Toyama, Mendel's law* of heredity as applied to the silk-worm Grosses. p,33 



rations, though bred f'rom similar parents. The most careful 

 selection pursued through five generatiqns was not sufficient to get 

 rid of the occurrence of the antagonistic characters. 



Summary. 

 From the foregoing facts and consideratiöns we come to the 

 t'ollnwing conclusions: 



1. Of varions characters of silk-worms, some strictly follow 

 Mendel's laws (colours of cocoons, larval markings etc.), 

 while the others obey certain other laws which are not to he 

 exactly formulated as those of Mendel. (Compare the results 

 of Mendel, De Vries, Correns, Tschermak, Bateson 

 and Saunders etc.). 



2. No single instance has been observed in which irregulär cases 

 of Mendelian phenomena take place as Correns, Bateson 

 and Saunders, Tschermak etc. have observed in animals 

 and plants. 



B. As the result of crossing the disintegration of parent-cha- 

 racters takes place. Each character thus produced behaves 

 exactly like an independent character, breeding true to parents. 



4. Conversely, the combination of two characters takes place 

 and the form produced remains constant, when bred together. 



5. Thus we may accept Bateson's "allelomorphs" with good 

 reason. 



6. The behaviour of a character when crossed, however, depends 

 in some degree upon the characters of its ancestors, since in 

 one case (Siamese "whites" X Siamese "yellows") no disinte- 

 gration of the parent-characters takes place, while in other 

 cases (Japanese "whites" X Siamese „yellows" or Japanese 

 "whites" X European „yellows") it may often be observed (see 

 Weldon's opinion). 



7. Of characters belonging to Mendelian categories, we may 

 mention the following. — 



With regards to the colours of cocoons, the yellow character 

 is the most dominant, next comes the flesh coloured one and 

 then the greenish white, and lastly the pure white which is 

 therefore to be considered as recessive. 



As to the larval markings, the striped one comes first, next 

 ranks the Japanese normal one and lastly the no-marking one. 

 Thus the first, and the last are to be considered as absolutely 

 to be dominant and recessive respectively, those intermediate 

 are relatively so, for they act as recessive towards the first, 

 but as dominant towards the last. 



Those characters exhibit every combination of characters 

 according to the Mendelian formulae. 



