86 CORRELATED VARIATIONS. 



of a black colour, suffer severely from eating the root of 

 Lachnanthes tinctoria. Similarly, buckwheat when in 

 flower is highly injurious to white or white-spotted pigs, 

 if they are exposed to the heat of the sun, but quite in- 

 nocuous to black pigs. 



These few instances suffice to show how widespread 

 and apparently capricious may be the range of correla- 

 tion. Until careful observations have been made, ac- 

 companied when possible by measurements, one can 

 never on a priori grounds assume that there is no cor- 

 relation between particular organs or parts of an organ- 

 ism, and that an agency acting on one part may not at 

 the same time be thereby indirectly modifying another. 



In their effects on the modification of species, prob- 

 ably by far the most important cases of correlation are 

 those in which the reproductive system is concerned. 

 Until recently comparatively little attention was paid 

 to such phenomena, and probably, even now, they are 

 far from being estimated at their true value by many 

 biologists. Perhaps the reason of this lies in the fact 

 that the physiological condition of an organism, or its 

 relative degree of sexual compatibility with other or- 

 ganisms of its own and of different species, is so 

 exceedingly difficult, if not impossible, to estimate. 

 Thus the degree of fertility cannot be tested in more 

 than one or two instances with each individual or- 

 ganism at least in the higher animals and so the 

 desired information can only be acquired by carry- 

 ing out most lengthy and laborious series of observa- 

 tions. There are sufficient data at our disposal, how- 

 ever, to indicate that the reproductive system is no 

 less subject to variation than any other part of the 



