Size iulieiitauce and tlie pure Hue theory. 235 



germination. Tiie pure line work with beans is accordingly at present 

 in the same position as that with Paramecium before CAiiXms and 

 Gregory undertook to verify it. It is quite possible that another 

 observer repeating it might reach very different conclusions. 



Since there exists general agreement concerning the facts of size 

 inheritance in animals and plants, it has been necessary to consider 

 only their theoretical significance. The currently accepted explanation, 

 which its supporters choose to call "Meudelian", rests upon the idea of 

 gametic purity in Mendelian crosses. It assumes that Meudelian unit 

 characters are unchangeable and unvarjdng, and that when they seem 

 to vary this is due to a modifying action of other unit characters (oi- 

 factors). It assumes further that genetic variation can occur in no 

 other way than I)y the gain oi- loss of unit characters (or factors) by 

 the germ cells. These assumptions are not an original part of Mendelism; 

 they are not found in Mendel's original papers or in the early "Reports" 

 of Bateson. They are an after-growth and if they deserve the name 

 Meudelian, it is only in the qualified form neo -Mendelian. But what 

 is more important, these basic assumptions lack any adequate experimental 

 support. The idea of unit character constancy is a pure assumption. 

 In numerous cases unit character inconstancy has been clearly shown, 

 as in the plumage and toe characters of poultry according to the obser- 

 vations of Bateson and Davenport, and the coat characters and toe 

 character of guinea-pigs in my own observations. Unit -character in- 

 constancy is the rule rather than the exception. How then can this 

 observational fact be reconciled with the idea of unit-character constancy? 

 Only by supposing that in cases of observed modification something has 

 become associated with the unit character which modifies its somatic 

 appearance. In a few cases such associated modifying factors have 

 been demonstrated. I have found them in the case of the rough coat 

 of guinea-pigs, and less certainly in the hooded coat pattern of rats. 

 These modifiers are demonstrable because detachable. But does it 

 follow that all modified unit-characters result from detachable modifiers? 

 We are not warranted in thinking so unless the supposed modifiers can 

 actually be detached and the modifications synthesized anew. But this 

 is possible iu a very small percentage of cases. As Mendelian characters 

 are being subjected to more careful and critical study, it is found that 

 the same unit character may assume a variety of forms. These are 

 now called multiple allelomorphs. One way of looking at these is 

 to consider them simple variants of a single genetic unit-character. But 

 such a view is incompatible with the idea of unit character constancy, 



