i8 95 . ORIGIN OF SPECIES AMONG FLAT-FISHES. 171 



or on the origin of the actual differences between existing forms of 

 Pleuronectidae. Bateson mentions cases of the abnormality in question 

 in six species belonging to three genera. In no case is the variation 

 diagnostic of a distinct race, variety, or species. Bateson argues that it 

 is discontinuous, that is to say, normal parents may produce an indi- 

 vidual presenting the abnormality fully developed without intermediate 

 stages. There is every reason to believe that this is true. But 

 supposing this condition were to become constant and peculiar in 

 separate races, what would be the result ? The difference between 

 these individuals and normal individuals is a difference not in 

 specific or generic, but in family characters. Supposing the condi- 

 tion became constant and were regularly inherited, we could not even 

 say that we had a new family or sub-family, because the quasi- 

 symmetrical turbot would still be a turbot, the quasi-symmetrical 

 flounder still a flounder, and so on. We should simply have 

 dimorphic species, or species which presented two distinct but 

 parallel forms. If, however, the condition in question became the 

 normal condition, and the condition which is now normal became rare 

 or extinct, then the characters of the family would be changed, and 

 changed by a sudden leap from the one condition to the other. 

 Bateson might suggest that the normal characters of the Pleuronectidae 

 arose by a sudden change from the symmetrical condition, just as the 

 abnormal individuals at the present time arise without gradation from 

 normal parents. But this would be an untenable supposition. In the 

 individual Pleuronectid at the present day the family characters 

 arise by a gradual change in development ; and we should have to 

 believe that this change took place in certain individuals completely 

 the first time it occurred. But the occasional variation we have been 

 considering is probably due, as suggested by Bateson and myself, to 

 the occasional manifestation in the organism of a tendency to 

 symmetry, in virtue of which the characters of one side are repro- 

 duced in the other. This phenomenon, which Bateson calls 

 homceosis, is exhibited in asymmetrical animals, where the two sides 

 are normally different. But we have no evidence of any converse 

 tendency, of a general tendency to asymmetry, by virtue of which one 

 side may be entirely different from the other. We have therefore 

 no facts which suggest the probability that the first flat-fish arose 

 directly, without any intermediate steps, from parents whose two 

 sides were quite similar. Here, for the present, we may leave the 

 discussion of these abnormalities, merely recurring to the main point 

 insisted upon, that they affect, not specific or generic characters, but 

 family characters, and, moreover, family characters which are 

 adaptive. 



Let us turn to the consideration of specific diversity in this 

 family. For the details to be discussed I am largely indebted to a 

 very able paper by David Starr Jordan and David Kopp Goss, 

 entitled " A Review of the Flounders and Soles (Pleuronectidae) of 



