To the number of extant native chronicles the Mahārāja's library contains an important accession in the shape of a small palm-leaf MS. of a Vamśāvalī discovered by me shortly before I left Nepal. Since my return to England owing to the kind negotiation of Col. Loch, I have not only received an excellent copy of the MS., but also the favour from H.H. the Mahārāja of the loan for three months of the original, so that I have been able to collate and photograph all important passages.

My use of the chronicles has been adversely criticized 2 by some scholars, though countenanced by others; 8 but I venture to think that the discovery of the present MS. puts matters in a somewhat new light.

Though written continuously in a single handwriting corresponding with the time (reign of Jayasthiti-malla, A.D. 1380-1394) at which the chronicle ends 4 (see the Plate annexed, figs. 3-10), the new Vamśāvali really contains three distinct chronicles, designated accordingly in the present essay V1, V2, V3.

VI is in the form of brief annals of the successive reigns not unlike the other Vamśāvalis, but giving a much greater number of dates, in addition to the lengths of the reigns. The leading events of each reign are also noticed in some cases with dates, at first in words and later on in numerals.

The leaf-numbering begins at f. 17 and this portion ends with 308. The language is no doubt intended for Sanskrit, but in obscurity and a perfectly wild absence of syntax it rivals the worst colophons of Nepalese MSS. that I have seen. I thought at first of printing the whole, but after studying my transcript and taking the advice of friends I came to the conclusion that I should either have to print the whole without spaces, which would be misleading and unsatisfactory, or to publish facsimiles. For the division of words and even sentences, when one had no fixed rules of grammar to help in the interpretation, seemed in many places quite doubtful. I have been consequently permitted by the Council of the Society to take the latter alternative, and have accordingly reproduced a selection of the most important leaves, of which I made legible negatives while the MS. was lent to me. It will be noted that the selected leaves begin with 34. I only publish now a portion of this leaf, as I reserve for future treatment the kings of Nepal before 879 A.D. in connection with my recently-discovered inscriptions.

From the prominent way in which temple-donations are recorded, it may be conjectured that this part of the MS. (VI) may be in some way connected with the records of the great shrine of Pasupati.

V2 is a document of different origin; it is a list of births of royal and other distinguished personages. The language is unfortunately old Newari; but one can make out the names and dates clearly enough. These extend, not always in strict chronological order, from N.S. 177 to 396. A specimen is given in the Plate, fig. 10. All the information given in this section of the Chronicle, so far as it relates to the kings, will be found condensed in the notes to the Table of kings below, Towards the end of the section other information beside births, deaths, and the like is introduced, but I have been able to make but little use of this owing to the difficulties of the language, for which I can get no adequate help either in Nepal or in Europe.

V3 is perhaps merely a continuation of Vo. I have called it a separate document, because a slight break with double dandas occurs in the original MS. at the end of fol. 36a, and because at this point there is a marked difference of style. The string of short paragraphs, each recording little more than a birth, is abandoned, and the annals become more expanded. The previous section had ended, as I said, with N.S. 396. This, however, begins with N.S. 379 and the history would not seem to be treated on a strictly chronological basis, as the irregularity of order in the dates noticed in Vs is here more pronounced. The latest date, as already noted, is N.S. 508 (f. 63b).

The events mentioned in V1 are sometimes described here in fuller detail. On the other hand, the chronological details though full seem not to be quite so trustworthy.

Nevertheless I feel sure that this section must contain much valuable information, and it is in the hope of drawing the attention of the few scholars skilled in the Himalayan languages to the matter that I reproduce a specimen-leaf (Plate, fig. 10). The passage refers to the invasion of Harisimha of Simraon about which I have more to say below.

Having thus indicated the materials of the present investigation, the divisions of the subject may be stated.

I.-The History of the Nepal Valley, A.D. 1000-1600 (i.e., Kathmandu, Patan, and Bhatgāon):

Chronological notes on the dynasties of the surrounding

states:

II.-Western Nepal. III. - Tirhut (Eastern and Western). To these notes I have added (as Table IV) a list of a dynasty, which I have not been able to recognize.

The main results of the enquiry are summarized in the Tables at the end of this article, which constitute of course its most important feature, and will probably provide most students with all that they require. The present notes are chiefly intended to elucidate the Tables and especially to bring out the relations between the dated series of kings obtained from MSS. and the dynasties detailed in the new Vamśāvali.

I. The present collection of MSS. affords an example [See Plate; fig. 2, 1. 2] of a datel earlier than any hitherto found referable to the Nepal era, but unfortunately no dated MS. with a king's name occurs earlier than those previously known.

