REDUCTION OF VARIABILITY, 107 



we can imagine how the black markings of the other tit are 

 useful to it in any way. How can the striking blacl^i and pearl 

 grey pattern of the hooded-crow be useful to the animal, if we 

 know that it only differs in this colour from the black crow, 

 whose territory touches his, which has the same habits and 

 breeding-places, and inter-breeds with it where they coexist? 

 Of what use is the black tail-tip to the stoat, if the weasel can 

 thrive without it? Why is the tree-martin pure for its yellow 

 throat, and the house-martin for its white one? 



When we see how the purity, the stability of characters in 

 species is as great in respect to trivial things as in respect to 

 important characteristics, we are made to think. The useful- 

 ness of a small thing may escape us, and on the other hand we 

 may over-estimate the usefulness of a seemingly important 

 thing. But even so, it is obvious, that there are untold charact- 

 ers for which species are pure, and which cannot possibly be 

 accounted for as useful. In such cases natural selection cannot 

 be depended upon to furnish the clue. Conscious selection in 

 plant-breeding work is a process which on the whole is far 

 more severe than natural selection. It is remarkable, that in 

 new plants or animals, which have been subjected to a rigid 

 selection in respect to one or several useful characteristics, 

 purity has not only been attained for these characters, but 

 equally as well for the most trivial things to which the selec- 

 tionist never even gave a thought. We are speaking of the shape 

 and the arrangement of small hairs on the seed of some cereals, 

 of the colour of the eggs in fowls, of the juvenile colour of rab- 

 bits, the shape of the leaf in coffee, the colour of the leaf of 

 sugaf-beets. 



We cannot fail to imderstand that there must be something 

 besides usefulness ,to cause stability, purity. And if we find a 

 way to account for specific stability, which does not in every 

 instance take into account fine shades of usefulness, in other 

 words, if we find the nature of the process which causes pmity 

 during the process of natural or artificial selection, we shall have 

 made an important step toward an insight in specific stability. 



