that is treated in a more concrete and 'physical' manner in the rest of the work. In regard to the relation between ethics and metaphysics, there are three alternative possibilities to be considered: (1) Ethics may be a mere derivative and off-shoot of metaphysics, consisting solely in the systematic application of metaphysical first principles to the subject-matter furnished by the facts of human conduct. (2) Or this relation may be reversed and ethics be regarded as the primary science, while metaphysics is employed with "the task of ascertaining what general conclusions about the nature of the universe can be drawn from the data supplied by ethics." (3) Or it may be that neither is derived from the other, that each is independent in subject matter and mode of treatment, although the two spheres may come into contact at special points (pp. 2, 3). After an enumeration of those who, with more or less qualification, may be regarded as exponents of each of these views, the author announces his own allegiance to the third position mentioned, maintaining that "ethics should be regarded as a purely 'positive' or 'experimental,' and not as a 'speculative' science." As the establishment of this view is the concern of all the rest of the treatise, and as the procedure is rather tortuous, it is of course impossible to follow the details in a mere review of the work. It may be sufficient simply to cite the author's own introductory statement of the threefold character of the argument. "We shall first of all offer some reflections of a general kind upon the points in which a science founded upon metaphysics ought to differ from one that is purely positive and experimental." "Next we shall try to meet and answer some of the reasoning by which the metaphysical moralists have sought to show that there can be no satisfactory theory of conduct apart from a metaphysical foundation. We shall then go on, in the main body of our essay, to show the impossibility of basing ethics upon a previous system of metaphysics, by a detailed examination of some of the principal facts of which ethics has to take account" (p. 5).
Probably the vast majority of ethical thinkers at the present time would unequivocally agree in maintaining that ethics is not to be viewed as an entirely derivative science, whose principles are to be a priori deduced from some formal metaphysical principle. Yet such a concession would not apparently enforce the conclusion that ethics is merely an empirical science, bearing precisely the same relations to metaphysics as physics or any other natural science does. Even granting Mr. Taylor's contention that "all knowledge is 'empirical,' in the sense of being concerned in the last resort with the description of matters of fact or experience," and without accentuating the distinction between scientific
