September 9, 1072 (India Office Library) 'Preliminary Sur- vey.. of the Statistical Materials in the India Office Library and India Office Records', which, though 'selective', according to the author, "is the first attempt to survey the statistical resources'' of these two havens of India-lovers over the world. As for the reliability, "at the moment it appears that insufficient use has made of these statistics to permit any general statements to be made". But studies conducted on the basis of trade, population, and sociological statistics have already indicated the nature of corrections necessary. On population, in particular, recent studies have indicated that "pre' 1871 estimates and the post-1871 full censuses are all subject to various errors". Among the many sources, still largely unused, according to Bingle are: (1) 10,000 volumes of European manuscripts consisting mostly of reports of civil servants and army officers, covering the period 1750-1950; (2) the papers of the Surveyer General of India (Mackenzie Collections: 1753- 1821) relating to the history of South- India and South East Asia, mostly relating to agriculture and trade; (3) Buchanan- Hamilton collections (1762-1821) consisting of the survey of seven districts of Bengal and Bihar; (4) Papers of Robert Clive and members of his family (1718-1831) and those of Warre Ma- let (1752-1815), and John Gross (I7tt3- 1771); these are important sources of private trade and "the way a 'nabob' amassed a fortune"; (5) Commercial Reports from India and the records of the Accountant General's department. Administrative reports, annual departmental reports, and ad hoc reports are among the three groups of reports that have not been much used for statistical purposes.