Pramāṇa · Search · About · API

The Early History of the Deccan (Parts I–XI) v1.c7.s4: Early Sātavāhanas—Chronology

The Early History of the Deccan (Parts I–XI) v1.c7.s4

Language
English

Contents

Early Sātavāhanas—Chronology

v1.c7.s4.1

THE Andhra-Sātavāhana kings may be divided into two groups, the one earlier and the other later. The break in their rule in Western Mahārāṣṭra, Gujerāt, and Malwa explains this division. Until very recently, not much was known of the history of kings 4 to 22 of the Matsya list. A closer study of the potin and copper coins of Western India bearing the legend Rano Sātakamnisa seems to reveal a second Sātakarṇi, the 6th of the Matsya, the 3rd of the Vāyu, and the 5th of the Viṣṇu lists. A large copper coin from the Central Provinces gives us the name of Apilaka, the 8th king on the Matsya list; and the Bṛhatkathā, Saptaśati, and Līlāvatī offer some account of Hāla, the 17th king on that list. Even of these rulers the available information is too meagre to give an idea of the extent of their dominions. After the rule of a few early Sātavāhanas their sovereignty over the western regions of their dominions gave way to that of foreign invaders. These latter held control over those lands for a considerable time, until finally Gautamīputra Sātakarṇi appeared on the scene, recovered these regions from the aliens, destroyed the Śakas, Yavanas, and Pahlavas, and put an end to the dynasty of the Khakharāṭas. Thus the rulers of this dynasty, from Simuka to Gautamīputra Sātakarṇi, are treated as Early Sātavāhanas and the rest as the Later Sātavāhanas.

v1.c7.s4.2

But like every other problem of Andhra-Sātavāhana history the question of their chronology is very difficult to solve. Except for the testimony of the Purāṇas and a couple of difficult but possible synchronisms, we are indebted entirely to palaeography for our scanty knowledge of the regnal periods of these princes. And palaeography seems to authorize such widely divergent views that we are never absolutely sure of anything at all. The Nānāghāṭ inscriptions of the Sātavāhanas, for instance, are assigned, for palaeographical reasons, to anywhere between the middle of the second century B.C. and the latter half of the first century B.C.,¹ to any time ‘between the Besnagar pillar inscription of Heliodorus (about the beginning of the first century B.C.) and the Hāthigumphā inscription of Khāravela (end of first century B.C.)’.²

v1.c7.s4.3

¹ R. P. Chanda, MASI, No. 1, H. C. Raychoudhari, PHAI, 4th edition, pp. 337 ff.

v1.c7.s4.4

² R. P. Chanda, op. cit., and D. C. Sircar, Select Inscriptions, p. 186, n. 1 and p. 206, n. 1. The dating of the Hāthigumphā epigraph itself varies from the beginning of the second century B.C. to the end of the first century B.C.

v1.c7.s4.5

The Purāṇas themselves hardly appear to improve matters, for they are replete with contradictory statements. The four Purāṇas—Matsya, Vāyu, Viṣṇu, and Bhāgavata—contain statements regarding several of the ancient Indian dynasties, the total duration of rule by each dynasty, the number of princes making up each of the ruling families, and in most cases their names as well. Allowing for a few discrepancies, they are all agreed that the Mauryas ruled for 137 years, the Śuṅgas for 112, and the Kaṇvas for 45 years. Of the Andhras, the *Vāyu*, *Viṣṇu*, and *Bhāgavata* declare that there were 30 princes, while according to *Matsya* there were only 29 of them. Their total regnal period was 460 years according to *Matsya*; 456 years according to *Viṣṇu* and *Bhāgavata*; and 411 or 300 years according to *Vāyu Purāṇa*. Though not much discrepancy is discernible here among their general statements, this is far from the case when they begin to give more exact particulars of the individual reigns. The following table illustrates these disagreements:

Andhra Kings

v1.c7.s4.t1.1

| | *Vāyu* | *Matsya* | *Viṣṇu* | *Bhāgavata* |
| | No. of princes | Total reign | No. of princes | Total reign | No. of princes | Total reign | No. of princes | Total reign |
| General statements | 30 | 411 or 300 | 29 | 460 | 30 | 456 | 30 | 456 |
| Details | 17 | 272½ | 30 | 448½ | 24 | .. | 22 | .. |

v1.c7.s4.t1.2

It will be noted in the above table that the number of names actually enumerated by each authority differs from the number given in the general statement, and that consequently the total of their regnal years also differs from the total duration of each dynasty mentioned in a general way.

v1.c7.s4.t1.3

Sir R. G. Bhandarkar sought to reconcile the conflicting statements of the *Vāyu* and *Matsya Purāṇas* about the duration of the dynasty and the number of princes, by supposing that the longer period and the corresponding list of princes are made up by putting together the reigns and names of the several branches of the Andhra dynasty. He accepted the particulars given in the *Vāyu Purāṇa* in preference to those given in the *Matsya Purāṇa*. Commenting on a statement of *Vāyu*, ‘Andhras will have possession of the earth for three hundred years’, Sir R. G. Bhandarkar writes: ‘the period of three hundred years and the seventeen names given in the *Vāyu Purāṇa* refer probably to the main branch. The *Matsya* seems to me to put together the princes of all the branches and thus makes them out to be thirty. . . . Thus then both the *Vāyu* and the *Matsya Purāṇas* each give a correct account, but of different things.’¹

v1.c7.s4.t1.4

¹ EHD, pp. 57-8.

v1.c7.s4.t1.5

This argument is hardly tenable. The compiler of the *Matsya Purāṇa* might well be spared this somewhat arbitrary justification of his alleged inability to distinguish the crowned from the uncrowned princes of the Sātavāhana line. A glance at the table will show that both the *Vāyu* and the *Matsya* list are in agreement about the names of the kings from Paṭimāvi the 5th of the former and Puḷumāvi the 15th of the latter to Gautamīputra the 13th and the

The Matsya and Vāyu Lists of Andhra Kings

v1.c7.s4.t2.1

(See EHD, pp. 55-6; CIC, pp. lxvi-lxvii)

