*THE POPULAR SCIENCE MONTHLY.*

vious he had read to the British Association a paper containing the fundamental positions of his later work, "Theory of Political Economy." In 1865 appeared the treatise "On the Coal Question," dealing with the problem of the exhaustion of the English coal-mines, the calculations of which were adopted by Mr. Gladstone and Mr. Mill in their treatment of the subject. In 1869 appeared the small work, "Substitution of Similars the True Principle of Reasoning;" in 1870 he read a paper before the Royal Society "On the Mechanical Performance of Logical Inference;" and about this time he completed his well-known Logical Machine. In 1870 appeared the first edition of the "Elementary Lessons in Logic." The "Theory of Political Economy" was published in 1871; and the author's great work, "The Principles of Science," was issued in two volumes in 1874.

"The Principles of Science" is a comprehensive treatise on pure and applied logic, or on the formal theory of inference and the methods of scientific investigation. The first book resumes the author's previous researches in pure logic, and carries them a step further. All inference is regarded as essentially reasoning from similars to similars, affirming that what is true of one thing is true of its like. The rules of inference flowing from this general principle, and the symbolical notation employed to express all the forms of thought, are stated and exemplified with great fullness. The particular novelty introduced is the view of induction, which Prof. Jevons regards as merely the inverse process of deduction. Thus, in deduction, we have given to us certain relations among terms or notions, and by the application of the formal laws of thought we develop all the possible combinations which are consistent with given relations. In induction, on the other hand, the combinations of terms are given, and we require to reason backward to the possible relations from which they may result. Insisting strongly upon his view of inductive inference. Prof. Jevons is led to criticise and reject the ordinary accounts of the process. He declines to admit that inductive research necessarily involves the idea of causation, and assimilates it more nearly to the mathematical doctrine of probability. The chapter in which he expounds the philosophy of inductive inference is peculiarly valuable, and deserves more careful criticism than it has yet received. As final result we have the complete subordination of induction to deduction; all inductive research, according to Prof. Jevons, consisting of three steps—framing an hypothesis as to the general law, deductively inferring results from it, and comparing the inferred conclusions with real fact.

The subordinate points involved in this theory of induction, such as the principles of combination and the general method of calculating probabilities, are treated very elaborately. The problem of inverse probability, which is, in Prof. Jevons's view, identical with the problem of induction, receives most careful attention. Some attempts have recently been made to carry out one or two of the elaborate