146 LECTUKES ON BIOLOGY 



succeed in a comparatively short time in producing, more or less 

 exactly, the desired results. 



Suppose, now, that the breeder desires to impart to his 

 carrier-pigeons a certain colour. He will, therefore, proceed in 

 precisely the same manner, selecting among his carriers, for 

 breeding purposes, only such as exhibit the desired colour in the 

 most pronounced manner. Continuing along this line, a skilful 

 breeder will gradually produce pigeons which are widely differen- 

 tiated from their parents on almost every point. The production 

 of new races is, therefore, not, as is generally assumed, the result 

 of the crossing of different animal species, though crossing may 

 lead to the formation of new races, but by the patient, deter- 

 mined accumulation through numerous generations of slight 

 variations. 



What is said here of pigeons applies to all other domestic 

 species. We need only mention the breeds of dogs, sheep bred 

 for wool or flesh, rabbits with long lop-ears or short erect 

 ears, domestic cattle and pigs, and many other products of 

 domestication. 



It must, of course, be understood that only those parts and 

 characteristics can be subject to conscious variation that may 

 be controlled from outside. It is, therefore, frequently found 

 that domestic species exhibit, together with widely divergent 

 external characteristics, a surprising similarity of internal 

 structure. It is obvious that by artificial selection generally 

 only those parts are transformed the transformation of which is 

 desired, whilst the rest of the body retains its original qualities. 

 For instance, in the various breeds of strawberries and goose- 

 berries flowers and leaves are alike, but the fruits differ both in 

 appearance and taste. In the numerous species of cabbage, the 

 flowers, which are of no account to man, are uniform, but the 

 edible leaves show a great variety. In tulips and other garden 

 flowers the process is reversed ; flowers are widely differentiated 

 in form, size, and colour, whilst there is no difference between 

 the leaves. Finally, in the various breeds of silk-worms, cater- 

 pillar, pupa, and imago are surprisingly alike, but the cocoon 

 that is, the part which is of value to man has been greatly 

 transformed by breeding. 



How enormously artificial selection may increase certain 



