and endings in o in other cases ; again at Shahbazgarhi, XIII line 8" Amtiyoko Turamaye nama ... Alikasudaro "where out of the five Greek names Turamaya ends in e (against the Girnar Turmayo) while two end in o. In view of the evidence of actual usage from stone all notions of Prakrit grammar to the contrary have to be given up, and it cannot be alleged that the occurrence of a masculine o-ending form side by side with e-ending forms is manifestly impossible.

Acho, Ache, or Acha will, however, denote one and the same name and the occurrence of any will support me equally. It is important that Dr. Barnett takes Acha as a proper name. I fail to see why it should be against my hypothesis if it is conceded (for the sake of agument) that Acha was the real name (and Dravidian if you 1ke) and Aja a sanskritization. How does it affect the question of identification? To call the Puranic Aja "legendary" is to beg the whole question in issue. When the vast majority of the names of Puranic list, from Bimbisara downwards, have been confirmed by inscriptions, coins or independent literature, when names both before and after Aja are proved and have had to be treated as historic al, how can we pick out one and call it legendary?

It would be convenient here to discuss Dr. Barnett's assertion that the Purāņas say nothing about a king called Varta Nandi. Now let us review the whole situation. He admits that there is Nandi Vardhana in the Puranas. There were two royal houses in the time of the Buddha and Maha-Vira with reference to whose regnal years the Buddhists and Jains date the great events in their early ecclesiastical history. These were the Magadha and the Avanti (Ujjain) houses. The Kingdom of Avanti lasted, from Pradyota (a contemporary of the Buddha and Bimbisāra) to Ajaka or Aja1 and Nandi Vardhana, for 138 or 128 years (Pargiter pp. 18, 19; J.B.O.R.S., I, 108.) From Bimbisara up to the end of Udayi 111 years, and that of his successor Nandivardhana the Ajeya, 151 years passed in the Magadha line according to the text of Mr. Pargiter (pp. 68, 69). 1 Pargiter, P.T., p. 19.

1Puranas:

=

This is thus evident that the two Nandivardhanas are undoubtedly contemporaries and that the Avanti dynasty came to an end in the time of Śiśunāka Nandi-Vardhana Ájeya of Magadha and Nandi-Vardhana, son of Aja, of Avanti. The Nandi-Vardhana of the Avanti list, apart from being a contemporary of the Nandi-Vardhana of Magadha, is expressly called a Śiśunāka (Eka-vimsat samā rājyam Ajakasya (V. Br., Vi. Süryakas tu, M.) bhavis hyati Šis'unākah nṛipas trimsat tatsuto Nandivardhanah) in one of the two oldest manuscripts of the Matsya Purana which is dated 1729 (Wilson 21, Bodleian). In view of the reading of the Jones Manuscript of the Vayu, characterized by Mr. Pargiter as "very valuable " " Ajakaḥ sa karishyati” it will be Ajaka who is called a Siśunāka by the Matsya manuscript. In either case, Nandi Vardhana being called the son of the former, if one is a Siśunāka both have to be taken as Fiśunākas. This is confirmed by the readings of other manuscripts and by the Jain records. The latter place after Palaka and 60 years the Nandas of Magadha. Pālaka was the son of Pradyota according to the Purāņas which place one more successor (Viśākhayupa) before Aja; and Palaka and Visakhayūpa have 74 years between them (Pargiter, p. 68). In other words, the Jains count the Magadha rule in Avanti with or in the reign of the Aja of the Purāņas.' It should be remembered that the Purāņas have dealt with the Pradyota family in the Magadha list as a sort of footnote. For a long time they had been lost amongst and mistaken for Magadha kings. I believe it was in the "S'aiśunāka Chronology" published in the first number of this Journal and in Mr. Pargiter's Text that the Avanti list was separated for the first time. In separating them I saw and pointed out the identity of the two Nandivardhanas.

1

Jains:

74 Palaka, 60 years.

Pradoyta

Palaka, 24.

Visakhayupa, 50.

A ja

Nandas of Magadha.

Nandivardhana

Aja is given 21 years in Avanti by the Purapas.

The point is that the S'iśunākas and the Pradyotas are read together. Now the Jones manuscript of the Vayu (called e Vayu by Pargiter) which is a unique document giving a very early version of the Vayu, closes the Avanti kings with Ajaka (Pargiter, p. 18), it does not give Nandivardhana and the total, and it reads immediately the line which is given by Mr. Pargiter as the first line of the next (the S'iśunāka) list: hatva tesham yasah kṛitsnam Sisunāko bhavishyati; "having destroyed completely their glory he will be a S'iśunāka”. It reads this line exactly as I had proposed it to be read as an emendation (J.B.O.R.S., I, 108). Several manuscripts of the Matsya (Pargiter's K) as well omit Nandivardhana and read the line quoted above after Ajaka (spelt as Suryaka). The result is that some manuscripts close the Avanti kingdom with Ajaka, calling him a S'is'unāka and some with his son Nandivardhana calling him a Siśunāka. Then again, the Asiatic Society edition of the Vayu and all other editions of that Purana unanimously call the son of Ajaka of Avanti Varti Vardhana 1. It is thus definite that Varti Vardhana, son of Ajaka or Aja, and Nandi Vardhana, son of Ajaka or Aja, denote one and the same king. At the same time no one would suggest that Varti can be a misreading for Nandi. Varti and Nandi have therefore to be taken as double designations, either one as a Viruda and the other a personal name, or both as personal

names.

