THE ORIGIN AND THE EXTINCTION OF SPECIES 357 



IS able to protect itself against extermination by any other; the 

 increased fertility, the increased rapidity of locomotion, the increased 

 intelligence and similar qualities, may all be insufficient, and tlien 

 extinction follows; not, however, because tlie species has become 

 ' sende,' but because the variations possiljle to its organization do not 

 suffice to maintain it in the struo-o-le. 



In discussing germinal selection I mentioned the view expressed 

 l)y Emery, tliat excessive variation in the same direction from intra- 

 germinal causes has not rarely been the cause of the extinction of 

 species. I also mentioned the very similar view of Doderlein, who 

 could not refer at that time to germinal selection, but assumed internal 

 compelling forces, which pressed a variation irresistibly forward in 

 the direction in which it had started, even l)eyond the bounds of what 

 is useful for the desired end, and which miglit thus brino- about the 

 extinction of the species. I cannot entirely agree with these views, as 

 I have already indicated, because I do not believe that the impulse 

 to variation can ever become irresistible and uncontrollable. If it 

 could, then we should not see, as we do, innumerable cases in which 

 the augmentation or diminution of a part has gone on precisely to the 

 point at which it ceases to be purposeful. Even the degeneration of 

 organs only proceeds as far as is necessary to accomplish a particular 

 end, as we see plainly from the parasitic Crustaceans of different 

 orders. In many of these parasitic forms the swimming legs degene- 

 rate, but- in the female only, because these attach themselves by 

 suckers or in some other manner to their host, so that they cannot 

 let go again. But the males need their swimming legs to seek out the 

 females. The females too require them in their youth, in order to seek 

 out the fish from which they are to obtain their food-supply, and tluis 

 the degeneration of the swimming legs has come to a full stop exactly 

 at the point where they cease to be of use ; the}^ develop in early 

 youth and degenerate later, when the animal becomes sessile. In 

 accordance with the laAV of biogenesis we may sa}" that while the 

 degeneration is complete in the final stages of ontogeny, its retro- 

 gression was not continued back to the germ, but onl}^ to the young- 

 stages. From this it follows that the progress of a variation may 

 at any time have a goal fixed for it, and we have seen that tliis 

 is possible by means of personal selection, which accumulates the 

 never-failing fluctuations of the variation in the direction of plus 

 or of minus. In the individual id a determinant X may perhaps 

 decrease and possibly also increase without limit, althougli we 

 have no certain knowledge in regard to the latter point, but as 

 this determinant is contained in all the ids, there are always plus 



