now appears very significant. For instance, on the grounds that his informant was "a stupid fellow, and no other person has heard of such a tradition", he did not think it worth while to refer in his Report to the names "Hangsa Nagar" and "Hangsapur" mentioned to him in connection with Giriak. Fifty years later, the remains which still exist on the hill above Giriak were conclusively identified by Cunningham with the Goose Stupa and Monastery described by Hiuen Tsang. Similarly, the jungle-covered valley of Old Rajagriha seemed to him obviously so unfitted for the site of a city, being "surrounded on every side by arid rocks, which would render the heat intolerable" and the situation "to the last degree insalubrious ", that he did not trouble to investigate for himself the truth of the local belief that it was the site of the old city of Jarasandha. There can be little doubt that Old Rajagriha was actually proved to be an unpleasant dwelling-place, partly owing to the reasons mentioned by Buchanan and perhaps still more to lack of water at the hottest season of the year and that its abandonment and the establishment of New Rajagriha outside the hills were due to the comparative advantages of the latter site, rather than to the legendary reason as related by Hiuen Tsang. But that Old Rajagriha was at one time inhabited by a large population is a fact which cannot but be evident to anyone who examines the site even now ;* and the Journal shows that the reasons why Buchanan never noticed even so much as its massive walls were, first, that he did not ascend either Baibhargiri or Vipulagiri sufficiently far to get a proper view of the valley within the hills, and second, that when he did enter the valley in order to examine the Sonbhandar Cave, his path went past a part of the old city where its wall has been almost completely cut away by the western branch of the Saraswati stream.

The Journals of South Bihar show that during his tour in each district Buchanan kept up the practice *Notes on Old Rajagriha, A.S.R., 1913-14, pages 265-271.

adopted during the Mysore survey, which was, in the words of Sir D. Prain:

"To make a stated daily march, and in the morning before leaving camp to gather round him the leading people of the neighbourhood whom he questioned on the various points enumerated in his instructions. During his march and at the places where he halted, his own observations were carefully noted, and extensive botanical and geological collections were made."

Much of the material recorded in the Journal of Mysore, especially the observations on agriculture and botany, finds no corresponding place in these Journals, but has been incorporated direct into the Reports. There is one feature, however, which shows a marked development in this later series of Journals. This is the care with which the distance is estimated between each successive village, river or other notable feature of the country passed over during each day's march. Buchanan had discovered that the existing maps of the districts included in the Bengal Survey were all more or less. unreliable, and the details of distance which he sets down in his Journals were evidently intended for use in preparing the revised map of each district which he himself drew. This task, which he set himself for it formed no part of his instructions, must have involved much labour, especially as no trained surveyor was attached to his party; and practically the whole of the work which he carried out in this manner-both as an independent geographer and as the direct successor of Rennell--has escaped notice hitherto, owing to the fact that his manuscript maps, still preserved in the Map Department of the India Office Library, have never been published except in a very incomplete and unsatisfactory form. For this reason, a detailed account of those which relate to South Bihar may appropriately be set down here.

All internal evidence points to the conclusion that Buchanan had no maps of Bihar to consult during his Survey other than those contained in the second edition

of Rennell's Bengal Atlas, published in 1781. The first edition of this Atlas, which was published in 1779-80, contained only Plates I to XII, but there are references in the Bhagalpur Journal (January 2nd, 1811) to Rennell's plans of Mir Kasim's fortifications at Udhua Nullah near Rajmahal, and in the Patna Report to his plan of Patna City, which are included only in the later and more complete edition, as Plates XXI and XV respectively.

There is no indication that Buchanan knew anything about the series of larger maps (on the scale of five British miles to the inch) drawn by Rennell in 1773 and published quite recently by Major F. C. Hirst, Director of Surveys, Bengal, from the originals in the India Office collection. As Hirst points out in his accompanying Memoir, these maps must have been used by Rennell in preparing the Atlas, but they differ considerably from the latter in certain important details which Buchanan would undoubtedly have noticed, if he had had access to them.

