March, 1895. ORIGIN OF SPECIES AMONG FLAT-FISHES. 177 



reasonably hold that the explanation of the similarity of individuals 

 in a species is the constant intercrossing of individuals. In other 

 words, Weismann's condition of panmixia occurs within every 

 homogeneous group, and tends to produce not degeneration, but 

 uniformity. There is no reason why this condition should prevent 

 modification. A species may produce great variations, and these 

 may be gradually extended to all the individuals, so that the later 

 characters of the species may be very different from the original, but 

 unless there is segregation or isolation of one group of individuals 

 from another there will be no divergence, no splitting up of the one 

 species into two or more. 



Although the mode in which Eimer has expressed his views on 

 variation is sometimes obscure and involved, there can be no doubt 

 that he has perceived important phenomena which have been more or 

 less neglected by other naturalists. The relations of several of the 

 characters of Zeugoptems, the fact, that is, that they differ only in 

 degree of development, form examples of what Eimer calls " gene- 

 pistasis." But this is not to be considered as implying that, as 

 Eimer believes, the less developed characters represent stages actually 

 passed through in the development of the individuals of the species 

 where these characters are more developed. The development of 

 specific characters is a subject I am investigating, and I have not 

 actually traced the development of those of Zeugoptems ; but I hold 

 this part of Eimer's view to be exceedingly improbable. It is 

 evident, where the character depends on the number of parts, that, as 

 Bateson points out, the greater number is not usually formed by 

 addition to a smaller number, but by the greater subdivision of the 

 developing blastema when the parts are about to be formed. It seems 

 to me that the facts represented by the term " genepistasis " are 

 simply due to the common descent of closely allied species. The 

 generic characters are either derived from the common ancestral 

 species, or represent the similarity of variation due to the blood- 

 relationship of the species. The specific differences of degree are 

 merely instances of the general fact that the species have varied 

 independently, some developing characters of the ancestral species 

 more strongly, some less, while at the same time variations occur in 

 one species which are not represented at all in the others. 



Plymouth. J. T. Cunningham. 



{To be continued.) 



