200 Rhodora [September 



while when rubrincrvis pollen is used the nanella-carrymg eggs of 

 Lamarckiana disappear. Will anyone be found willing to support 

 such an hypothesis? 



But the difficulties with a Mendelian interpretation of these crosses 

 have only begun. How are we to account for the fact that both 

 Lamarcktana and nanella from the Fi of Lamarcktana X nanella breed 

 true? On the Mendelian assumption it must be because in these 

 Lamarckiana plants the ?2rtNc//a-carrying germ cells all degenerate, 

 either in the pollen or the egg cells or both. Otherwise when selfed 

 they would produce nanella in F|. Is this degeneration a reasonable 

 assumption when we know that in rubrincrris X nanella some of the 

 rubrincrris plants appearing in Fi when selfed split out nanella in a 

 ratio which is, in some cases at least, close to 3:1? That is, we know 

 that rvbrinertu plants which are heterozygous for nanella develop 

 their two types of germ cells according to regular Mendelian expecta- 

 tion, and it would be, to say the least, highly improbable that the 

 closely related Lamarckiana would behave in an entirely different 

 manner and that its nanella germ cells (assuming that there are such) 

 would degenerate. 



Another fact which East must explain is this: Why is it that 

 Lamarckiana X nanella yields dwarfs in Fi while rubrincrris X nanella 

 only yields them in F 2 , or in other words why is it that the first result 

 is a mutation cross while the second is a Mendelian result as regards 

 the dwarf character? So far as I am aware, no Mendelian has 

 attempted to offer an explanation of this significant fact. 



There is yet another fact in this connection which has not even 

 been considered, still less explained, by the critics. This is that while 

 Lamarckiana gives rise to the mutant nanella, rvbrincrris has never 

 been known to do so in all the extensive cultures. Is it unreasonable 

 to connect these facts with those mentioned in the last paragraph? 



The mutationist conception on the other hand, while it may not 

 furnish a complete explanation, at least enables us to consider all 

 these facts under a consistent point of view and does not lead to any 

 of the absurdities which lurk in a Mendelian interpretation. More- 

 over, it offers an explanation of whole classes of facts which no 

 Mendelian writer has attempted to explain. Let us consider this 

 conception as it applies to the facts we have cited. DeVries has 

 assumed that pangens, or if you like, factors for the differences 

 between the mutants and their parent Lamarckiana, may exist in 



