NEW BOOKS. 595 The first kind is the subject of the "historical," the second of the " natural" sciences. " Physics " deals with the general presuppositions of natural science, " ethics " with those of historical science. Formal logic, being for the most part an empirical science founded on observation of the actual process of (verbal) thinking, is only a propaedeutic of philosophy. To become a "philosophical science" it must be extended into a general doctrine of scientific method. Psychology and anthropology contain both a propaedeutic and an application of philosophy ; but they cannot replace systematic philosophy itself, which is a presupposition for applied, an end for propaedeutic, philosophy. Logic as philosophical science is divided into the "Doctrine of Elements or Principles," or "Analytic of Knowledge," and the " Doctrine of Method," or " Synthetic ". Its " final principle " is the conception of knowledge as a totality or unity ; its " real principle " or starting point is " intuition " ; then, since thought, starting from intuition, knows truth, which yet is not in intuition itself, a third principle is required, viz., the activity of thought in knowledge (Erkenntnisskral't cles Denkens). This determines a threefold division of the " Doctrine of Elements or Principles," treated in section i. of the systematic logic (pp. 60-214). The " Doctrine of Method," which is the subject of section ii., is divided into two parts dealing respectively with " methodic thinking in general" (pp. 215-33) and "the methods of knowledge in the particular" (pp. 234-69). Under this last head conies the theory of inductive and deductive logic as ordinarily treated. Die Principien der Wahrscheinlichkeitsrechnung. Eine| logische Unter- suchung von JOHANNES VON KRIES. Freiburg i. B. : J. C. B. Mohr (Paul Siebeck), 1886. Pp. xii., 298. This is a new attempt to furnish the mathematical theory of probabilities with a logical basis. The author rejects, first of all, the doctrines that would make "subjective probabilities" the basis of calculation; the "degree of actual expectation" being, like intensity of feeling generally, incapable of measurement. That, as logicicians usually say, two events are equally probable when there is no assignable (objective) ground for ex- pecting one rather than the other, is true, but is insufficient to define the characteristic property of numerical probabilities. The principle the author himself proposes as a basis is, " that suppositions stand in a numerically assignable relation of probability when they embrace original fields of play (Spielraume) indifferent and comparable in respect of their magnitude, and that accordingly there result determinate probability-values where the totality of all possibilities can be exhausted by a number of such suppositions" (p. 36). This is described briefly as the principle of "Spielraume". All of the book that is not historical or critical, the greater portion, is taken up with the application of it. Heraklit von Ephesus und Arthur Schopenhauer. Eine historisch-philoso- phische Parallele. Von GOTTLOB MAYER. Heidelberg : C. Winter, 1886. Pp. 47. A comparison of Heraclitus and Schopenhauer as regards the theoretical basis of their philosophy, its pessimistic outcome and its determining factors. The author finds that the phenomenalism and pantheism of both philosophers rendered it impossible for them to attain to any true ethics and made their pessimism inevitable. Lotzes Aesthetik. Von FRITZ KOEGEL, Dr. phil. Gottingen : Vandenhoeck und Ruprecht, 1886. Pp. 138. An exposition of Lotze's aesthetic views. The author intends at present to give only the outlines, and does not deal with questions relating to the
