But what evi- dence warrants us in asserting that the Gupta kings had their capital at Kanauj? I can find none. It will not be disputed that the belief in Kananj being the Gupta capital originated in Prinsep’s designation of the gold coins as the ‘ Kananj series,’ and in his assertion that they were most commonly found at Kanauj. But the statistics given in the last preceding section of this Paper, which are indisputable so far as they go, prove that Prinsep was mistaken as to the fact, and that the coins in question are not most com- monly found at Kanauj. Attention has already been drawn to tho circumstance that Prinsep subsequently corrected his earlier and less guarded assertion, and bracketed Jaunpur and Gayé with Kanauj, as the places where the Gupta gold coins were found in greatest abundance ; and, from the first, he was careful to note that no history connected the Guptas and Kanouj. Prinsep’s error, therefore, so far as it was an error, was not a grave one, and his statements offer a very slender foundation for the categorical assertion that Kanauj was the Gupta capital. It seems to me that Prinsep’s misapprehension on the subject can be very easily explained. ‘The fine Retreating Lion coin of Chandra Gupta UI was obtained by Lieut. Conolly at Kananj; and the fact that the publication and study of this coin led to the decipherment of the rest of the series appears to have impressed Prinsep’s imagination, and to have infuenced him in giving the name of ‘Kanauj Series’ to this class of gold coins, However this may be, the solid fact remains that out of 37 coins described by Prinsep, the find-spots of 25 are known more or less accu- rately, and of these latter only 3 can be traced to Kanauj. Nor have I been able to find a record of a single hoard of Gupta gold coins found at that city, and it need hardly be observed that the occurrence of hoards in certain places is more valuable as evidence for the purposes of the his~ * Archol. Survey of W. India, II, p. 80. + Num. Chron. Il, 3rd 8. (1882) p. 158.158 V. A. Smith—Gold Coins of the Imperial Gupta Dynasty. [No. 2 torian than the finding of isolated coins, which may have reached their resting places in any of a hundred different ways. I think, therefore, that the evidence now presented fully warrants the assertion that the find-spots of the Gupta gold coins in no way support the statement that Kanauj was the Gupta capital. I am not aware that evidence of any other kind has ever been ad- duced in support of that statement, which has been passed from one writer to another apparently without examination. I do not deny that Kananj was in existence during the rale of the Gupta kings, nor that it was included in their dominions. Little appesrs to be known about its early history, but it has always been reputed one of the most ancient of Indian cities, and we know that it was an import- ant place in 400 A. D. when Fa Hian visited it, and it appears to have been known by name to the geographer Ptolemy about A. D. 140. It is also certain that it was the capital of the eastern dominions of the great Harsha Varddhana in A. D. 634, but all these facts in no wise prove it to have been the Gupte capital.* I am quite willing to admit that Sir K. C. Bayley is right in calling Kanauj ‘the Dehli of the Hindus,’ if that title be restricted to the centuries between 600 A. D. and the Muhammadan conquest, but I can find no authority for the antedating of this claim to precedence.t The conclusion arrived at so far is a purely negative one. I shall now consider whether any positive result as to the position of the mints and capital may be obtained from » study of the find-spots of the Gupta gold coins and other evidence. It may safely be affirmed that the records of the localities, both where hoards and where individual coins were found, indicate unmistakably that the Gupta gold coinage was struck and chiefly current in territories far to the oast of Kanauj, and that these territories may be roughly de- scribed as the Province of Benares, with some adjoining districts. It seems to me impossible to draw any other conclusion from the evidence which has been set forth in the section on find-spots.* * These references are quoted from Genl. Cunningham's Archmol. Rep. I, 280 Sir B.C. Bayley informs me that in the Basle edition of Ptolemy (1583) the namo which is supposed to mean Kanauj is written Kavayépa, + The phraso ‘ the Delhi of the Hindus ’ is quoted from a letter on this subject with which Sir E. C. Bayley favoured me. His theory about the supposed dates in the Gupta era on the medimval coinage of Kabul (Num. Chron. 8rd Ser. Vol. II, pp- 128-165 and 291-204) is of very doubtful correctness, and even if it were proved, does not contribute to the solution of the question discussed in the text. I see no reason for supposing that theuse of the Gupta era was connected with the sovorcignty of Kanauj. t The scanty evidence as to thé provenance of the Gupta copper coins (ante