v1.c7.s4.t2.2

| | Matsya | | | Vāyu | |
| *No.* | *Names of kings* | *Years of reign* | *No.* | *Names of kings* | *Years of reign* |
| 1 | Śiśuka (= Simuka of Inscr.) | 23 | 1 | Sindhuka | 23 |
| 2 | Kṛṣṇa (Kaṇha of Inscr.) | 18 | 2 | Kṛṣṇa | 10 or 18 |
| 3 | Śrī-Mallakarṇi | 10 | | | |
| 4 | Pūrṇotsaṅga | 18 | | | |
| 5 | Skandastambhi | 18 | | | |
| 6 | Śātakarṇi¹ | 56 | 3 | Śātakarṇi¹ | 56 |
| 7 | Lambodara | 18 | 4 | Apilavā | 12 |
| 8 | Apilaka | 12 | | | |
| 9 | Meghasvāti | 18 | | | |
| 10 | Svāti | 18 | | | |
| 11 | Skandasvāti | 7 | | | |
| 12 | Mṛgendra Svātikarṇa | 3 | | | |
| 13 | Kuntala Svātikarṇa | 8 | | | |
| 14 | Svātikarṇa | 1 | | | |
| 15 | Puḷomāvi | 36 | 5 | Paṭimāvi or Paṭumāvi | 24 |
| 16 | Gaurakṛṣṇa | 25 | 6 | Nemikṛṣṇa | 25 |
| 17 | Hāla | 5 | 7 | Hāla | 1 |
| 18 | Maṇḍulaka | 5 | 8 | Maṇḍulaka | 5 |
| 19 | Puriṇdrasena | 5 | 9 | Purikasena | 21 |
| 20 | Sundara Svātikarṇa | 1 | 10 | Śātakarṇi | 1 |
| 21 | Cakora Svātikarṇa | ½ | 11 | Cakora Śātakarṇi | ½ |
| 22 | Śivasvāti | 28 | 12 | Śivasvāti | 28 |
| 23 | Gautamīputra | 21 | 13 | Gautamīputra | 21 |
| 24 | Puḷomat (Puḷumāvi of Inscr.) | 28 | | | |
| 25 | Śiva Śrī | 7 | | | |
| 26 | Śivaskanda | 7 | | | |
| 27 | Yajñaśrī | 29 | 14 | Yajña Śrī | 29 |
| 28 | Vijaya | 6 | 15 | Vijaya | 6 |
| 29 | Caṇḍaśrī | 10 | 16 | Daṇḍaśrī | 3 |
| 30 | Puḷomāvi (Puḷumāvi of Inscr.) | 7 | 17 | Puḷomāvi | 7 |

v1.c7.s4.t2.3

¹ Other variants of this name are Svātikarṇa, Sāntakarṇa, Śātakarṇi, &c. In the following pages the last form is adopted.

v1.c7.s4.t2.4

23rd respectively of those Purāṇas. With slight differences, not of much significance, both of them give these nine kings the total regnal period of 126½ years and 126½ years respectively. Again, the names of the kings beginning with Yajñaśrī Śātakarṇi and ending with the last of the line are fairly similar. But Vāyu is silent as regards three princes whose reigns intervened between those of Gautamīputra and Yajñaśrī. Bhandarkar would ask us to believe then that these three—Puḷomat, Śiva Śrī, and Śivaskanda—were not crowned kings at all, but that they died as princes without succeeding to the ancestral throne. But Bhandarkar himself admitted that ‘Puḷumāyi’ (Puḷomat of Matsya) reigned in his own right for at least four years.²

v1.c7.s4.t2.5

² EHD, pp. 59-60. Puḷumāyi = Puḷumāvi = Puḷomāvi = Paṭimāvi. In the text the form ‘Puḷumavi’ is adopted.

v1.c7.s4.t2.6

It is clear, therefore, that the omission of thirteen names and a corresponding reduction in the total duration of the Andhra rule must be accounted for in a different way. Might it not be that the *Vāyu* was paying greater attention to the *more important phase* of the Sātavāhana rule which began with Paṭimāvi or Puḷumāvi (I)? Dr. V. A. Smith¹ pointed out that the difference between the *Viṣṇu's* 456 years and the *Vāyu's* 300 years is nearly made up by total duration assigned by the *Purāṇas* to the Śuṅgas (112) and the Kaṇvas (45), and that the shorter reckoning of *Vāyu* must be taken to date from the close of the Kaṇva dynasty. If *Vāyu* mentioned four other predecessors of Paṭimāvi this was in consonance with its own observation in the preface that it would mention only the leading names (*Pradhānyataḥ Pravakṣyāmi*).²

v1.c7.s4.t2.7

¹ ZDMG, 1902, 1905.

v1.c7.s4.t2.8

² JBORS, xvi, 264.

v1.c7.s4.t2.9

The *Matsya*, therefore, may be accepted in preference to the other *Purāṇas*, for it is fuller in the details it furnishes as regards the princes and their names. For the duration of their reigns, however, *Vāyu's* version may be accepted, where available and not contrary to epigraphic evidence.

v1.c7.s4.t2.10

All the *Purāṇas* agree in stating that the Andhra dynasty followed that of the Kāṇvas. But the statements of the *Vāyu* and *Matsya Purāṇas* regarding this are held by Sir R. G. Bhandarkar to warrant a conclusion that the Śuṅga and Kaṇva dynasties ‘reigned contemporaneously, and hence the 112 years that tradition assigns to the Śuṅgas include the forty-five assigned to the Kāṇvas’. If in agreement with Kern we take it that the coronation of Chandragupta Maurya took place in 322 B.C., then the simultaneously uprooting of the Śuṅga and Kaṇva dynasties and the rise of the Andhra dynasty will have to be assigned to 322—(137+112) = 73 B.C.³

v1.c7.s4.t2.11

³ Bhandarkar, EHD, pp. 53-4.

v1.c7.s4.t2.12

The passage in question runs as follows:

v1.c7.s4.t2.13

काण्वायनस्त (नं) तो भृत्यः सुशर्माणं प्रसद्य तम्।
शुङ्गानां चैव यच्छेषं चययित्वा बलं तदा॥
सिन्धुको अन्ध्रजातीयः प्राप्स्यतीमां वसुन्धराम्॥

v1.c7.s4.t2.14

Sir R. G. Bhandarkar translated it thus: ‘A servant of the race of the Andhras having destroyed Suśarman of the Kāṇva family by main force, and whatever shall have been left of the power of the Śuṅgas, will obtain possession of the earth.’

v1.c7.s4.t2.15

Since the Kaṇvas are spoken of as Śuṅgabhṛtyas, Sir R. G. Bhandarkar concluded that ‘when the princes of the Śuṅga family became weak, the Kāṇvas usurped the whole power and ruled like the Peshwas in modern times, not uprooting the dynasty of their masters but reducing them to the character of nominal sovereigns...’⁴ We may remark at once that the reference to the Kāṇvāyanas as the servants of the Śuṅgas does not imply their joint rule with the Śuṅgas. The analogy of the Peshwās is unfortunate, because, in spite of all their power, they acknowledged themselves vassals of the family of Śivājī. No evidence can be adduced to show that the Kāṇvāyanas ever admitted any such fiction of legal suzerainty in their relation to the Śuṅgas. The Purāṇas clearly state that before Vāsudeva Kaṇva assumed power, he slew the Śuṅga king Devabhūti, and this is fully corroborated by Bāṇa’s *Harṣacarita*.¹ We cannot, therefore, ignore the dynastic succession as given in the *Purāṇas*. The verse under discussion has to be understood as having a different significance. When the Kāṇvāyanas usurped the throne, they did not secure all the territory that the Śuṅgas had acquired, but only a large part of it. Whatever part of the original power of the Śuṅgas was obtained by the Kaṇvas was destroyed by Sindhuka² when he slew Suśarman, the last of the family of the Kāṇvāyanas. The Śuṅgas and the Kaṇvas were thus successive and not contemporary dynasties. The fall of the Kaṇvas and the accession to power of the Sātavāhanas must then be placed in 322-(137+112+45) = 28 B.C.

