¹ I am indebted to Professors J. Duncan M. Derrett, Robbins Burling, and P. S. Jaini for criticisms and suggestions on a draft of this paper, though of course the final version is not to be taken as representing their views. Research was supported by the American Council of Learned Societies, the National Endowment for the Humanities, and the Horace H. Rackham School of Graduate Studies, University of Michigan.

² [e*][hy e]hīty upaguhya bhāvapiśunair utkarṇitai romabhiḥ

sabhyeṣūcchvasiteṣu tulya-kulaja-mlānānanodvīkṣi[ta]ḥ |

[sne]ha-vyāḷulitena bāṣpa-guruṇā tattvekṣiṇā cakṣuṣā

yaḥ pitṛābhihito ni[rīkṣya] nikhi[lāṃ pāhy tva*][m urvī]m iti ‖

The text of Fleet, J. F., Inscriptions of the early Gupta kings (Corpus Inscriptionum Indicarum, III), 1888, 6ff.Google Scholar, with the emendations of Chhabra, B. Ch. (“Chandragupta's abdication”, Indian Culture, 14, 1948, 141 ff.)Google Scholar. Fleet's translation follows the grammatical structure of the text more closely, but in that it gives the reactions of the courtiers and princes before the embrace which causes those reactions it is aesthetically false. The order of ideas in the original is logical and well motivated.

³ Op. cit.

⁴ See Bhandarkar, D. R., “Identification of the princes and territories mentioned in the Allahabad pillar inscription of Samudragupta”, IHQ, 1, 1925, 250–260Google Scholar.

⁵ anu-Gaṅgā Prayāgam ca Sāketam Magadhāṃs tathā |

etān janapadān sarvān bhokṣyante Gupta-vaṃśajāḥ ‖

(Pargiter, F. E., The Purāṇa text of the dynasties of the Kali Age, O.U.P., 1913, 53)Google Scholar. Goyal, S. R. (A history of the Imperial Guptas, Allahabad, 1967, ch. iiGoogle Scholar) has a good discussion of the problem of the original home of the Guptas. I differ with him, however, when (p. 51) he adopts the reading anu-Gaṅgā Prayāgam Māgadhā Guptāś ca bhokṣyanti in the Puraṇic passage above, following the Viṣṇu Purāṇa (Pargiter, 53, n. 8), and takes it to mean that the Licchavis (Māgadhas) and Guptas jointly ruled the territory along the Ganges up to Prayāga. This reading will not do: the line does not scan. Of 4,562 half-ślokas from the Sabhāparvan analysed by Barend A. van Nooten of the University of California, Berkeley (“The śloka in the Sabhāparvan”, J. Or. Inst., Baroda, 17, 4, 1968, 353–362)Google Scholar, only one occurrence of ⌣--⌣ in the final pāda was recorded.

⁶ “Notes on the Gupta coinage: I”, J. Numis. Soc. India, 19, 2, 1957, 135–144Google Scholar.

⁷ S. R. Goyal, A history of the Imperial Guptas, 89–97; “The attribution of Candragupta-Kumāradevī coin-type”, Ind. Numis. Chron., 4, 2, 1965, 104–115Google Scholar. The phrase Bhavanāga-dauhitra occurs in the Chammak copper-plate inscription of Pravarasena II, 1. 7: Sircar, Dines Chandra, Select inscriptions bearing on Indian history and civilization, I, 2nd ed., University of Calcutta, 1965, 442Google Scholar; Fleet, op. cit., 236 ff.

⁸ See de Coulanges, Numa Denis Fustel, The ancient city [La cité antique, 1864], New York, n.d., 77Google Scholar. This classic reconstruction of IE law is much neglected by Indianists.

¹⁰ yasyāstu na bhaved bhrātā na vijñāyeta vā pitā |

nopayaccheta tāṃ prāj aḥ putrikādharmaśaṅkayā ‖

¹⁰ See Kane, MM. Pandurang Vaman, History of Dharmaśāstra, Poona, 1930–, III, 657–659Google Scholar, for this and other passages.

¹¹ Pathak, op. cit., 141.

¹² Summarized in Kane, III, 685.

¹³ janayitur asaty anyasmin putre sa eva dvipitṛko dvigotro vā dvayor api svadhārikthabhāg bhavati.

