No. 503] SHORTER ARTICLES AND CORRESPONDENCE 729 



gation" by which domesticated types are induced to change 

 some of their characters, hut it docs not appear that this in- 

 tensive segregation is a condition of evolution. Differences 

 between small groups are more obvious and more readily de- 

 finable because small groups are generally more uniform, like 

 the pieces with uniform patterns which may be cut from a 

 variously figured fabric. Isolation is the shears that splits the 

 species, not the loom that weaves it. The weaving is done when 

 the fabric is broad. The larger and more diversified species 

 make the truly constructive evolutionary progress. 



The evolution of a species is in no way dependent upon its 

 being split into smaller groups, but is more likely to be hind- 

 ered by narrow subdivisions. If the groups are too small they 

 degenerate and become extinct, instead of continuing their evo- 

 lution. Isolation, though making more species, impedes evolu- 

 tion. In like manner, selection favors adaptation, because it 

 keeps species from evolving in non-adaptive directions. Isola- 

 tion and selection may still be considered as evolutionary factors 



if this time-h >rcd reckoning is too sacred to be changed, but 



they must stand as negative factors instead of positive, if my 

 interpretation is correct. 



Darwin saw in his later years that evolution is not alto-ether 



Darwin was concerned to show that species multiply and 

 diverge in nature, for this is good evidence of the general Pact 

 of evolution. Nevertheless, it should not be assumed that all 



