122 REDUCTION OF VARIABILITY. 



geno- variation, and how it is possible for a trained observer to 

 pick out groups without potential variability. Such groups are 

 really remarkably stable and uniform. A trained nurseryman 

 will be able to name correctly hundreds of Appleclones, no 

 matter where he meets them, and that even in winter, by the 

 looks of the twigs. All hyacinth bulbs may look alike to a casual 

 observer, with the exception that some are a little narrower, 

 others a little browner, small differences, which may be 

 thought due to differences of age or soil. But a trained bulb- 

 grower will give you the name of dozens of hyacinths, if you 

 hand him the dry bulbs. And the wheat expert will know his 

 wheats back, no matter where you grew them. 



Any species becomes pure for its type, inevidentiy and auto- 

 matically, and the only thing which counteracts it, is crossing. 

 Given a certain geno-variation due to crossing, and resulting in 

 the variation of some important organ, selection may come 

 into play and decide the ultimate genotype for which the 

 group will become stable. 



If we find that practically all the house-rats in Holland are 

 black, we may think that it is advantageous to them to be 

 black, and if the house-rats in Java are mainly agouti, we may 

 think that to be agouti is an advantage for rats in Java. But 

 if we see that in the same city, small colonies of house-rats are 

 nearly always homogeneous in respect to colour, and different 

 one from the other, we must see that it cannot be the useful- 

 ness of any colour which determines the purity of the group in 

 respect to it. And this is what Lloyd found in British India, and 

 what we found in the farm-houses in Holland and France, and 

 later in the rice-godowns and tobacco-warehouses in Java, and 

 in respect to the rat-population of ships. 



In those plants or animals which form colonies from small 

 beginnings, and thus habitually isolate themselves in smaller 

 groups, such colonies speedily become pure for whatever geno- 

 type is given in the potential variability. Isolation is a very 

 important factor in species- formation. A group which is com- 

 pletely isolated from crossing with individuals from outside 