v1.c7.s4.t2.16

⁴ EHD, p. 53.

v1.c7.s4.t2.17

¹ Nirṇayasāgara edition, p. 199.

v1.c7.s4.t2.18

² The *Matsya Purāṇa* gives the name as Śiśuka. In the Nānāghāṭ inscription it is written as Simuka. According to Pargiter, Simuka was misread ‘Sisuka’ and then Sanskritized into ‘Śiśuka’.

v1.c7.s4.t2.19

If then we accept 28 B.C. as the starting-point of the Sātavāhana power and work out the individual reigns of the princes on the basis of the *Matsya Purāṇa*, we shall be taking their rule well into the fifth century A.D., which is absurd. Hence Smith, Jayaswal, Bakhle, and other scholars are inclined to accept the general statement of the *Purāṇas* where they give the total number of princes with 400 years and more for the complete inclusive duration of their reigns, and to ignore their formula that ‘Andhra Simuka would slay Suśarman’, the last of the Kāṇvāyanas. These scholars would ascribe to some descendant of Simuka this seizure of the Imperial power of the Kaṇvas. Mr. V. A. Smith says: ‘the independent Āndhra dynasty must have begun about 240 or 230 B.C., long before the suppression of the Kāṇvas about 28 B.C.; and the Āndhra king who slew Susarman cannot possibly have been Simuka.’³

v1.c7.s4.t2.20

³ EHI, 4th edition, pp. 216-17.

v1.c7.s4.t2.21

Jayaswal’s conclusions are somewhat similar to those of Smith.⁴ But he drew support for his views from the Hāthigumphā inscription, where it is said that in his second year Khāravela sent his armies west, heedless of the Sātakarṇi. This Sātakarṇi is generally identified with the third in the Purāṇic list and with the one mentioned in the Nānāghāṭ inscription. As the relievo epigraphs in the Nānāghāṭ cave mention only Simuka and Sātakarṇi, leaving out Kaṇha (Kṛṣṇa), it has been presumed that Sātakarṇi was the son of Simuka and not of Kṛṣṇa as stated by the *Purāṇas*.

v1.c7.s4.t2.22

⁴ JBORS, xiii, 221-46.

v1.c7.s4.t2.23

But though the Sātakarṇi of *Vāyu Purāṇa* stands third in the list, he cannot be given a like position in the order of succession. We cannot identify him with the similarly named king of the Nānāghāṭ records, the performer of innumerable Vedic sacrifices, for in that case the *Purāṇas* could not possibly have mistaken his name: for the *Matsya* calls its 3rd king Mallakarṇi, and the *Bhāgavata* styles him Śāntakarṇi. The *Viṣṇu* has a Sātakarṇi for its 5th king, in place of the 6th of *Matsya*. The *Vāyu*’s 3rd must be taken to correspond to this Sātakarṇi. For among the *Purāṇas*, the *Vāyu* and the *Matsya* provide the regnal periods of these kings. If the reigns tally exactly with regard to any king, then the strong presumption is that both the *Purāṇas* must be referring to the same individual. In this case Sātakarṇi, the 6th of *Matsya*, is given a reign of 56 years and so is the 3rd of *Vāyu*. If we remember that *Vāyu* names only 17 out of the 30 kings with whom it invests the dynasty, it is natural to assume that between Kṛṣṇa and this Sātakarṇi *Vāyu* must have omitted the mention of certain rulers whose contribution to the greatness of the dynasty was either non-existent or comparatively insignificant.

v1.c7.s4.t2.24

The *Purāṇas* are quite clear that Kṛṣṇa was the father of his successor Sātakarṇi. To this categorical statement is preferred the evidence of the six Nānāghāṭ relievo inscriptions. But this ‘evidence’ is no more than a mere conjecture, based on the common location of the two names. Of the six epigraphs, four bear distinct terms of relationship. Nāyanikā is the *Devi* of Rājā Śrī Sātakarṇi. The next Bha(ya) is a *Kumāra*; Hakusiri and Sātavahana are again *Kumāras*. Mahārathi Tranakayiro, coming in between them, is taken to be the father of queen Nāyanikā (Nāganikā). But no such terms appear with the name of Simuka. He is called merely Rāya Simuka-Sātavāhano Sirimāto. The inference is obvious. If these figure-labels represent the family of Sātakarṇi, Simuka finds a place there only as the founder of the line. Otherwise some term of relationship would have been affixed to his name. This idea is not alien to the evidence, especially when we recall the recent suggestion that the missing passage in line two of Nāganikā’s record was

v1.c7.s4.t2.25

Simuka-Sātavāhanasa *Vamsa-Vadhanasa*.¹

v1.c7.s4.t2.26

¹ Dr. Sircar, *Select Inscriptions*, p. 187, n. 2.

v1.c7.s4.t2.27

It is evident, therefore, that the Sātakarṇi of the Nānāghāṭ record was not the son of Simuka or of Kṛṣṇa. He certainly came later than Sātakarṇi, the son of Kṛṣṇa, and third in the Purāṇic list of the Andhra kings. He was the contemporary of Khāravela for the ‘alphabet of the Nānāghāṭ inscription agrees generally with that of the Hāthigumphā epigraph. . .’. But Chanda and other scholars believe that the Hāthigumphā record was later than that at Nānāghāṭ, for palaeographical reasons. Mr. Chanda identified Hāthigumphā Sātakarṇi with No. 6 of Pargiter’s list and tentatively assigned him to the period between 75 and 20 B.C.² This date, however, is not acceptable. Chanda’s argument runs directly against that of Dr. Bühler who had given the Nānāghāṭ record the date 200-150 B.C. on palaeographic grounds.³ Now palaeographic evidence as to date can be reliable if we allow a margin of about fifty years on either side. So we agree with Dr. Bühler regarding the date of the record and Sātakarṇi’s contemporaneity with Khāravela of Kaliṅga, and we agree with Chanda that this Sātakarṇi must have been the sixth of the *Matsya* list.

v1.c7.s4.t2.28

² *IA*, 1919, 214; *MASI*, no. 1.

v1.c7.s4.t2.29

³ *ASWI*, v, 75.

v1.c7.s4.t2.30

But so far as the rise of the Sātavāhana power is concerned, the place of this Sātakarṇi in the *Matsya* list does not make any considerable difference. Khāravela’s date, and therefore that of the Sātakarṇi king, may now be fixed. The Hāthigumphā record speaks of Yavanarāja Dimita who was defeated and forced to withdraw owing to the valour of Khāravela, in the 8th regnal year of the latter. This Dimita was Demetrios, son of Euthydemos, who flourished during the first half of the second century B.C. The *Gārgi Saṃhitā* hints at the retreat of the Greeks from India due to troubles at home. The allusion is evidently to the struggle between Eukratides and Demetrios in Bactria. According to Meyer, Demetrios retired in 175 B.C.¹ This was in the eighth year of the Kaliṅga king. His expedition to the west in defiance of the contemporary Sātakarṇi took place in his second year, i.e. 181 B.C. Whether this Sātakarṇi be the third or the sixth of the *Matsya* list, the conclusion is inevitable that the slayer of Suśarman in 28 B.C. was not Sindhuka or Simuka, but some one of his descendants.

