Poole, R. B. Fires, Protection, Insurance. U.S. Education Report, 1892-93, vol. i. p. 724. For articles, see Cannons, D 46, Fire Prevention, Insurance. =54. Accounts:= Brown, J. D. Manual of Library Economy, Ed. 1, 1903, p. 30. Hopper, F. F. Order and Accession Department. _In_ A.L.A., Man. of Lib. Econ. Preprint of chapter xvii., 1911. For articles, see Cannons, E 6-7, Accounts. CHAPTER IV THE MATERIAL OF THE ANNUAL REPORT. STATISTICS, WORKING RECORDS. FORM OF THE ANNUAL REPORT =55. Statistical Methods.=--It seems desirable to describe here, as being concerned with committee work, the various statistical and other methods adopted to show the operations of the library. Every business concern of any consequence has what approximates to a statistical department, in which records are kept and analysed of every transaction of the business, and from these useful deductions are made. Such statistics, with the necessary proviso that the keeping of them should not impede more obvious work, are desirable for libraries, and although they are necessarily of a quantitative rather than qualitative character, they are nevertheless of value as showing the use made of the several departments and of the various classes of the stock. Statistics usually kept include: the stock; accessions; the issue of books and material in each department according to their classes; the number of readers’ tickets in force; and the attendances at the reading rooms. Sometimes records are kept of the occupations and ages of readers, and the wards from which they are drawn. Hitherto all these statistics have been of doubtful comparative value, owing to the great divergence in the methods of computation adopted, and the methods of administering the library have a definite effect upon the resultant figures. For example, in the few libraries where so brief a time as a week is allowed for the reading of a volume, the issue figures will be higher than in those where a fortnight or more is allowed. Then, the amounts charged as fines for undue detention of books, ranging from 1d. a week or part of a week to 1d. a day beyond the time allowed, make a difference of some moment. Again, in some libraries it is usual to record the number of borrowers continuously, only counting off the lapsed tickets, while in others only those borrowers who are actually using the libraries within the year are counted. It would be well if an effort were made to standardize all these methods. A scheme for standardizing the form in which they are to be presented has been reached, and is described below (Section 62).