NATURAL SELECTION 35 



Now if the organs of the body are so intimately con- 

 nected with one another that what affects one may affect 

 also the others, we see at once that changes produced by 

 natural selection in any organ of the body because of the 

 usefulness of such change, might very likely bring about 

 correlated changes in other organs, though these latter 

 changes be not in themselves useful. The secondary modi- 

 fications would not be directly due to natural selection 

 and so would not necessarily have to be useful. Their 

 connection with a useful modification would be enough 

 to account for their presence. This principle of correlation 

 is undoubtedly of great importance, but it is often difficult 

 to understand the details of its operation in particular cases, 

 since the nexus between the different organs, postulated by 

 this principle, may be so intimate and subtle as to be ex- 

 ceedingly difficult to study. 



As a very evident example of correlation think for a 

 moment of the great weight of the antlers of an elk and 

 the great strength required in the ligamentum nuch&, the 

 ligament which stretches from the top of the skull along 

 the back of the neck to the vertebrae between the shoul- 

 ders. The strength of this ligament must have increased 

 as the weight of the antlers which it supported increased, 

 the two being correlated. In this instance it is easy to 

 see the nature of the connection between the two struc- 

 tures, and that natural selection has probably produced the 

 correlation. In many cases, however, it is very, difficult 

 to understand the relation between correlated structures, 

 as in the case of the reproductive organs and the organs 

 affected by their extirpation in the domestic fowl. Wallace, 

 in his delightful book, Darwinism, says: "In Paraguay, 



