Modification of Unit Characters 109 



selection has modified steadily and permanently a 

 character unmistakably behaving as a simple Mendelian 

 unit." The importance of this conclusion is evident. 

 Mendelism has been based upon the conception that 

 unit characters carmot be modified. Mendelians have 

 granted only two possible methods for the origin of new 

 races: (i) by recombinations of existing characters by 

 hybridizing; (2) by the sudden and complete dropping 

 out of an existing unit or the equally sudden addition of 

 a new unit, both of- which possibilities may arise from 

 mutation. No Mendelians will grant, however, the 

 possibility of modifying an existing unit character, the 

 thing which Castle claims to have done, and bases his 

 claim upon well controlled experimental breeding. If 

 Castle's contention is true it must result in the fun- 

 damental modification of Mendel's law. The whole 

 mechanism will have to be modified or new fields of 

 variation not known to exist will have to be taken into 

 account. 



The statements of the Mendelians in reference to this 

 situation should be considered. Their explanation is 

 based in effect upon the situation we have already 

 developed in connection with cumulative factors. The 

 claim is made that Castle started with a pecuUar 

 character which fluctuates continually and has never 

 been brought to as small a variabiUty as have most 

 other characters. The question is raised whether 

 Castle's assumption that this variability is merely due 

 to fluctuation is altogether justified. May not the 

 variability be due to varying doses of cumulative 

 factors? Suppose for the moment that this were the 

 case; it would not be surprising that Castle could 



