THE MECHANICAL SYSTEM AND ITS LIMITS 369 



Against heredity the objection has been raised that 

 ordinary variations were not preserved, and that it was 



cally, the variations must depend absolutely on mechanical grounds. 

 If it is assumed that the variations only follow definite directions and 

 are restricted — if, in other words, there are limits in the organisation of 

 animals that prevent the variations from passing a certain measure — 

 they are no longer mechanical ; in that case there is a force in 

 organisms that avoids unserviceable and aimless changes — a purposive, 

 teleological force. Variations do not then depend on chance — which 

 alone would be mechanical — but are directed according to definite 

 internal principles. 



Thus we see that variations cease to be mechanical as soon as they 

 are assumed to be definitely directed and limited. Natural selection 

 is then no longer a mechanical principle, and a unified conception of 

 the world is impossible. As soon as a Mechanist recognises variations 

 of that kind he abandons the possibility of a harmonious system, and 

 is no longer a Mechanist. It is just the same if he recognises other 

 principles that can be shown to be teleological, as we shall show 

 presently of sexual selection and the Lamarckian principle. If he 

 would remain a Mechanist he must abandon them in this case. That 

 is mere logic and consistency. But even if one holds the mechanical 

 system to be impossible and unattainable, these theories are none the 

 less useless, if they are teleological. He considers the development of 

 living things to be due to a purposive internal force, and so must 

 regard natural selection, the Lamarckian principle, etc., as merely 

 subsidiary. These theories have real value only for a Mechanist, and 

 they must, therefore, be mechanical. 



It is said that if there are no limits fixed in the nature of the animals, 

 natural selection would be able to equip the horse with wings and 

 create all the fantastic forms of our legends and fairy tales. How do 

 we know that it cannot ? Is the legendary dragon more wonderful than 

 the ichthyosaurus or the plesiosaurus ? Does it not show an unlimited 

 capacity when a sort of tape-worm is developed from a spider, and a 

 being that scarcely looks like an animal out of a crab? The point is 

 worthless on other grounds. The transformations of the animal world 

 are due to the fact that those survive which are most in harmony with 

 their conditions at the time. Hence when it is said that certain animals 

 ought to have arisen, one must have had some knowledge of the 

 conditions of life of the evolving species at the time. But we do not 

 know them — in the case of the horse, for instance. Horses with wings 

 would only be developed if the conditions of the horse were such that 

 the best leapers survived (compare the origin of the insects). But that 

 can hardly have been the case. 



2 A 