v1.c7.s4.t2.31

¹ *JBORS*, xiii, p. 242.

v1.c7.s4.t2.32

We may here hazard a suggestion that the Andhra-Sātavāhana king who destroyed the Kaṇvāyanas was very likely Puḷumāvi, the fifteenth prince of the *Matsya*, Paṭimāvi, the *Vāyu*’s fifth, and Paṭumān, the ninth of the *Viṣṇu Purāṇa*. The close correspondence between the versions of the *Vāyu* and *Matsya Purāṇas* with regard to the number, names, and total duration of the reigns of the princes following this Puḷumāvi (I) has been made clear above. A study of the table of the Sātavāhana princes enumerated by the different *Purāṇas* will also convince us that next to the name Sātakarṇi that of Puḷumāvi was one of the most popular among the Sātavāhanas. The *Matsya* gives as many as three of the line bearing this name. The importance of the name Puḷumāvi is further emphasized by the statement in the *Matsya*, viz.: ‘as to the Āndhras, they are the Puḷomas, that is, those succeeding the king Puḷumāvi.’² The significance of this parenthetical statement seems to be largely lost on those scholars who were intent on proving that the Sātavāhanas were rulers of the Andhra country only from the time of Puḷumāvi (II), the son of Gautamīputra Sātakarṇi. The truth seems to be that with Puḷumāvi (I), the fifteenth of the *Matsya* list, the kingdom of the Sātavāhanas ceased to be confined merely to the Deccan. It took long strides and extended itself rapidly, when this Puḷumāvi, destroying the last of the Kaṇvas, helped it to assume the role of an All-India power.

v1.c7.s4.t2.33

² Pargiter, *Dynasties*, p. 58; *JBORS*, iii, 248.

v1.c7.s4.t2.34

Puḷumāvi (I), then, was ruling over the Sātavāhana kingdom of Dakṣiṇāpatha when he extinguished the Kaṇva power of Magadha in 28 B.C. The *Matsya Purāṇa* gives this king thirty-six years of rule, while the *Vāyu* makes Nāsik cave inscription of Gotamī Balaśrī, the mother of Gautamīputra Sātakarṇi.¹ It is dated in the 19th regnal year of her grandson, Vāśiṣṭhīputra Puḷumāvi, and refers to the glories of her son’s achievements. Gautamīputra is credited with victories over the Śakas, Yāvanas, and Pahlavas. He is also stated to have destroyed the race of the Khakharāṭa (Kṣaharāṭa).

v1.c7.s4.t2.35

¹ EI, viii, Nāsik No. 2.

v1.c7.s4.t2.36

It is fairly certain that among these Khakharāṭas must be counted either the Kṣatrapa Nahapāna, or one of his successors, not far removed from him in time. The conquests of Gautamīputra, recapitulated by his mother, include those lands where the influence of Nahapāna’s son-in-law, Uṣabhadāta, and of Nahapāna’s minister, Ayama, had earlier been paramount. Again a large hoard of coins, all bearing the name of Nahapāna, has been discovered at Jogalthembi. Most of these coins are found restruck by Gautamīputra. If this does not definitely mean that Nahapāna himself was supplanted and lost his throne, it is at least quite certain that his successors were replaced by Gautamīputra. Nahapāna’s last known date, found on the record of his minister Ayama at Junnar, viz. year 46, may in fact actually indicate the year of Gautamīputra’s conquest of the Khakharāṭas. But this year 46 is not referred to any era, nor do we know the time lapse between Nahapāna’s end and the said victories of Gautamīputra.

v1.c7.s4.t2.37

Perhaps it was this very Nahapāna that the *Periplus* intended when it referred to Mambares or Nambanus. But the date of this work is variously fixed, and the Mambares-Nahapāna identity is questioned.

v1.c7.s4.t2.38

If Nahapāna’s last years could be fixed with any reasonable accuracy, the earliest limit of Gautamīputra’s reign could then be determined from the year when the Khakharāṭas were vanquished. But even here we are confronted with a great many difficulties. Ptolemy, the Greek geographer, seems to suggest that Caṣṭana of Ujjain and Puḷumāvi of Paiṭhan were contemporaries. But Caṣṭana’s dates are not known, except that he cannot have reigned long after A.D. 130, the date of the Andhau inscription which shows his grandson Rudradāman holding equal power with him. Writers like Prof. Rapson think that there must have been a break in between the reigns of Caṣṭana and Rudradāman. But opinions are not wanting that they may have ruled conjointly.

v1.c7.s4.t2.39

Ptolemy died in A.D. 161. But scholars differ as to the dates or the periods of which he wrote. If his information could be exactly dated, it would not be difficult to find the dates of Puḷumāvi’s reign. We might then proceed to identify the earliest date of Puḷumāvi as being also the latest of his father, Gautamīputra. But here the Bhandarkars postulated a joint rule for father and son. This, if accepted, introduces another uncertain element.

v1.c7.s4.t2.40

These several difficulties must then be got over and reasonably explained in any scheme of chronology suggested for the later Sātavāhanas. It will be seen in the sequel that our date, A.D. 62-86, for Gautamīputra Sātakarṇi meets each of the requirements.

v1.c7.s4.t2.41

The question of Nahapāna’s exact date is regarded as important for the student of later Andhra chronology. But unfortunately there is considerable uncertainty about it. The inscriptions at Nāsik and Kārle of Nahapāna’s son-in-law, Uṣabhadāta, and that of his minister Ayama at Junnar, show that it ran from year 41 to 46 of an unknown era. As the term Varṣa stands for the year everywhere in the inscriptions of the Western Kṣatrapas, as it does in the majority of the Śaka dates,¹ Prof. Rapson referred the dates of Uṣabhadāta’s and Ayama’s inscriptions to the Śaka era. He argued

v1.c7.s4.t2.42

¹ Kielhorn, IA, xxvi, 153.

v1.c7.s4.t2.43

. . . it would seem improbable that Nahapāna’s reign could have extended much beyond the last recorded year 46 = A.D. 124. Gautamīputra’s conquest of Nahapāna seems undoubtedly . . . to have taken place in the 18th year of his reign. We therefore have the equation: Gautamīputra’s year 18 = A.D. 124, or A.D. 124+x. On this synchronism, on the recorded regnal dates in the inscriptions of other Andhra sovereigns, and on the known date, 72 = A.D. 150, of Rudradāman as Mahākṣatrapa rests at present the whole foundation of the later Andhra chronology.²

v1.c7.s4.t2.44

The supposition is that Nahapāna did not live long after A.D. 124, when he came into direct conflict with Gautamīputra Sātakarṇi. The latter’s triumph was so complete and for Nahapāna so disastrous that Gautamīputra well deserved his mother’s praises that he was the Khakharātavasa niravasesakara. It is sought to strengthen the view that Gautamīputra destroyed Nahapāna himself by citing the evidence of the hoard of coins discovered at Jogalthembi. Of the 13,270 coins, about 9,270 had been countermarked by Gautamīputra; but not even one of those 13,000 and more bears the name of any Kṣatrapa other than Nahapāna. If there were any Kṣatrapa princes between Nahapāna and Gautamīputra, they would have struck coins in their own name, as Bhūmaka had done before Nahapāna. Nahapāna cannot have been a remote predecessor of Gautamīputra. If additional support for this view were required, we have it in Nāsik No. 4 and Kārle No. 19. The latter refers to the gift of a village of Karajaka to some monks by the king—whose name is missing. But this village was certainly the same as that earlier granted to the same monks by Nahapāna’s son-in-law, Uṣabhadāta.³ Nasik No. 4 transfers a field held by Uṣabhadāta, presumably Nahapāna’s son-in-law, to the monks mentioned therein by an order to the minister in Govardhana. Prof. Rapson therefore declares that ‘Gautamīputra’s conquest of Nahapāna seems undoubtedly to have taken place in the 18th year of his reign’,² which, as shown above, was A.D. 124. This places Gautamīputra’s accession in A.D. 106.

