FACTORS OF EVOLUTION 439 



fishes adaptations exist, but they do not follow, in the majority 

 of instances, the lines of classification : in some instances, e.g.. 

 Skates and Flat-fishes they do, but these are rather the excep- 

 tions than the rule. It may be suggested that the apparently 

 useless or indifferent characters are correlated with the useful, 

 but of this there is no sufficient evidence. Again, it might be 

 suggested that the distinguishing characters are or were origin- 

 ally the direct result of conditions, but we have no proof of 

 this : we cannot say that fresh water produces bone, seeing 

 that many of the most bony forms now exist in the sea. We 

 are driven then to substitute for the terms variation and selec- 

 tion, the terms spontaneous variation and survival: we know 

 that the new types which have arisen have survived and multi- 

 plied, but we have no right to assert that the distinguishing 

 characters have been directly necessary or useful in the struggle 

 for existence. 



With regard to adaptations the question is different : here, 

 admitting the advantage, we may doubt whether the variations 

 were spontaneous, that is to say, independent of the conditions 

 of life. If, for example, the bladder arose as a spontaneous 

 variation we may ask how it is that it arose only when the fish 

 was compelled to breathe air, why has no sign of it ever 

 appeared in the shark type in the sea. Why is it that various 

 adaptations for atmospheric respiration have arisen in different 

 groups of fishes which were compelled to breathe air? Why is 

 it that in the same fish in the individual lifetime the structure 

 changes gradually when the habits change, as in the flat- 

 fishes ? It is true that to some extent the cartilaginous 

 skeleton succeeds the notochord and the bone succeeds the 

 cartilage, in individual development, but these changes do not 

 correspond to changes of habit. In some cases we see similar- 

 ity of habits producing similarity of form in fishes belonging 

 to different groups, as in the numerous eel-like fishes which 

 hide themselves in holes or burrow in the mud, for example 

 cyclostomes, lung-fishes, electric eels, true eels, and blennies, 

 fishes far apart in classification, and we cannot perceive that 

 the fundamental differences between these types were all ori- 

 ginally due to differences of habit. Enough has been said to 

 show that the various theories of evolution are far from afford- 

 ing a satisfactory explanation of all the problems arising in the 



