206 



THE AMERICAN NATURALIST 



[Vol. XLIV 



waste products, such as carbon dioxide and ammonia. 

 Of course, only the former can be of significance in in- 

 heritance, and it is probable that whatever determines 

 in how far the former process of tyrosin oxidation shall 

 take place, will be found to control or determine these 

 quantitative differences in the end-products of the germ 

 cells. 



It is interesting and suggestive to note in this connec- 

 tion that Morgan ( '09) has adopted a quantitative inter- 

 pretation of the sex-determining factors in tiie germ cells 

 of insects and other organisms. 



It is important to enquire what is the nature and origin 

 of this quantitative difference, such as must occur in the 

 germ cells of the hybrids I have described. It is most 

 reasonable to suppose that it originates at the time the 

 germ cells are formed, i. e., during the reduction divi- 

 sions, for the whole individual from a very early stage 

 of the seedling to mature development shows one or the 

 other of the alternative characters. Hence, the germinal 

 difference must date back to the fertilized egg, and if 

 this be true we must logically take it back farther still, 

 on account of the definite Mendelian proportions, to the 

 individual germ cells which united in fertilization, and 

 hence probably to the reduction divisions. 



The question of the relation of the chromosomes to 

 Mendelian behavior has been so often discussed that I 

 will only touch upon it here, to point out that the theory 

 of qualitative hereditary differences among the chromo- 

 somes is not incompatible with the view that differential 

 Mendelian characters are properties of the germ cell as 

 a whole. If, during the reduction division, certain of the 

 chromosomes which are segregated into separate colls 

 are chemically unlike, then different groupings will 

 arise and the whole germ cells whose nuclei the different 

 groups enter will soon experience the effect 

 differences, and accordingly such germ cells would be 

 expected to become unlike as a whole. This follows 

 neeessarilv, not onlv from the continual active inter- 