v1.c7.s4.t2.45

² CIC, p. xxvii.

v1.c7.s4.t2.46

³ EI, vii, Kārle No. 13.

v1.c7.s4.t2.47

Before we discuss the point of view presented above, we may, however, note that Gautamīputra’s conquests did not result literally in the extinction of the Khakharāta race. Nahapāna was sovereign of more extensive regions than were conquered by Gautamīputra. The place-names in Balaśrī’s record at Nāsik certainly include Ushabhadāta’s sphere of political influence which comprised South Gujerāt, the Northern Konkan from Broach to Sopārā, and the Nāsik and Poona districts. Even after these were conquered by the Andhras, the Khakharātas still retained the districts farther north ‘since Puṣkara in Ajmer, the place of pilgrimage to which Ushabhadāta resorted after his victory over the Mālayas, may be supposed to have lain within the dominions of his lord (bhaṭṭāraka)’. Nor is there any evidence in Balaśrī’s record to show that Gautamīputra destroyed Nahapāna himself. The mother of a mahārāja and the grandmother of another mahārāja, Gotamī Balaśrī could not be ignorant of the significance of her son’s victory over Nahapāna, who was not less important than any other of his contemporary monarchs. We may, therefore, remark with Dr. V. A. Smith that ‘it is not necessary to believe that Gautamīputra Āndhra fought with Nahapāna personally’.¹

v1.c7.s4.t2.48

¹ EHI, 4th edition, p. 232.

v1.c7.s4.t2.49

With reference to the name Ushabhadāta appearing in the Nāsik record, No. 4, Dr. Bhandarkar felt that too much had been presumed on the ground that ‘this name (Ushavadata) is not very common.’² Prof. K. A. Nilakanta Sastri says: ‘it is very unfortunate that all the considerations urged so far about the Ushavadata inscriptions should have the effect of throwing serious doubts on what appeared a fairly secure foundation for Andhra chronology, viz. Professor Rapson’s equation: Gautamīputra’s year 18 = 46 Nahapāna + x = A.D. 124 + x (catalogue xxvii).’³ The record referred to (Nāsik No. 4) registers a grant of a field in Western Kakhadi, a field which was held by Ushabhadāta. The relevant passage runs thus: ya khetam ajakālakiyam Ushabhadātena bhūtam. Bühler and Bhagavanlal Indraji took the crucial word ajakāla for the Prākrit form of adyakāla. Bühler understood the passage as ‘the field which has been possessed by Rishabhadatta up to the present time’. Senart denied that ajakālakiyam stood for adyakālikam and held that this word was the name of the field.⁴ Ajakālaka as a proper name does appear in a Bhārhut inscription.⁵ While this looks plausible enough, Prof. Nilakanta Sastri denies that this ‘Ushabhadāta’, who held a single field in Govardhana, was the son-in-law of Nahapāna.³ Ushabhadāta may not have been such a rare name. But there is nothing in the record to suggest or imply that this Ushabhadāta did not hold many more or even all the other lands of the village. These districts had certainly formed part of Ushabhadāta’s dominions during Nahapāna’s reign. The objection raised by Prof. Sastri ignores the context of the grant. It would surely not be required of the donor of a field which had earlier belonged to another individual to recount all the other possessions of that individual, or to make out a list of all those owners with properties similar to or other than the subject of the grant. More serious is Senart's objection to Bühler's and Indraji's equation of *ajakāla* with *adyakāla*. If this objection be valid then we cannot be certain that Gautamīputra came into the possession of those lands so soon after Uṣabhadāta. The *ajakāla* of the passage may, however, stand for *adyakāla*. In that case, the passage may mean 'the field which had been Uṣabhadāta's of old'. This would suggest a lapse of time between Uṣabhadāta and Gautamīputra.

v1.c7.s4.t2.50

² IA, 1918, p. 73 and Senart, EI, vii, 56.

v1.c7.s4.t2.51

³ JRAS for 1926, p. 657.

v1.c7.s4.t2.52

⁴ EI, viii, p. 72.

v1.c7.s4.t2.53

⁵ Lüders, No. 795.

v1.c7.s4.t2.54

Far from supporting Prof. Rapson's theory, 'the study of the great Jogalthem bhi hoard of more than 13,000 coins of Nahapāna proves that the coinage extended over many years, although always bearing the name of Nahapāna, who . . . was dead before Gautamīputra extirpated his family or clan.'¹ Pandit Bhagavanlal Indraji believed that Nahapāna lived to a ripe old age, the changing features of the face on the coins corresponding to the increase in his age. The Rev. H. R. Scott's was the most minute and thorough study of the coins.² He found that the coins showed 'indeed faces varying greatly in age and not in age only, but in every feature'. He suggests that 'we may have here the coins not of a single king, but of a series of kings, sons and grandsons perhaps of Nahapāna, who retained on their coins the name of their great ancestor as a title of honour . . .'. The Rev. H. R. Scott finally suggests that 'judging from the condition of the coins, I should say that they must have been a very long time in circulation, and that both before and after being counter-struck'.

v1.c7.s4.t2.55

¹ V. A. Smith, EHI, 4th edition, p. 232.

v1.c7.s4.t2.56

² J BBRAS, 1907, pp. 223 ff.

