,g^. SOME NEW BOOKS. 545 



internally by gill arches, or the blood vessels which convey the 

 blood to be oxygenised in the branchial apparatus. It is really a pity 

 that the author was not more careful to prevent misconception on the 

 part of the general readers for whom his book is intended. Mr. 

 Romanes claims to belong to a class of naturalists whom he calls 

 speciaHsts in Darwinism, and says that the opinion of those who have 

 done good work in other departments may be destitute of value in 

 questions of evolution. However this may be, it is certain that the 

 specialist in Darwinism who undertakes to expound the embryo- 

 logical argument for evolution should not rely almost exclusively on 

 the writings of Haeckel. He would find Balfour's " Comparative 

 Embryology " a safer guide. Mr. Romanes would have been wise to 

 submit this unfortunate chapter to the revision, if not of a specialist 

 in embryology, at any rate of some zoologist with a competent know- 

 ledge of that department, for instance to that of his friend Professor 

 Lankester, whom he quotes with great admiration. 



The second more truly Darwinian section of the book is much 

 more satisfactory than the first. The theory of Natural Selection is 

 clearly and fairly stated. It is pointed out that in its main elements 

 the theory is merely a statement of observable facts, the fact of the 

 excessive rate of reproduction leading to the constant pressure of 

 numbers in each species, the fact of competition, or struggle for exis- 

 tence and consequent survival, the facts of hereditary transmission 

 and individual variation. We welcome the candid admission that 

 the theory only explains changes in organisms so far as these changes 

 are of use ; in other words, that it is a theory of the origin of adapta- 

 tions, not of species. This admission occurs in an interesting dis- 

 cussion of misconceptions of the Darwinian theory which are current, 

 not among its opponents, but among its supporters. Another of 

 these misconceptions is that the theory can explain all cases of 

 modification, whereas in some cases it is not logically possible that it 

 can apply ; others are that it follows deductively from the theory 

 itself that Natural Selection must be the sole means of adaptive 

 modification, and that all hereditary characters are necessarily due 

 to Natural Selection. 



In the chapter on the evidences of the Theory of Natural Selec- 

 tion we have three general or main arguments : — The observed fact 

 of the extermination of forms in the struggle for existence ; the con- 

 sideration that we cannot find an instance of a structure or instinct 

 developed for the exclusive benefit of another species ; and, thirdly, 

 the facts of domestication. This last class of facts is illustrated by 

 fourteen pages of figures drawn from actual specimens of domestic 

 breeds, illustrations which form the most novel and distinctive 

 feature of the book. These figures, although not very beautiful, are 

 certainly of great practical use, and enable one to appreciate the 

 peculiarities of the several races better than the most elaborate 

 description without figures. As far as we can judge, they are not 

 only vigorous, but accurate. 



Then follows a discussion of some of the detailed applications of 

 the theory to the explanation of adaptations, the author having 

 decided to select all his instances from a single class, namely, that 

 which may be generically termed defensive colouring. This choice 

 is, it seems to me, extremely unfortunate at a time when so many 

 popular treatises have just been produced on the same particular 

 subject. As we have so recently had to study Mr. Poulton's discus- 

 sion of the subject on one side, and Mr. Beddard's on the other, it 

 would have been a relief to find in Mr. Romanes' book the exposition 



2N 



