260 SPECULATIVE SCIENCE 



cross down a donkey's back, is useful to it. It seems possible, 

 even probable, that these things are the unavoidable consequences 

 of the elementary combination which will produce the quagga, 

 or a beast like it. Darwin himself appears to admit that corre- 

 lation will or may produce results which are not themselves 

 useful to the animal ; thus how can we suppose that the beauty 

 of feathers which are either never uncovered, or very rarely so, 

 can be of any advantage to a bird ? Nevertheless those concealed 

 parts are often very beautiful, and the beauty of the markings on 

 these parts must be supposed due to correlation. The exposed 

 end of a peacock's feather could not be so gloriously coloured 

 without beautiful colours even in the unexposed parts. According 

 to the view already explained, the combination producing the 

 one was impossible unless it included the other. The same idea 

 may perhaps furnish the clue to the variability of abnormal 

 organs and widely diffused species, the abnormal organ may 

 with some plausibility be looked upon as the rare combination 

 difficult to effect, and only possible under very special circum- 

 stances. There is little difficulty in believing that it would more 

 probably vary with varying circumstances than a simple and 

 ordinary combination. It is easy to produce two common wine- 

 glasses which differ in no apparent manner ; two Venice goblets 

 could hardly be blown alike. It is not meant here to predicate 

 ease or difficulty of the action of omnipotence ; but just as 

 mechanical laws allow one form to be reproduced with certainty, 

 so the occult laws of reproduction may allow certain simpler 

 combinations to be produced with much greater certainty than 

 the more complex combinations. The variability of widely 

 diffused species might be explained in a similar way. These 

 may be looked on as the simple combinations of which many 

 may exist similar one to the other, whereas the complex com- 

 binations may only be possible within comparatively narrow 

 limits, inside which one organ may indeed be variable, though 

 the main combination is the only possible one of its kind. 



We by no means wish to assert that we know the above 

 suggestions to be the true explanations of the facts. We merely 

 wish to show that other explanations than those given by Darwin 

 are conceivable, although this is indeed not required by our 

 argument, since, if his main assumptions can be proved false, 



