III. 



The Origin of Species among Fiat-Fishes. 



THE purpose of this paper is to consider how far, by the help of the 

 various principles which have been suggested or advocated, we 

 can obtain a satisfactory explanation of the evolution of the family of 

 flat-fishes. We may assume that descent with modification is the 

 process of organic evolution. But the question is, how far we can 

 really explain the modification, and especially how we can explain the 

 peculiarities of its results, the relations of structure which form the 

 subject-matter of taxonomy, the relations between the characters 

 which divide organisms into species, genera, families, etc. Thus, by 

 the evolution of flat-fishes I mean, not what is usually meant — how 

 they came to lie on the bottom on one side and have both eyes on 

 the upper side ; but how and why they diverged into a multitude 

 of distinct species arranged in a number of genera, which form higher 

 groups called sub-families. This method of studying the problem of 

 evolution, namely, by examining all the problems presented by the 

 morphological relationships of a particular kinship, from the family 

 down to the individuals, or from the individuals to the variety, from 

 varieties to species, species to genera, etc., has very seldom, if ever, 

 been followed. 



Zoologists may be divided into three classes:— (i) The investi- 

 gators of evolution who, with Darwin as their pattern, follow the 

 inductive method, and having arrived at some general principle give 

 a general survey of the biological facts which support it. (2) The 

 embryologists and comparative anatomists, who investigate and de- 

 scribe morphological facts and structural phenomena, and trace homo- 

 logies : their work consists almost entirely in investigating the results 

 of descent with modification, and its historical course ; they have 

 little concern with the origin of modifications, though they usually 

 assume one theory or another. (3) The systematists and naturalists, 

 who investigate the minutiae of structure and habit which characterise 

 the different kinds of organisms, from main divisions down to the 

 smallest distinctions by which a variety or race can be recognised : 

 these as a rule do not trouble themselves about evolution at all ; their 

 method is simply empirical. We may roughly characterise the three 

 classes as the followers of Darwin, the followers of Cuvier, and the 

 followers of Linnaeus. The theorists of the first class are too apt to 



