The Evidence from Variation. 153 



feet, and the cells which are involved will therefore have 

 a greater tendency than those involved in generic char- 

 acters to throw off gemmules. These characters will 

 therefore be more variable in the descendants than 

 generic characters. 



Another law, the evidence for which is given by Dar- 

 win on page 44 of the " Origin of Species," is that 

 " species of the larger genera in eacli country vary more 

 frequently than the species of the smaller genera." 



When a country contains a great number of species 

 of a genus it is generally safe to conclude that they have 

 recently varied and diverged from each other. As the 

 tendency to vary is in itself hereditary, and as one 

 variation is in itself a cause of other variations, our 

 theory of heredity would lead us to expect species which 

 have recently undergone considerable change to show a 

 tendency to vary still further, and we should therefore 

 expect the species of large genera to be, as a rule, more 

 variable than the species of small genera, although there 

 is no reason why this rule should be absolute. 



A still more interesting law is that " a part developed 

 in any species in an extraordinary degree or manner, in 

 comparison witli the same part in allied species, tends to 

 be highly variable" (" Origin of Species," p. 119). 



When one species of a genus agrees with the other 

 species in most particulars, but differs from them all in 

 some one respect, we may conclude that the peculiar 

 organ or feature has recently been modified. Natural 

 selection has therefore had less time to perfect the ad- 

 justment between this part and the remainder of the 

 body than it has had to perfect the relations between other 

 parts, or between the same parts in the other species. 



This peculiar part will accordingly be in a favorable 

 state for the production of gemmules, and it will there- 



