The librarian must develop a sense of proportion and beware of the library patron with a hobby, and of the trustee who is interested in building up only one side of the library, and of the scholar who thinks that “only solid reading for the immortal mind should be placed before old and young.” In buying for an average public library the aim should be to choose general treatises rather than those covering only special phases or special subjects. The special treatises would be the more desirable for a university library, where they would be in demand both by faculty and students, as authorities on detailed points, as aids or as sources in further investigation. Too much money should not be locked up in expensive volumes that will be seldom used. The librarian should estimate the average cost of his books per volume and, except in the case of reference books, should not go too far beyond this average cost. He should avoid partisanship and develop catholicity of taste and breadth of sympathy. He should try to have something on his shelves for every patron in town, real or potential, but should not allow the library to be drawn into any sectarian propagandist movement. He should avoid controversial works, sensationalism and the latest fad, and put off the purchase of the book of the hour until he feels fairly sure that the demand for it will not die within the hour. The efficient librarian does not think too much of the sum total of accessions, but is mindful of the fact that it is quality not quantity that counts. “It doesn’t matter how many but how good books you have.” This was said by Seneca, but the same truth has been stated by many modern librarians. “I should as soon tell how many tons the books in the Astor Library weigh as to tell how many volumes there are,” was a sage remark of Dr. Joseph G. Cogswell. “Strength does not lie in mere numbers; this fact is as true of books as of soldiers,” said Mr. W. E. Foster. “One thousand carefully picked are worth two thousand assembled at random.”