NEW BOOKS. 285 symbolising the infinite, afterwards passing into Natiire-worship, is supported neither by inductive (archaeological) nor by deductive (psycho- logical) evidence, but is a foregone conclusion, theologically, and not scientifically, determined. Appendix C, on " The Linguistic Method of the Mythologists " (pp. 830-837), is new. Mr. Spencer here contends that the assumption of philologists " that there exists in all cases, or in nearly all cases, a rational root for a word" is rendered inadmissible by the knowledge that " at present there goes on what may be called by contrast an irrational genesis of words ". In the absence of historical evidence of their origin in particular facts, entirely wrong etymologies (as is shown by illustrations) would seem plausible for words now in use. Similar forma- tions cannot have been absent in primitive times ; and there is no criterion by which one set of formations can be distinguished from the other. Those who advocate the mythological theory, for example, Prof. Max Miiller, also set out with the postulate " that there were originally certain roots super- naturally given," and would, further, maintain " that mankind lost their original ability to frame abstract ideas and use the corresponding abstract words"; but this implies that the movement of thought is from abstract to concrete, or the reverse of that which it actually is. In the present edition, the means of verifying the statements contained in each paragraph (at first omitted for the sake of avoiding foot-notes) is provided in a section of " References " (pp. 839-68) compiled according to the method that has been followed in all the later Parts (separately published) of the Principles of Sociology. While references have thus been supplied (with no ordinary labour, for .which the author has to thank Mr. H. R. Tedder and the late Mi'. P. R. Smith), slight errors have also been removed in the quotations originally taken from Descriptive Sociology, " which, though not diminishing the value of the extracts as pieces of evidence, rendered them inexact ". Lastly, a Subject-Index (pp. 869-83, due to Mr. F. H. Collins) has been added, greatly facilitating the use of a work so full of matter. Lectures on Philosophy. First Series. By THOMAS MAGUIRE, Professor of Moral Philosophy ; Fellow and Tutor, Trinity College, Dublin. London : Kegan Paul, Trench, 1885. Pp. 263. These racy Lectures " all grounded on the fact familiar to anyone that understands Plato or Hegel, that all knowledge involves two opposite ele- ments, never separate and always distinct" were, we believe, published last autumn, but have only now come to hand. They might have come sooner, occupied as some of them are so directly with the work of writers in MIND. When it it said that " they are intended for students in Logics and Ethics in Trinity College," the meaning appears to be that they have been delivered to that auditory, but under what circumstances and with what continuity (or discontinuity) is less apparent. The first of them, on " Some Facts of Perception and their Significance," reads like the inaugural lecture which it will have fallen to Prof. Maguire to deliver when he assumed his chair some years ago ; but a whole year seems to have passed before it was followed by the next, on " The Will : in reference to Dr. Maudsley's Body and Will". The others, on "Materialism," "Ethics founded on End " (Aristotle, Butler), " Transition from Ancient to Modern Philosophy " (Schoolmen, Descartes, Leibniz), " Kant " may have followed at any time in the interval till last year, when Prof. Webb and a number of .writers in this Journal or other periodicals gave identifiable occasion for the remaining five lectures (one of these, however, being no more than a note of three pages on Mill's Nameable Things, designed to show that the Daily Telegraph was not justified in saying, at some time or other, that Mill was a man whom Plato would have called o-o$6r). It is plain that the
