1 62 DARWINISM AND THE PROBLEMS OF LIFE 



every other living thing. If the eye is to retain its 

 efficacy, no part must be greatly altered unless the other 

 parts change in the same sense. If one of the parts of 

 the eye is suddenly and separately developed in an 

 animal that absolutely needs good sight, it fails in the 

 struggle for existence, and with it disappears the 

 inharmoniously developed eye. There is only one way 

 of improving the eye, but there are plenty of ways of 

 spoiling it. Hence if selection no longer watches over 

 the harmony of an organ, each part will vary on its own 

 account ; one part will advance in this direction, 

 another in that, and the end of it will be the spoiling 

 of the whole organ. 



It is clear, then, that panmixis, or the mixture of all 

 possible animal variations without weeding out the bad 

 ones, must lead to the degeneration of an organ. But 

 does an organ become smaller when selection ceases to 

 act? Diminution of size is the chief characteristic of 

 rudimentary organs. 



Certainly, when natural selection no longer prevents 

 the reduction of an organ by cutting off all variations 

 towards diminution, these will persist and be transmitted 

 like the rest. But are there not always variations in 

 the direction of increased size ? Why should the 

 reducing tendencies suddenly prevail .■* Natural 

 selection cannot influence variations ; it can only 

 accept or reject those that are offered to it. 



As a matter of fact, it has been proved beyond 

 question by a number of investigations that variations 

 towards increase and decrease — plus and minus 

 variations, as we call them — are as a rule equal in 