Now let us take up the consideration of Vardhana. The Puranas alternatively call Nandi Vardhana, "Nanda Vardhana". The Bhagavata MS. dated 1407 reads Nanda. Mr. Pargiter describes this manuscript as "generally accurate" and " very valuable ". The Puranas giving 100 years collectively to the "Nandas", count from Nandivardhana, like the Jains (J.B.O.R.S., V, 98). The Jain author Hemachandra calls the successor of Udayi" Nanda" only. He does not use

1 Only two manuscripts of unknown dates used by the editor of the Asiatic Society edition and one by that of the Anandasrama edition give different readings Vardhi and Kirti which are manifestly easy misreadings.

Vardhana. Vardhan is, again, used by the Puranic writers (Vishnu) in case of As'oka (Asokavardhana) while we know from inscriptions that Vardhana was no part of his name. It has therefore to be taken as a title in the Puranio writers 1. In view of the "Nanda" and "Nandas" of the Jains and Purāņas and the use of Vardhana with As'oka, I am entitled to treat Vardhand as a title used by the Purāņas to distinguish Nandi from other Nandas. We thus get Nandi and Varti alone as names, personal and Viruda, or alternative.

In face of these facts, in the existence of the indisputable Varti Vardhana and Nandi Vardhana as denoting one and the same king, can it be said that "the Puranas say nothing at all about a king called Varta 2 Nandi" and that " Mr. Jayaswal's effort to identify bis Vața Nandi with Puranic NandiVarshana must be pronounced a failure"? We find Simuka (or its misreadings) in the Purānas but not the other name Satavahana, while below his statue we have "Simuka Satavahano" (Bühler, A.S.W.I., 4.) Does any one doubt the identification of Simuka Satavahana with the first king of the Puranic Andhras? In the Puraṇas we have only As'oka. in the Ceylonese chronicle" Priyadarsana," and in inscriptions discovered up to this time, Asoka or Priyadarsin. Will the identification be challenged if we found in future As'oka Priyadarsi together? The Oriss MSS. have only either the Viruda or personal name of kings and now in inscriptions both are found together. Are the identifications of those kings to be doubted?

For my thesis it would have been enough to find Aja or Ajaka and his son Nandi- Vardhana even in the Avanti list with the express mention about one of them to be a S'is'unāka. In addition to that we have the identification confirmed by the S'is'unāka list. Nandi Vardhana in the Magadha list is the successor of

1

Compare also the use of Vardhana joined on to Harsha.

2 Vața may equally represent Varta or Varti.

3 Also Dr. Barnett's as he admits that it may be read without violence as Vata Nandi.

Udayi according to all the Puranas except the Bhagavata. The latter gives in place of Udayi "Ajaya" and calls Nandivardhana "Ajeya". To any one who knows Sanskirt Grammar it is evident that a patronym Ajeya can only be formed from ja and therefore the preceding form “Ajaya" is to be regarded as corrupt. This was clear to Dr. Barnett and he has not questioned the Puranic existence of Aja alleged by me. This seems to have been clear to Mr. Pargiter who in giving the reading Ajaya said, "but 1 see note 38 ", and note 38 runs: "Bh (the Bhāgavata ) gives him (Nandivardhana) the patronymic Aieya. It has been however questioned in the pages of the Indian Antiquary (1919, page 35) and as the objection has the tacit support cf Mr. Bhandarkar, the Sar skrit knowing editor, I may be permitted a short digression to deal with it. It has been boldly asserted there that Aja does not exist in the Sifunāka list" as one may satisfy himself by looking at Pargiter's Purāna Text, pages 20-22". The objector derives "Ajeya" from "Ajaya via" ajeya", for both in his opinion mean "invincible ". But a derivative suffix is not attached to the meaning but to the word itself. Similarly it is elementary enough that taddhita suffixes are added to a noun and not to a Viseshana as Ajeyn is (verbal adjective). The form Ajeya here in the S'ais'unāka and "Aja and Ajaka in the Avanti list prove beyond controversy that the other name of Udayi is to be read as Aja orjaka in place of the reading "Ajaya”.

now.

The question of readings and palæography may be discussed Dr. Barnett reads the Aja inscription as Bhage Acha chhanivike, against my reading Bhage Acho chhonidise. The material difference is about the last word, and there too it is narrowed down to the last two letters and the vowel mark in the first one. In the second inscription he agrees with my reading Vata Nandi and agrees with me in treating it as a proper name. Acha also, as I have pointed out already, he regards as a proper name. I have thus the

1 Italicization is by me.