42 GENERAL PRINCIPLES OF ZOOLOGY 



existence and thus fixed as permanent characters. It is not easy to 

 see how many cliaracters, especially those used in classification, can be 

 of use to their owners. It can only be said that they have developed in 

 correlation with other important characters. But useful characters 

 must be considerable in order to be seized upon by natural selection. 

 Fortuitous variations with which Darwinism deals are too inconsiderable 

 to be utilized liy the organism and so to be of value in the struggle for 

 existence. In most cases, too, alteration in one organ alone is not enough 

 to be of value; usually a whole series of accessory structures must be 



15B. — Paradisea apoda, female (after Levaillant). 



modified. In short, there must exist a harmonious co-adaptation of parts, 

 which presupposes a progressive and well-regulated development extend- 

 ing through a long space of time, during wliich the struggle for existence 

 could have exerted no directing intlucnce. Thus, the wing of a bird in 

 order to be used for flight must have alrcad}^ reached a considerable size; 

 the muscles for moving it, the supporting skeletal parts, the ner^■es 

 running to it must have a definite formation and arrangement. Then 

 there are diificulties in that most animals are bilaterally or radiallv sym- 

 metrical, many in addition segmented. In all these cases the same organ 

 is repeated two or more times. Organs which are repeated symmetric- 

 ally and usually those which are segmental agree in general in structure. 

 One must therefore admit that the alterations of chance must have oc- 

 curred at at least two points siTnultaneously and in exactly the same wa)'. 

 A further oljjection is that the action of natural selection would, under 

 ordinary conditions, be negatived by unhindred crossing of the varving 

 forms. If, for example, we do not isolate fantails from other jiigeons, 

 they will cross with these, and their descendants will soon resume the 



