36 ORGANIC EVOLUTION 



horses with curled hair occur, and these always have hoofs 

 exactly like those of a mule, while the hair of the mane 

 and tail is much shorter than usual. Now, if any of 

 these characters were useful, the others correlated with it 

 might be themselves useless, but would still be tolerably 

 constant because dependent on a useful organ. So the 

 tusks and bristles of the boar are correlated and vary in 

 development together, and the former only may be useful, 

 or both may be useful in equal degrees." If, in case of the 

 boar, the conditions of life became such that increase in 

 the size of the tusks would be useful, there might be de- 

 veloped a race of boars with larger tusks, and at the same 

 time the length and coarseness of the bristles would 

 probably increase, not because better developed bristles 

 are needful in themselves, but because of the correlation 

 between large tusks and coarse, long bristles, a correlation 

 the reason for which we are unable to understand. 



In the case of the regular patterns in the skeletons of 

 many unicellular animals and plants, to which we have re- 

 ferred, it is possible, I will not say probable, that the regular- 

 ity of their arrangement may be due to the constitution of 

 the protoplasm of the cells which form them, to some regular 

 arrangement of the constituent particles of this protoplasm, 

 especially as regards its chemical activity, so that the 

 skeletons will be regular, not because of any utility in their 

 regularity, but because they are each formed by a bit of 

 protoplasm so constituted that, if it is to form a skeleton 

 at all, it must form a regular skeleton. Thus the regular- 

 ity of the diatom shell may be due to correlation with a 

 kind of protoplasmic structure which is itself useful. 



But, though natural selection is a factor in evolution* 