v1.c7.s4.t2.57

The theory of a Gautamīputra-Nahapāna synchronism thus appears to be insecurely founded. It is shattered completely by the discovery of the Andhau (Cutch) inscriptions,³ dated in year 52 (Śaka) = A.D. 130 and in Rudradāman's time. The quaint wording of the inscriptions, viz. 'Rano Cashtanasa Yasamotikaputrasa, rano Rudradāmasa Jayadāma putrasa vasa 52 . . .', permits the inference that Chashtana and Rudradāman ruled conjointly. This suggestion by Prof. R. C. Majumdar is accepted by Dr. D. R. Bhandarkar.⁴ Prof. R. D. Banerji, taking the earlier suggestion by Dr. Bhandarkar that the word *putrasya* was omitted, refuses to have anything to do with the theory of such a conjoint rule.⁵ But as Dr. Gopalachari pointed out, unlike the other records of the Western Kshatrapas of the Chashtana line, in the Andhau inscriptions, the names of Chashtana and Rudradāman precede those of their fathers.⁶ The practice of father and son ruling concurrently as Mahākshatrapa and Kshatrapa respectively is found among the later Western Kshatrapas. Chashtana was a Mahākshatrapa. Jayadāman is known to have been only a Kshatrapa. Rudradāman 'won for himself the name of Mahākshatrapa'. While Chashtana was a Mahākshatrapa, Jayadāman predeceased him as a Kshatrapa, to which place Rudradāman succeeded. After Chashtana's death (some time between 52 and 72) the Kshatrapa Rudradāman would have become a Mahākshatrapa in his turn.¹

v1.c7.s4.t2.58

³ Edited by D. R. Bhandarkar and R. D. Banerji, see J BBRAS, xxxiii; EI, xvi, 19 ff.; JRAS, 1917.

v1.c7.s4.t2.59

⁴ IA, xlvii, 154, n. 26.

v1.c7.s4.t2.60

⁵ EI, xvi, 22-3.

v1.c7.s4.t2.61

⁶ EHAC, p. 52.

v1.c7.s4.t2.62

¹ Prof. K. A. Nilakanta Sastri, *JRAS*, 1926, p. 660.

v1.c7.s4.t2.63

But Prof. Rapson believes that there was a break between the reigns of Chashtana and Rudradāman.² According to him this interval, during which there was no Mahākshatrapa, may have been the result of some defeat, probably at the hands of Gautamīputra. Mr. Hari Charan Ghose has delivered a vigorous attack on the theory of joint rule propounded above.³ This theory is based upon ‘unproven and gratuitous assumption’, he declares. His main arguments are that it ‘ignores the fact that Jayadāman ruled as an independent Kshatrapa and also assumes that he died before Chashtana’; the wording of the Andhau inscriptions makes no distinction at all between the titles of Chashtana and Rudradāman, and if conjoint rule existed, at any rate one of the two must have been pre-eminent. Again Cutch is one among the several conquests of Rudradāman recounted in his Girnār inscription dated in the year 72 (A.D. 150).

v1.c7.s4.t2.64

² *CIC*, p. cxvii; *IA*, 1918, p. 154.

v1.c7.s4.t2.65

³ *IHQ*, vi, 1930, p. 755.

v1.c7.s4.t2.66

‘Nowhere have we any evidence that Cutch was within the territory of Chashtana. Coins do not prove it, while Ptolemy's evidence distinctly runs counter to it, who places the Cutch region not under his sway . . . The Andhau records belong to the reign of Rudradāman alone . . . The records show that in *c.* A.D. 130 Rudradāman alone was ruling. So his father Jayadāman ruled before him and before Jayadāman ruled Chashtana.’

v1.c7.s4.t2.67

This contention of Mr. Hari Charan Ghose leaves the wording of the record unexplained: why should the names of Chashtana and Rudradāman have appeared together with similar titles and why should Jayadāman's have been mentioned with no ceremony at all except to suggest that he was the father of Rudradāman? If no earlier evidence proves Chashtana to have been lord of Cutch, and if the inscription gives identical titles to Chashtana and Rudradāman, the explanation is simple. Cutch must have been brought under Chashtana's rule very recently by the exertions of Rudradāman. *Chashtana was still living in the year 52 (A.D. 130).* But he was so advanced in the infirmities of old age that he had practically retired from active rule. Hence the theory of the joint-rule of Chashtana and Rudradāmana, which is indicated by the Andhau inscriptions, stands unimpaired. Ptolemy noticed the contemporaneity of Chashtana and Puḷumāvi. It is admitted on all hands that the year 52 of the Andhau inscriptions is dated in the Śaka era. If Nahapāna's last known year, as noted in the Junnar record of Ayama, also belongs to that era, the interval of six years is too short for (1) the end of Nahapāna's reign, (2) the destruction of the Khakharāṭas, (3) Gautamīputra Sātakarṇi's rule and his death, (4) the accession of his son Puḷumāvi, (5) Chashtana's reconquest of the lands between Cutch and Malwa, and (6) his rise to power in these regions as Mahākshatrapa.

v1.c7.s4.t2.68

Mr. Hari Charan Ghose demonstrates ‘a territorial conflict which cannot be explained away on any hypothesis¹—a conflict which arises from the assumption that Uṣabhadāta’s and Ayama’s years were dated in the Śaka era. The history of the Nāsik and Kārle regions illustrates this quite clearly.² These important districts are not specifically mentioned in Balaśrī’s record. Prof. Rapson explains it on the plea that not all the dominions of Gautamīputra are included in the place-names found in the epigraph. The better and more reasonable assumption would be that one of those geographical names comprehends the Nāsik and Kārle districts as well. Such a name was Aparānta. The Mārkaṇḍeya Purāṇa places the Nāsik region in the Aparānta country. In the Sahyādrikhaṇḍa, the country of Aparānta is extended to include places other than those in Konkan. A record from Kānheri (certainly a place in Aparānta) refers to the donor as a Nāsikaka.³ We may now proceed with an account of these regions; and the following table may be studied with advantage:

v1.c7.s4.t2.69

¹ H. C. Ghose, *IHQ*, vi, 750 ff.

v1.c7.s4.t2.70

² For Prof. Rapson’s summary of the history of N. Mahārāṣṭra and Aparānta, see his introduction to *CIC*, p. 695. Also Banerji in *JRAS*, 1917, p. 276. Prof. Nilakanta Sastri is unable to accept Mr. Rapson’s theory, see *JRAS*, 1926, p. 659, n. 2.

v1.c7.s4.t2.71

³ Lüders, No. 985; *IHQ*, vi, 752.

v1.c7.s4.t2.72

| *Gotamī Balaśrī’s Nasik record dated in 19th year of Puḷumāvi. Gautamīputra’s dominions.* | *Girnār Inscription of Rudradāman dated 72 (150 A.D.) Rudradāman’s dominions.* | *Rudradāman’s conquests from the Andhras.* | *Retained by the Andhras after Rudradāman’s conquests.* |
| Asika, Asaka, Muḷaka, Suraṭha, Kukura, Aparānta, Anūpa, Vidabha, Ākara and Avanti. Mountains Vijha, Chavata, Pāricāta, Sahya, Kanhagiri, Maca, Siritana, Malaya, Mahida, Seṭagiri, and Cakora. | Eastern and Western Ākarāvanti, Anūpa, Anartta, Surāṣṭra, Srabhra, Maru, Kachcha, Sindhusauvira, Kukura, Aparānta, Niṣāda, and other territories. | Suraṭha, Kukura, Aparānta, Anūpa, Ākara, and Avanti. | Asika, Asaka, Muḷaka, Vidabha, and other portions of the Deccan. |

v1.c7.s4.t2.73

The Aparānta country which included Nāsik and Kārle was a western Kṣatrapa dominion until the year 46. In that year, which was his 18th, Gautamīputra wrested it from them. He held it for the remaining six years of his rule. Puḷumāvi’s inscription dated in his twenty-second year was found at Nāsik, a fact which shows that he retained those districts in his power. Two of Puḷumāvi’s successors, who together ruled for fourteen years according to the Purāṇas, continued in possession of Aparānta. Then came Yajña Śrī Sātakarṇi, whose records show him as lord of Nāsik in his 7th year. Thus from the year 46 to 101, Aparānta must have been under the Andhras.⁴ But the Girnār inscription presents Rudradāman as Lord over the Aparānta country actually in the year 72 (A.D. 150). Here is a riddle which defies solution so long as Ushabhadāta’s and Ayama’s years are also referred to the same (Śaka) era.

