SIMILARITY IN ACQUIRED ORGANS. 2$l 



only must these small variations be of use, but in 

 order that they should be selected, they must be of 

 so much use as to lead the individuals with this 

 variation to flourish in conditions where other indi- 

 viduals will have difficulty in living, and thus to 

 cause the modified variety to supplant the unmodi- 

 fied form ; and when, further, we remember the 

 results which must arise by crossing of modified and 

 unmodified individuals, it is simply impossible to 

 believe that this peculiarity could have been devel 

 oped by the natural selection of minute variations. 

 Or, take the case of the development of the horse's 

 foot. Granting that it is of great advantage to the 

 horse to have one large toe rather than five small 

 ones, it is incredible that any single ancestor, hap- 

 pening to have a slightly larger middle toe, would 

 so far surpass all others that it should event- 

 ually supplant the unmodified form. If we can 

 believe that many horses have varied simultaneously 

 and rapidly as regards the shape of the toe, it is 

 quite easy to see that natural selection might have 

 developed the modern horse's foot ; but so long as 

 there are only indefinite variations, it is no longer 

 within the realm of possibility. 



Similarity in Independently Acquired Organs. 



Natural selection finds further difficulty in ex- 

 plaining the similarity of independently acquired 

 organs. It is the verdict of one scientist that " it is so 

 improbable as to be practically impossible for two 

 exactly similar structures to have ever been inde- 

 pendently acquired." It is found that, as a rule, 



