THE PRODUCTION AND FIXATION OF NEW BREEDS. 



35 



the contributions made by the two parents are similar in character or 

 when they unite to form a blend in the offspring, we are apt to over- 

 look their separate existence. It is easiest to demonstrate the exist- 

 ence of units when a different contribution is made by either parent 

 and when the inheritance is alternative between the two parental con- 

 tributions. It was in this category of cases that Mendel's discovery of 

 the principle of unit characters was made. 



The practice of breeders had long accorded with the principle of unit 

 characters, because that practice gave desired results. Now came a 

 wholly unpractical man of science, a recluse, but a man who could give 

 a reason, and that reason (the principle of unit characters) marks the 

 first and greatest step yet taken in the transformation of breeding from 

 an art to a science. Breeders knew that crossing of dissimilar parents 

 induced variation ; from variations so induced selection might be made 

 of such variations as were desirable and these might be fixed by in- 

 breeding. Further than this the analysis of the matter had not been 

 carried. We now know, thanks to Mendel, that the variation following 

 crossing is due largely to the production of new combinations of unit 

 characters. Thus, in rats, to take an illustration from my own experi- 

 ence, coat pigmentation and coat pattern are distinct unit characters 

 separately heritable. A cross between parents unlike in both par- 

 ticulars makes the matter entirely plain, leading to the production in 

 offspring of the second generation of all the possible combinations of 

 the units in question. A wild self-gray rat mated with a tame black 

 hooded 1 one produces offspring all with the wild color and wild color 

 pattern (though usually slightly modified) ; but when these young are 

 bred together, there are produced in the next generation, four sorts of 

 young representing the following combinations: 



(1) Wild color and wild pattern. 



(2) Tame color and wild pattern. 



(3) Wild color and tame pattern. 



(4) Tame color and tame pattern. 



By selection, any or all of these combinations may now be fixed. 

 The empirical breeder's idea that fixation of a new variation can be 

 secured only by inbreeding, we now see, rested on a misconception. It 

 is not the relatedness of parents which results in fixation, but the fact 

 that they represent like combinations of characters. We might cross 

 one pair of rats in Massachusetts and another one in California, and 

 bringing the hybrids together in pairs secure complete fixation of 

 the new varieties produced without ever pairing relatives. Here we 

 get from the principle of unit characters a practical suggestion worth 

 remembering, which allows fixation of characters without inbreeding. 



Not all characters give alternative inheritance; in the case of some 

 a blend (or condition intermediate between that of the two parents) 



1 A " hooded " rat is white all over except on the head and shoulders 

 and along a median dorsal back-stripe, in which regions it bears either 

 black or gray pigmentation. 