¹⁴ aputreṇa parakṣetre niyogotpāditaḥ sutaḥ |

ubhayor apy asau rikthī piṇḍadātā ca dharmataḥ ‖

¹⁵ This statement needs some qualification, e.g. some jurists say the adopted son is dvyāmusyāyaṇa if some saṃskāras including the tonsure (caula) are performed by the genitor, and others including the initiation (upanayana) by the adoptor (see Kane's discussion, pp. 685–7). But this does not affect the putrikā-putra.

¹⁶ The Bhitari pillar inscription of Skanda Gupta, Fleet, 53 ff.; Sircar, 321 ff.

¹⁷ Pargiter, op. cit., 50.

¹⁸ Op.cit., 135–9.

¹⁹ Manu 10.5, 6, and ff.

²⁰ Idha bho brāhmaṇo ubhato sujāto hoti mātito ca pitito ca saṃsuddhagahaṇiko yāva sattamā pitāmaha-yugā akkhito anupakkuṭṭho jāti-vādena, Dīgha Nikāya (P.T.S. ed.), I, 120; cf. Majjhima Nikāya, II, 165 and ff.

²¹ Culled from Sircar, Dines Chandra, The successors of the Sātavāhanas in the lower Deccan, University of Calcutta, 1939Google Scholar.

²² Lüders, H., Mathurā inscriptions, Göttingen, 1961, nos. 115, 176; 18, 168; 162; 187Google Scholar.

²³ Chattopadhyāya, K. P. in JASB, 1927, 503 ff.Google Scholar; 1939, 317–39; criticized in Raychaudhuri, Hemachandra, Political history of ancient India, 6th ed., University of Calcutta, 1953, 409, n. 4Google Scholar.

²⁴ The chiefs of the Naṅguḍi Veḷḷāḷar: Dumont, Louis, Hierarchy and marriage alliance in South Indian kinship (Occasional Papers of the Royal Anthropological Institute of Great Britain and Ireland, no. 12), 1957Google Scholar.

²⁵ Sircar, D. C., Studies in the society and administration of ancient and medieval India, I, Calcutta, 1967, 204–209Google Scholar.

²⁶ Sircar himself notes this case: Successors, 248–9. Kṣatriyas were of course permitted some of the inferior forms, but one cannot adopt such a solution in the case of brahmins.

²⁷ See Gharpure's, J. R. exposition, Sāpiṇḍya or the law of Sapiṇḍa relationship (Collection of Hindu Law Texts, vol. 27), Bombay, 1949Google Scholar.

²⁸ Caste and kinship in Central India: A village and its region, University of California Press, 1966, 184–186Google Scholar.

²⁹ Ibid., 186.

³⁰ Ibid., 202.

³¹ Sircar, Studies, 208. The author raises the same question of gotrāntara in regard to Prabhāvatī Guptā, who refers to her natal gotra rather than that of her husband Rudrasena II in her epigraphs (Select inscriptions, 436, n. 9), but the whole point of the passage is to establish the illustriousness and purity of her descent. Accordingly she gives her patrilineage back to Gupta and refers to her mother's patrilineage when she calls her “crest-jewel of the Nāga family”. What is more natural than reference to her natal gotra in such a context?

The recurrence of the name Nāga at this point in Vākāṭaka epigraphy raises a fascinating problem. Samudra Gupta, as we have seen, claims to have exterminated the rulers of more than one Nāga kingdom, and yet we find his son married to a Nāga princess, as if to cement the amicable conclusion of his father's wars against one of those kingdoms. About the same time Vākāṭaka Gautamīputra married a Bhāraśiva Nāga princess, whose family lent his son Rudrasena important military assistance, as I have inferred. Subsequently Rudrasena's son's son married Prabhāvatī Guptā, who was a Nāga on the distaff side as a result of the marriage referred to above. Was Prabhāvatī's maternal line related to Bhavanāga ? The law of prohibited degrees which neither the Guptas nor the Vākāṭakas are otherwise known to have violated probably rules out too close a relationship, but it is tempting to see in Prabhāvatī's Nāga connexion the renewal of an old alliance between the Vākāṭakas and the Nagas, as well as the start of a new one with the Guptas.

The son of this marriage, Pravarasena II, names his mother and maternal grandfather Deva Gupta (Candra Gupta II), though, fortunately, the term dauhitra is not used. The illustriousness of the mother's family is the obvious motive here.