v1.c7.s4.t2.74

⁴ Mr. Ghose evidently gives only twenty-two years for Puḷumāvi and dates the last year of Andhra hold on Nāsik as 96 (if year 46 was its beginning).

v1.c7.s4.t2.75

Any chronological scheme which treats Ushabhadāta’s and Ayama’s years as of the Śaka era is thus found to be very unsatisfactory. Two ways were open to scholars to get over the difficulty created by such a scheme. One was to attribute those years to some other era. This was actually done and we shall come to it presently. The second way out was to postulate the theory of a conjoint rule for Gautamīputra Sātakarṇi and his son Vāsiṣṭhīputra Puḷumāvi. Sir R. G. Bhandarkar suggested this long ago and Dr. D. R. Bhandarkar supported and restated it in 1914 and 1918.¹ Prof. Rapson’s equation and the epigraphic evidence which gives Gautamīputra Sātakarṇi a reign of twenty-four years, take his last date to A.D. 130 (A.D. 124+x was Gautamīputra’s 18th year), which is the date of Rudradāman’s Andhau inscriptions. If Ptolemy’s statement that Chasṭana and Puḷumāvi were contemporaries be valid—and none has so far doubted it, then Prof. Rapson’s scheme, by which the Bhandarkars swear, requires the conjoint rule not merely of Chasṭana and Rudradāman but of Gautamīputra and Puḷumāvi as well. This point having been noted, we may now enter on a critical analysis of this theory of the Bhandarkars.²

v1.c7.s4.t2.76

¹ JBBRAS, xxxiii, ‘Epigraphic Notes and Questions’; IA, 1918, ‘The Deccan in the Sātavāhana period’.

v1.c7.s4.t2.77

² Sir R. G. Bhandarkar did not have before him the evidence of the Andhau inscriptions. These were discovered later and used by D. R. Bhandarkar.

v1.c7.s4.t2.78

We have already observed that Sir R. G. Bhandarkar sought to explain the differences among the *Purāṇas* in the reckoning of the total duration of the Andhra dynasty by the plea that the shorter period referred to the main branch of the family which succeeded to the throne at Dhanakaṭaka, while the longer period included the reigns of subsidiary princes who ruled at Paiṭhan but died before they became kings at Dhanakaṭaka. Nāsik inscription No. 2 calls Gotamī Balaśrī the mother of the Great King and grandmother of the Great King, these monarchs being identified as Gautamīputra Sātakarṇi and Vāsiṣṭhīputra Puḷumāvi respectively. Nāsik record No. 3 refers to Gautamīputra as *Dhanakaṭasami*. The inference is that both father and son were ruling simultaneously, the former at Dhanakaṭaka, identified with Dharanikoṭa, and the latter at Paiṭhan. If her son and grandson were great kings at different times, he argued, Gotamī Balaśrī could not have any ‘special claim’ to honour, for ‘every queen belonging to a dynasty in power is the mother of a king and grandmother of a king’.³

v1.c7.s4.t2.79

³ EHD, pp. 32-3 and n. 17.

v1.c7.s4.t2.80

Prof. Nilakanta Sastri has pointed out that Nāsik inscription No. 5 in the year 24, which is a continuation of No. 4 dated in the eighteenth year of Gautamīputra, specifically mentions Gotamī Balaśrī as *Jivasutā*, and that her son was then living, whereas a postscript to Nāsik inscription No. 2, dated in the nineteenth year of Puḷumāvi, records a gift, the merit being made over to the period.

v1.c7.s4.t2.81

father (*Pitupatiyo*). As pointed out by Senart, ‘such an application of merit can be made only in favour of deceased persons’.¹

v1.c7.s4.t2.82

¹ *EI*, viii, 65.

v1.c7.s4.t2.83

Sir R. G. Bhandarkar would argue that ‘if it was a fact that Gautamīputra was dead, when the cave-temple was dedicated and Puḷumāyi alone was reigning’, one certainly would not expect ‘that a king who had been dead for nineteen years should be highly extolled in the inscription and the reigning king altogether passed over in silence.’ But, as Prof. Śastri remarks, this absence of a eulogy on Puḷumāvi ‘cannot by itself form the foundation of so far-reaching a theory as that of the simultaneous rule of these two princes’.² We may with greater reason ask, if it be true that her son was alive at the time, how it was that Gotamī Balaśrī, who so elaborately extolled her son, failed to date the record in *his* regnal year. The relevant inscriptions come from the same place and are even found in the same cave. With Mr. Hari Charan Ghose we may observe that ‘it is certainly very astonishing that kings who were supposed to be ruling together should never have been mentioned together’.³ The explanation seems to be that, as Prof. Dubreuil pointed out,⁴ the excavation was begun under the orders of Gautamīputra and that this inscription was ‘the funeral oration on the great king delivered by an inconsolable mother’. The solicitude of Puḷumāvi for the maintenance of the ascetics living in the Queen’s cave also points to the fact that he (Puḷumāvi) was then ruling alone.

v1.c7.s4.t2.84

² *JRAS*, 1926, p. 647; see also R. D. Banerji in *JRAS*, 1917, p. 281.

v1.c7.s4.t2.85

³ *IHQ*, vii, 1931, p. 117.

v1.c7.s4.t2.86

⁴ *AHD*, pp. 38 ff.

v1.c7.s4.t2.87

One more argument remains to be considered which is inferred by Dr. D. R. Bhandarkar from the Nāsik inscriptions. Nāsik No. 5 records an order in the name of Gautamīputra and his mother. It runs: ‘we have here on the mount Tiraṇhu, formerly given to the mendicant ascetics dwelling in the cave which is a pious gift of ours, a field in the village of Kakhaḍi.’ This record is dated in the 24th year. The queen-mother had dedicated this very cave to the monks of the same hill in the 19th year of Puḷumāvi. This latter, therefore, must have preceded the former. If then Gotamī was a *jivasutā* (having a son living) at the time of the later record, would she be otherwise six years earlier when the dedication was originally made in the 19th year of Puḷumāvi? He goes even further and states that although the Nāsik record No. 5, dated in the 24th year, records a donation of Gautamīputra, it must be referred not to his, but to Puḷumāvi’s reign. But, as shown by Prof. Nilakanta Śastri, the answer to this argument is minutely archaeological and even architectural. This argument is admirably treated by Mr. R. D. Banerji.⁵ He says

v1.c7.s4.t2.88

⁵ *JRAS*, 1917, pp. 281 ff.: and ibid., 1925. For an account of the cave, see Havell, *AMAI*, pp. 80-1.

