170 NATURAL SCIENCE. March, 



take illustrative facts from various provinces of the animal kingdom 

 without reflecting that these facts may be all of one class, and that 

 there may be facts of quite a different class which they ignore. The 

 question whether the principles of the theorists are in harmony with 

 the details of the empiricists has yet to be thoroughly examined. 



The most recent important contribution to the study of evolution 

 in this country is Bateson's " Materials for the Study of Variation, 

 treated with especial regard to Discontinuity in the Origin of Species." 

 In this work, as in its predecessors, the method is to search the animal 

 kingdom for instances of one class of facts, instead of taking a portion 

 of the animal kingdom and ascertaining how many classes of facts it 

 presents. Valuable as the work is, and important as are the aspects 

 of variation to which the author draws attention, one cannot help 

 being astonished at the fact that Mr. Bateson writes as though, apart 

 from adaptation to the environment, no explanation of the discontinuity 

 of specific forms had been offered before he himself took up the pro- 

 blem. He does not mention that, so long ago as 1886, Romanes main- 

 tained that Natural Selection was a theory of the origin of adaptations, 

 by no means a theory of the origin of species ; or that two years later 

 Gulick published most important evidence of the origin of varieties by 

 mere divergent variation through isolation, without any adaptation at 

 all. In fact the modes in which variations, instead of being individual 

 peculiarities within a species or other taxonomic group, are to become 

 constant and characteristic of a distinct group, are never discussed by 

 Bateson. Except in one or two rare instances, he entirely omits to 

 consider the relation between the variations, whose description makes 

 up the body of his book, and the taxonomic characters of the forms in 

 which they occur. There is some logic in his contention that, since 

 environments are often continuous, their influence cannot always 

 explain the discontinuity of species. But it is clear that, if progressive 

 modification in different directions goes on in two groups of a single 

 species which are isolated so that no interbreeding takes place, then 

 two species will be formed, the discontinuity between which will 

 become greater at every generation. This result will follow, however 

 gradual and continuous may be the modification in each group. Mr. 

 Bateson's argument is defective, therefore, in two respects : first, 

 there is no necessity for discontinuity of variation to explain discon- 

 tinuity of species ; secondly, he has not attempted to show where 

 discontinuity of modification probably did occur historically in the 

 evolution of any particular group. 



In reference to the flat-fishes (Pleuronectidse), Bateson (p. 466) 

 discusses only one class of variations, namely, the partial or complete 

 coloration of the under side, and the abnormality of the head associated 

 with the complete condition. He argues that this kind of variation 

 cannot be explained simply as reversion to an ancestral condition. In 

 this I entirely agree with him. But he does not consider the question 

 how the occurrence of this variation bears upon the specific characters, 



