THE ORIGIN OF SPECIES 71 



dots and half would be without, but some of those which had 

 black dots would be red and some would be uncoloured, and some 

 of those which had no black dots would be red and some would 

 be uncoloured. There would thus be four kinds of balls, viz. : 

 (1) red with black dots, (2) red plain, (3) uncoloured with black 

 dots, (4) uncoloured plain. Thus every time that members of 

 two races distinguished from one another by two differentiating 

 characters are crossed with one another, four kinds of germ-cells 

 are produced ; two of these varieties of germ-cell are the same 

 as the germ-cells borne by the two parental races, but two carry 

 new combinations of characters, and all these germ-cells if they 

 unite with others of the same kind produce offspring with stable 

 characters. Thus by crossing two such races two new stable 

 varieties can be created. If two races are crossed which differ 

 from one another in three definite characters, the hybrid pro- 

 duces six (i.e. 3x2) different types of germ-cells ; if in four 

 characters, eight (4 x 2) different kinds of germ-cells. It will 

 thus be seen that by crossing two races differing in a large 

 number of characters a very large number of new stable com- 

 binations can be formed. This discovery is of enormous practical 

 importance, because it enables breeders to combine the desirable 

 qualities found into different races and to produce a new 

 stable variety bearing all the wished-for qualities. But there 

 is grave reason to doubt whether any such process has played a 

 part in the formation of natural species. Bateson, the most 

 brilliant member of the Mendelian school, has shown that in 

 most cases which are closely examined the differentiating 

 character separating two races consists in the presence of one 

 race of some substance or quality which is absent in the other 

 race. The albino, or white mouse, for instance, is deficient in 

 a substance necessary for the production of colour which the 

 ordinary mouse possesses. Nearly all the differences between 

 domestic races with which the Mendelian biologists have ex- 

 perimented are of this character. The wild race, or original 

 type of the species from which the domestic race has been 

 derived, is characterised by the possession of all the differ- 

 entiating characters ; the domestic race by the absence of one 

 or more of them. All domestic races experimented with are 

 therefore in a sense cripples, and their differences from the 

 parent species of a different character from those which usually 