It is interesting to note that the king Raghavadeva mentioned by Cunningham as the traditional founder of the Nepal era of 879-80, but passed over in the Vamśāvali of D. Wrights and by Kirkpatrick,4 is duly recorded in the new chronicle. Not only so, but the years of reign assigned to him and his immediate successors quite accord with the tradition of his having founded the era. Thus, if we add together the duration of his reign and his five successors down to Laksmikāmadeva we get about 135 years. This, again, added to 879-80 brings us to the second decade of the eleventh century, when we know from a colophon that Laksmikāma had commenced to rule at all events as jointsovereign, becoming sole king later on.

The earliest king of Nepal mentioned with a date in the Catalogue is Bhāskaradeva; and it is very satisfactory to find that this date already noticed by Pandit Haraprasād (J.A.S.B. for 1897, Pt. I, p. 312) is verifiable. Prof. Kielhorn has kindly calculated it for me and it corresponds to 24th September, 1046. The new chronicle duly records this king with a rather obscure note 1 as to his 'repairing his paternal crown.' The other chronicles make him the founder of a new dynasty. Of the next king, Baladeva (called in Vl Balavantadeva), we have a dated MS.8

Of Harsadeva's reign we have now two MSS.4 A third date has been added from the Chronicle, which says of this reign merely: राजा श्रीहर्षदेव वर्ष १४ ऊनविंशति सम्वच्छर समय परिवर्त्तमानं ॥ Interpreted in the light of the two other dates this rather crudely expressed notice gives good sense, if we take it to mean that Harsa died in N.S. 219 current. This fits also quite well with the duration of the next reigns as given in our Chronicle. The credibility of the dates in this part of the chronicle is further enhanced by its mention of the completion in 239 (date in words) of a tank by Sivadeva, the next sovereign.

This was called after the Yuvaraj (जौराज) Indra or Mahendra-deva, Mahendra-saras. It was otherwise known as Madanasaras. It will be seen that this date falls in the year before the writing of a MS. in the same reign.

I have lingered over these somewhat minute details for two reasons: (1) Because it forms a new feature of the present chronicle to find so early as this dates expressed both in words and figures that accord with the contemporary evidence of the scribes; (2) because doubts has been expressed whether the Nepal Samvat (of 879-80) was actually in use in the eleventh century A.D. It is satisfactory to note that our chronicle, following the tradition already known from Kirkpatrick, does mention the aforesaid Indradeva both as yuvarāja and rāja, as we have now a MS. of his reign. It will be seen, however, that the number of years (12) assigned to his reign is probably excessive. The dates of the next two reigns overlap one anther. If this is not a case of subdivision of the kingdom of which there are so many instan-ces, it may be quite well explained by the tradition preserved in the records of Wright and Bhagvanlal, that Mahādeva retired early in his reign from active sovereignty and Narendra (or Narasimha) became his regent. Of the reign of the next king, Ananda, MSS. are now numerous. It is curious that the other chronicles either give his name wrongly (Wright, Bh ) or omit his reign altogether (Kirkpatrick). It is found, however, correctly spelt at f. 316. of our MS. chronicle. Of Rudra's reign no MSS. are extant. The years assigned by our MS. to his reign (8) seem to show the origin of the curious error in the length of the reign (80 years) assigned by Kirkpatrick. Equally correctly given is the form Amrta, which is now verified by a MS.7 The 'great dearth' re

1. 4.

1 The event is again chronicled with the same date in V. See Plate, fig. 9,

6 Kirkpatrick's dates begin later (thirteenth century). Those in Wright only become correct somewhat later (invasion of Harisimha).

8 A. Foucher, Iconographie Bouddhique, p. 28, n. 1

• The records preserved by Wright and Bhagavanlal (Ind. A. XIV. 413) pass this king over.

5 The actually discrepant date is that supplied by the colophon in the Cat., p. 62. Here there can be no doubt as to the interpretation of the chronogram, through some of the terminations of the other words must be corrected for the scansion. But the date seems not to work out. The obscure phrase rājārājasādṛśye may quite possibly refer to regency.

6 The common mistake 'Nanda-'deva is found at f. 25a (plate, fig. 5, 1.1.). Rājendralāl Mitra makes the same blunder in his text of Astas. Pr. pref., p. XXIV. note.

7 Cat. p. 65 (ম). I find from a tracing sent from Nepal by Col. Loch that the Pandit's reading of the year (296) is quite correct, and in that in the next line Sri Amrtadevasya is quite clear.