374 EXPERIMENTAL STUDY OF INHERITANCE 



reinvigoration ; when the results of the cross are inbred, forms 

 like the original parent will reappear. 



Old-established form. " Fresh blood." 

 A B 



If A be dominant, A(B); or B(A), if B be dominant. 

 x x 



A A 



I! II 



n A + n A(B) n A + n B(A) 



Or if A(B) be inbred the result will be n A + 2 n A(B) + v B. 

 Or if B(A) be inbred the result will be n A + 2 n B(A) + n B. 

 There is obviously no theoretical danger of losing A. 



No one can, of course, at present say that these " simple 

 equations " will apply to the introduction of fresh blood into 

 a herd of cattle, but the time has come for more daring ex- 

 periment on Mendelian lines. It might obviously happen that 

 the " fresh blood " (B) introduced was quite incompatible with 

 the pure-bred (A), and the progeny was an undesirable freak. 

 But do not such casualties happen under the present instinctive 

 or empirical regime followed by most breeders ? 



7. Other Experiments on Heredity 



Our survey of cases must be supplemented by reference to 

 the works of Bateson, T. H. Morgan, Punnett, De Vries, and 

 others ; but we have said enough to show, (i) that Mendelian 

 phenomena are well illustrated in certain cases -e.g. peas, mice, 

 rabbits, poultry, snails ; (2) that in other cases, while there are 

 clear Mendelian phenomena according to some observers, dis- 

 crepant results have been reached by others e.g. silk-moths ; 

 (3) that in other cases, while there are hints of Mendelian pheno- 

 mena, the results^ cannot be readily interpreted in conformity 

 with Mendelism e.g. pigeons. 



Of great interest are Johannsen's experiments with pure lines 

 of beans, which go to show that small quantitative fluctuations 

 are in this case unimportant, the real evolutionary steps being 

 attained by brusque variations, corresponding to Darwin's sports 

 and De Vries's mutations. Selection continued generation after 



