THE CHROMOSOMES IX HEREDITY. 247 



of salt. The results of chemical analysis show that one of the 

 most characteristic features of chromatin is a large percentage 

 content of highly complex and variable chemical compounds, the 

 nucleo-proteids, and therefore if, as assumed in the theory here 

 advanced, the chromosomes are .the bases of definite hereditary 

 characters, the suggestion of Bateson becomes more than a 



o o 



merely interesting comparison. 



We have seen reason in the case of the true-breeding hybrids 

 to suspect that the transmission by the hybrid of heterozygote 

 characters may be due to permanant union of the homologous 

 chromosomes. From this it is but a short step to the conclusion 

 that even if, as is normally the case, the chromosomes do not fuse 

 permanently, the very fact of their association in the same liquid 

 medium may allow a possibility of a certain degree of chemical 

 interaction. This must normally be slight, since its effects do 

 not appear to be visible in a single generation ; but the slightest 

 of variation as a result of repeated new association, even though 

 it tend in diverse directions, must in time, guided by natural 

 selection, result in an appreciable difference in a definite direction 

 between a chromosome and its direct descendant and hence be- 

 tween the characters associated with them. In this we have a 

 suggestion of a possible cause of individual variation in homol- 

 ogous chromosomes which we have already seen reason to sus- 

 pect (pp. 221 and 226). 



Finally, we may briefly consider certain observations which 

 seem at first sight to preclude the general applicability of the 

 conclusions here brought out. If it be admitted that the phe- 

 nomenon of character-reduction discovered by Mendel is the ex- 

 pression of chromosome-reduction, it follows that forms which 

 vary according to Mendel's law must present a reducing division. 

 But the vertebrates and flowering plants the very forms from 

 which most of the Mendelian results have been obtained have 

 been repeatedly described as not exhibiting a reducing division. 

 Here, therefore, is a discrepancy of which I venture to indicate 

 a possible explanation in the suggestion first made by Fick l and 

 more recently by Montgomery.- This is to the effect that in 



J Fick, R., " MittheilungueberEireifungbei Amphibien," Supp. Anat, Am., XVI. 

 2 Montgomery, T. H., Jr., loc. cit. 



