238 THE TRANSFORMATIONS OF THE ANIMAL WORLD 



become incapable, under any conditions, of again 

 forming a complete hind limb, and of reassuming 

 quadrupedal functions. Should the circumstances 

 of their surroundings become modified in a manner 

 unfavourable to their natatory existence, the 

 Manatee and the Siren would suddenly become 

 extinct, but without leaving survivors adapted to 

 different functions. In the same way the Ammon- 

 ites, such as the Pinacocerata, in which the line of 

 suture has acquired an elegant degree of complica- 

 tion exceeding, doubtless, that of all the other 

 branches of the Cephalopods, died out at the 

 end of the Triassic period, without perpetuating 

 themselves in the branches with simpler partitions 

 of the commencement of Jurassic times. 



By the side of the law of Irreversibility, it is 

 right to make an interesting place for an idea, 

 already ancient, but which has acquired a new 

 lustre from the recent writings of Rosa. I refer 

 to the law of the progressive reduction in variability. 

 Haeckel had already shown that groups on the 

 road to extinction produced no new varieties, and 

 taking, with Wallace, the standpoint of Darwinian 

 selection, it must be admitted that the chances of 

 survival of a type are in direct ratio to the number 

 of favourable varieties it produces. Rosa proves 

 that every series of forms specialized in one direc- 

 tion is doomed to extinction, for the reason that 

 these forms are no longer competent to vary suf- 

 ficiently. It is perfectly true that the number and 

 extent of the variations diminish as fast as the 

 specialization increases. Palaeontology can furnish 

 numerous proofs of this. The great group of 



