226 THE AMERICAN NATURALIST [Vol.XLIV 



suppose that marked variation in the composition, and 

 consequently in the functions, of the chromosomes may 

 have occurred as a result of evolutionary changes. It 

 may be that some of the substances with which we have 

 dealt are produced only by particular chromosomes, but 

 we can hardly decide this question with the evidence at 

 hand. In any case, particular chromosomes must be re- 

 sponsible for the deficiencies of the substances in order to 

 get Mendelian phenomena. As previously stated, my 

 object is simply to show that it is possible to make as- 

 sumptions that will give us Mendelian phenomena without 

 unit characters. The assumptions made seem to be con- 

 sistent with the facts of physiological chemistry. 



Even if we assume that the various substances with 

 which we have been dealing are each produced by a single 

 pair of chromosomes rather than by all the organs of the 

 cell we can not look upon the chromosomes as the heredi- 

 tary units spoken of by the de Vriesians and Weismann- 

 ians, for we here look upon each chromosome as play- 

 ing possibly an important part in the development of 

 every feature of the organism. 



The fact that in some species the number of pairs of 

 Mendelian characters found is greater than the number 

 of chromosomes is not an argument against the validity 

 of the assumptions I have made. In Table IV we have 

 assumed that the same pair of chromosomes may be re- 

 sponsible in one race for the deficiency of a certain sub- 

 stance and in another race for the deficiency of another 

 substance. It happens in this case that the two sub- 

 stances concerned are both necessary to a particular reac- 

 tion. If each chromosome play its part in the develop- 

 ment of all parts of the organism, we might, in different 

 races of a species, have a good many Mendelian characters 

 dependent upon the same set of chromosomes, and in- 

 stances of this kind are not wanting. For instance, Bate- 

 son found a particular flower color and a particular shape 

 of the standard in sweet peas to be alternative to each 

 other in inheritance. In other words, in the terms of our 

 theory we have here two pairs of Mendelian characters 



