trinsically significant, and far more influential in shaping Leibniz's thought as a philosopher, mathematician, and theologian than has been hitherto realized; and he has given us a most full and able account of what these labors were in all their aspects. While greatly impressed with the strength of the evidence which the book also presents in support of the conclusion that Leibniz's logic was the real foundation and center of his whole system, the present reviewer is not prepared, without further investigation and an examination of the logical writings of Leibniz, published by M. Couturat since the appearance of this book, to regard that conclusion as fully established. In the case of a universal genius like Leibniz, it is easy to make out a strong case for this or that interest as being primary and central; for example, it would not be difficult to make out a plausible case (cf. p. 165, n. 2) in support of the view that Leibniz was primarily a theologian and that his logical and his mathematical labors alike were but auxiliary to his theology. This is said not to disparage the evidence adduced by M. Couturat in support of his revolutionary conclusion, but to lead to its being soberly weighed.
George Martin Duncan.
Yale University.
The book has several claims to consideration: first, it is a fairly comprehensive review of the literature and status of comparative psychology. Mr. Hobhouse has made use of many authorities even the most recent. Yet there are very singular omissions: for example, it is singular that James's name should not occur in a book which discusses instinct, habit, the limits and methods of acquisition, etc. Second, it contains new experimental and critical matter. Mr. Hobhouse has endeavored to conduct experiments on animals (including monkeys) under conditions which more nearly fulfill the normal ones than many heretofore carried out (e.g., Thorndike's). In this he is fairly successful, and his effective criticism of Thorndike, for example, is supported by experimental results. At the same time, the 'naturalness' of the conditions is certain to be criticised by the extreme advocates of exactness. The present reviewer's opinion, however, is on the side of Hobhouse and Mills—that artificiality in such experiments is the extreme to be avoided. Better 'natural history' observation in such a complex thing as an animal's learning processes, than an artificial exactness which paralyzes the learning process or ren-
