238 NATURAL SCIENCE. April, 



the bones and teeth being much more strongly developed on the blind 

 side, in accordance with the general habit of feeding on small inverte- 

 brate forms which live on the ground, and which the fish seize from 

 above with the lower side of the mouth. But many of the characters 

 are neither adaptive, nor known to be correlated with the adaptive : 

 such as the narrow symmetrical pelvic fins differing from those of the 

 Rhombinae ; the free edge of the preoperculum differing from the adnate 

 edge in Soleniae ; the slight anterior extension of the dorsal fin. 



We return again to the family characters : are they all adaptive ? 

 Certainly the Pleuronectidae appear to offer strong support to the 

 view that families are distinguished by different adaptations. But 

 among the adaptive characters there is evidence that one at least, the 

 absence of pigment from the lower side, is simply a consequence of the 

 condition of life, not an advantage, and therefore not due to selection. 

 There are other family characters which are not known to be 

 adaptive at all, namely, the well-developed condition of the pseudo- 

 branchia, and the character of the tail, the Pleuronectidae differing 

 strongly in these respects from the Gadidae, supposed to be their 

 nearest allies. In the Gadidae the tail very largely consists of dorsal 

 and ventral rays symmetrically corresponding to one another, while 

 the terminal portion of the tail, composed of ventral rays which have 

 become terminal, is very small. In the Pleuronectidae very nearly 

 the whole of the tail is formed of the ventral rays corresponding to the 

 upturned end of the notochord. 



Moreover, if we examine the diagnostic characters of other 

 families of fishes, we find that the most salient among them are by no 

 means obviously adaptive. We have no evidence that the mode of 

 swimming of the Gadidae is such that three dorsal fins and two anal 

 are necessary to its due performance. The adipose fin characteristic 

 of the Salmonidae appears to have no function at all. The 

 Scombridae are distinguished by minute posterior dorsal and ventral 

 fins, which are not known to have any function. When we 

 ascend to still higher divisions, the dominion of adaptation is by no 

 means complete. For instance, we cannot say that there are 

 differences in mode of life between Teleostei and Elasmobranchii 

 which exactly correspond to their striking differences of structure. 

 It may be said that the lungs of the amphibian are adapted to breathe 

 air, the gills of the fish to aquatic respiration, but who shall say that 

 the branchial apparatus of the Teleostei is adapted to a different 

 kind of respiration from that which the branchial apparatus of the 

 Elasmobranchii performs ? 



I cannot now discuss the limits of adaptation in the main 

 divisions of the animal kingdom. I will content myself with 

 suggesting that, although it is difficult to find anything in the 

 characteristics of birds which is not a modification of the reptilian 

 organisation specially adapting it to aerial life, still it is generally 

 held that the birds correspond only to a subordinate division of the 