So far at least as the Districts of Monghyr, Bhagalpur and the Santal Parganas are concerned, the existence of any maps later than those in the Bengal Atlas is conclusively disproved in the first paragraph of the following passage, which Martin omitted from page 2 of the Bhagalpur Report as published in Eastern India, Volume II:

"The turbulent state of the inhabitants, and the difficulty of access into the country, when Major Rennell made his survey, opposed obstacles which have rendered his map of this district less valuable than mest part of his excellent work, and I have to regret that a copy of a more recent survey, which had been deposited in the office of the Collector, has been lost. Owing to these circumstances, to the very uncommon manner in which many of the subordinate jurisdictions bave been intermixed, not only with each other, but with other districts, and to the disputed and undetermined nature of the The Surveys of Bengal, by Major James Rennell, 7.R.S. 1764-77, Calcutta, 1917.

boundaries, I have not been able to trace these in a manner that can be at all satisfactory, and in almost every case I have been under the necessity of proceeding by conjecture, and that in many cases of a very vague

nature.

"In calculating the proportion of various kinds of land and the extent of cultivation, I found the proprietors so evidently departing from the truth that I have in great measure been obliged to trust entirely to what I and my native assistants could actually observe, on which account we traversed the district in many directions and with much pains. Notwithstanding this labour, I am much less satisfied with the result than with those which I procured in the districts formerly surveyed. The conduct of my enquiries owing to this circumstance has in this district been rather disagreeable. The managers of the estates showed much alarm, a want of veracity that could only be equalled by their total indifference about it being discovered, and a degree, of intellect vastly inferior to the people of Bengal. No general statement could be. procured from the most intelligent, and the details which they gave were in such diametrical opposition according to the nature of the questions proposed that no reliance whatever could be placed on their assertions. If for instance a man was asked, why so much land was waste, he would assert that seven eighths of his estate were cultivated; but in explaining the heaviness of his burthens he would show an account in which, with an extent of ten or twelve miles square, he had not above one or two thousand biglas in cultivation. "

In addition to the passage just quoted, there is ample evidence throughout the Journals and Reports that Buchanan was by no means satisfied with the accuracy of Rennell's maps. Though he himself does not say so, it is interesting to note that the plan which he formed was to prepare a revised edition of the maps in the Bengal Atlas itself, so far as they related to his Survey. His own maps have been drawn on exactly the same scale as the somewhat inconvenient one employed by Rennell in Plates I to VIII of the Atlas, which are common to both editions. This scale was one of ten geographical or nautical miles to an inch, Rennell's marked preference for nautical units being explained

by Hirst in the memoir already cited as being due to his training as a Marine Surveyor. Rennell assumed that a nautical mile was 6,090 feet (6,080 feet is more correct), so that this scale should correspond to 11.53 statute or British miles to an inch. In the copies of his maps which I possess, the real scales are slightly smaller than this, owing doubtless to a certain amount of shrinkage in mounting, and the actual values are approximately 11.8 and 11.6 miles to an inch in Plates II and III3 respectively. Small though this difference may appear, it is not altogether negligible, as it corresponds to a discrepancy of about three miles between the extreme limits of the map which shows the boundaries of Monghyr, Bhagalpur, and the Santal Parganas.

Altogether, there are eight of these manuscript maps in the India Office Library, and very careful tracings of the three which refer to South Bihar have been made for me by Miss Anstey. The description of these is as follows:

(A) "M. S. maps

of Districts by Dr. Buchanan Hamilton. Drawn in about 1814. No. 1. Bhagalpur. Size 11 inches by 14.

Title on Map, in Buchanan's handwriting, "Bhagal pur". No scale of miles is drawn on this map, but comparing selected points on the tracing with Rennell's map, the scale is 11-67 miles to one inch.

(B) "Map of Zila Behar, including the City of Patna, drawn by Dr. Buchanan Hamilton about 1814. Scale about 11 miles to 1 inch; size 8 inches by 11."

Title on Map, in Buchanan's handwriting, as given above. Scale of miles drawn on map. On the tracing, 60 miles 511 inches, or 11.74 miles to one inch.

() "The Jungleterry District and the adjacent Provinces of Birboomi, Rajemal, Boglipour, etc., comprehending the Countries situated between Moorshedabad and Bahar.

(2) "A Map of South Bahar, including the course of the Ganges to Chuuargur."