240 BONY FISHES vm. 6 



This replacement of one type by another appears much more 

 remarkable when we reflect that by a 'type', say the palaeoniscid, we 

 do not mean a homogeneous set of similar organisms all interbreeding. 

 Quite the contrary; the palaeoniscids included many separate lines, 

 each with its own peculiarities. When one 'type' therefore is thus 

 replaced by another it must mean either that some one of these many 

 stocks gives rise to a specially successful new population, which ousts 

 the old ones or that all the members of the stock are changing their 

 type together. It is not easy with a record such as that of the fish, 

 which is far from continuous, to say which of these is true in any 

 particular case, but we have sufficient evidence to be sure that either 

 of them is possible. 



There is no certain example in the Actinopterygii of a single new- 

 type replacing all the former ones, but the origin of the Teleostei may 

 perhaps show a case of this sort. It is possible that the Teleostei is a 

 monophyletic group, arising from a single type such as *Leptolepis 

 and proving so successful that nearly all creatures of previous types 

 soon disappeared. 



The Actinopterygii provide also examples of the other and perhaps 

 even more interesting process, the parallel evolution of a number of 

 different lines. There can be no doubt that from Palaeozoic times 

 onwards several independent lines of fishes have shown similar 

 changes. The tail has become shorter and more nearly symmetrical, 

 the body has become flattened and deepened dorso-ventrally, while 

 the pelvic fins have moved forward and the scaly armour has been 

 reduced, all of these being signs of a more effective swimming and 

 steering system (p. 244). As we come closer to modern times and the 

 geological record becomes more complete we obtain more and more 

 critical evidence that such changes have occurred in separate popula- 

 tions. Thus in the descendants of the holosteans we can recognize at 

 least three such lines, that leading to the round-shaped *Microdo?i, 

 another leading through *Caturus to Amia, and a third through 

 *Leptolepis to the Teleostei; probably there were many others. The 

 important point is that although each line possessed peculiar speciali- 

 zations of its own, they all showed some shortening, development of 

 symmetry of the tail, and thinning of the scales. 



It is a considerable advance to be able to recognize such tendencies 

 within a group. We begin to see the possibilities of a general state- 

 ment on the matter. Instead of examining a heterogeneous mass of 

 creatures called Actinopterygii we can recognize an initial palaeoniscid 

 type and state that in subsequent ages this has become changed in 