v1.c7.s4.t2.89

‘The cave which is called *āmha-dhama-dāne leṇe* in inscription No. 5 by Gautamīputra Sātakarṇi is not cave No. 3 as it stands to-day. The cave to which the phrase . . . refers was not dedicated in the 19th year of Puḷumāvi. . . . The grants of land made by Gautamīputra Sātakarṇi which are recorded in inscriptions Nos. 4

v1.c7.s4.t2.90

and 5 were not given to the monks living in the larger cave (cave No. 3 as it stands to-day) excavated by his mother; but to the ascetics living in the smaller cave in the north-eastern corner of cave No. 3, who are called the Tekirasi ascetics in inscription No. 4.¹

v1.c7.s4.t2.91

¹ *JRAS*, 1925, p. 7.

v1.c7.s4.t2.92

In the 18th year of Gautamīputra the cave was planned on an ambitious scale but was constructed only in part. Six years later, when his health was on the decline, his mother took it up, but she took a long time over it and the whole construction was completed only in the 19th regnal year of Puḷumāvi. Then came her great inscription, Nāsik No. 2. The great eulogy of Gautamīputra's exploits, as well as the dating of the record in Puḷumāvi's reign, are thus easily explained. Prof. Nilakanta Śastri's interpretation of the Nāsik records on the basis of the method of their dating is quite revealing. In all those cases where the dates come at the end, the word *ānapayati* is found introducing a royal order. On the other hand those cases where dates appear in the beginning are records of private or non-official benefactions. The former must be referred to the reigns of princes issuing the order: the dates 18 and 24 of Nāsik inscriptions 4 and 5 would then belong to the reign of Gautamīputra and 'could not by any means be transferred to Puḷumāvi'.² Puḷumāvi must have succeeded his father *in due course*.

v1.c7.s4.t2.93

² *JRAS*, 1926, pp. 648-50.

v1.c7.s4.t2.94

If, then, Prof. Rapson's arrangement of the Andhra chronology cannot be sustained without the hypothesis of a conjoint rule of Gautamīputra and Puḷumāvi, the question which would naturally arise is: to which era should we attach the years of Ushabhadāta's and Ayama's inscriptions? If they are not of the Śaka era and furnish so very little information for the period of the later Sātavāhanas, have they in fact any great importance for the study of the earlier Sātavāhana history?

v1.c7.s4.t2.95

General Cunningham suggested a long time ago that Nahapāna's years must be referred to the Vikrama era, which brings Nahapāna's last known date to 12 B.C. This would mean a very long interval between Nahapāna and Chashṭana. This was not acceptable to Prof. Rapson,³ who would however place Bhūmaka, Nahapāna's predecessor, much near Azes in time, as the result of a study of Bhūmaka's coins.⁴ As he would have resigned himself to place him far later than Bhūmaka. Later studies, however, shortened this interval by a great many years, so that Nahapāna was placed much nearer to Azes than he would have been allowed to be by the chronologists of the Rapson school. Mr. R. D. Banerji is definite that Ushabhadāta's inscriptions must have been earlier than the Mathurā inscription of Soḍāsa, dated in year 72.⁵ Their characters are indeed so very much earlier that 'it is impossible to place Rudradāman twenty-six years after Nahapāna'. This view of Mr. Banerji finds ample corroboration in the observations of Bühler in his work, *Indian Palaeography*. According to him, while the script of Girnār prasasti agrees with the later southern alphabet, ‘the traces of the southern peculiarities are faint or entirely wanting’ from the records of Uṣabhadāta at Nāsik. What is more, the script of Uṣabhadāta’s epigraphs is ‘very similar’ to that of the inscriptions of Gautamīputra or of Puḷumāvi.¹ But ‘it is a similarity which is only to be expected’, says Prof. Nilakanta Sastri, ‘if the area of their location had been ruled by the Khakharāṭas for some time before Gautamīputra recovered it for the Sātavāhanas’.² He, therefore, found no difficulty at all in bringing Uṣabhadāta’s dates nearer to Soḍāsa’s date, viz. year 72.

v1.c7.s4.t2.96

³ *JRAS*, 1899, p. 365.

v1.c7.s4.t2.97

⁴ *JRAS*, 1904.

v1.c7.s4.t2.98

⁵ *IA*, 1908, p. 63.

v1.c7.s4.t2.99

¹ IA, 1904, App., pp. 41-3.

v1.c7.s4.t2.100

² JRAS, 1926, p. 652.

v1.c7.s4.t2.101

The coins of Nahapāna discovered at Jogalthembi definitely support the above hypothesis. As on the coins of the Parthians and on those of Hagāna and Hagāmasha, the Satraps of Mathurā, so also on Nahapāna’s coins the arrow and the thunderbolt make their appearance. The Satraps mentioned above are separated from Soḍāsa only by one more ruler, by Rājuvula, the father of Soḍāsa, of the year 72. Dr. V. A. Smith would, therefore, certainly connect the Western Kshatraps of Junnar with those of Mathurā.³ After a close study of the different language-scripts of the legends on Nahapāna’s coins, the Rev. H. R. Scott stated that ‘Bühler’s tables give specimens of Brāhmī ranging from 350 B.C. to A.D. 350.’⁴ The characters on our coins belong to ‘somewhere near the middle of that period’, which is nearly the time ascribed to Nahapāna if his years were of the Vikrama era. Nahapāna was distinctly earlier than Chasṭana, and, if further argument be required, we have it from Prof. Dubreuil, who claimed a very early date for the style of Nahapāna’s monuments.⁵ The architectural characteristics of Nahapāna’s monuments at Nāsik and Kārle ‘bespeak a date approximately contemporary with the Sāñchī Toraṇas’. Sir John Marshall assigned these to the latter half of the first century B.C.⁶ Messrs. Fergusson and Burgess believed that Nahapāna’s hall at Nāsik must have belonged to the same period.⁷ Now that Prof. Sten Konow has referred Soḍāsa’s year 72 to the Vikrama era,⁸ Prof. Nilakanta Sastri would accept Cunningham’s suggestion, because ‘it seems to accord best with all the evidence and to make an intelligible story of the relations between the Sātavāhanas and their foreign enemies’.⁹ Mr. V. S. Bakhle also argued on the above lines.¹⁰ He fixed 17 B.C. as Nahapāna’s earliest date and assigned the period between 25 and 20 B.C. for the conquest of Mahārāṣṭra by the Kshatrapas. Like General Cunningham and Prof. Nilakanta Sastri, Mr. Bakhle also ascribed Nahapāna’s years to the era that started with 58 B.C., but which according to him was founded by Azes I. Whether it was the era of Azes or the Vikrama era the conclusion looks extremely obvious. If Rudradāman’s Andhau and Girnār records of years 52 and 72 were Śaka years, and if the conjoint rule of Gautamīputra Sātakarṇi and his son Puḷumāvi is far from admissible, Nahapāna’s years and those of Chashṭana’s dynasty cannot possibly belong to the same era.¹

First 111 of 157 passages. Read on

Cites (17)