150 THE BREEDING OF ANIMALS 



to assume that there exists an insufficient quantity of 

 this substance and hence an incomplete transposition of 

 sugar to starch in the wrinkled pea. 



It is not at present possible to apply the presence and 

 absence hypothesis to all cases of apparent mendelian 

 inheritance. That it applies to a very large number of 

 characters is obvious, but as pointed out by Darbishire 

 there are many cases which cannot at the present time 

 be explained on this theory, and in fact will be obstacles 

 in the way of its general application. Some characters 

 seem to be dominant in one class of plants or animals 

 and recessive in another. The polled character in cattle 

 is unquestionably dominant, while the possession of horns 

 is a recessive. In the case of sheep the reverse seems to 

 be true. The white color of pigs is dominant to black, 

 but the black color of sheep is dominant to white. 1 



142. The theory of mutations. It is clear that all 

 improvements in the domestic animals must come through 

 variation. If the offspring was always an exact repro- 

 duction of the parent, improvement would be impossible. 

 But how have the improved qualities now possessed by 

 our domestic animals come about? Have these qualities 

 come through a gradual and continuous series of changes, 

 each better than the last, or have they come through 

 sudden and radical variations ? A study of the ancestral 

 history of both plants and animals gives clear evidence 

 that new types have originated by both kinds of variation. 



143. Two important classes of variation. Darwin 

 recognized these two distinct types of variation, but 

 believed that the most important changes in organic 

 beings were due to small but gradual and continuous 

 variations in a given direction. In his earlier writings 



^unnett, "MeDdelism," p. 29. 



