78 PROCEEDINGS O'F THE 



altogether too scanty. Mutation by loss of a factor is of sub- 

 sidiary importance to my theory ; mutation by addition of a 

 factor would be fatal to my conception that species are stable and 

 reproduce themselves unchanged so long as no cross interferes. 

 Let us therefore consider first 



Mutation hy Addition of a Factor. 



AYe all know that de Vries tliinks he has good proof for the 

 existence of this in the case of O'Jenothera LamarcJciana. There is 

 no need to give a resume of these universally known investigations. 

 It suffices to remind you that CEnothera Lamarchiana self-fertilized 

 throws plants unlike itself, the mutants of de Vries. Several of 

 these breed triie from their first appearance, others do not. 

 This, of course, reminds one of the behaviour of a hybrid indi- 

 vidual, although the proportions in which the aberrant individuals 

 appear are such that we cannot bring this throwing in direct line 

 with Mendelian segregation. Bateson and others, among whom 

 myself, have therefore long ago suggested that (E. Lamarc/ciana 

 might be a hybrid. I yet think so, and am greatly strengthened 

 in this opinion by the results obtained by Davis after crossing 

 Giaotliera grandijiora and biennis. It is true that he did not 

 obtain a plant quite identical with Lamarclciana, but as he says 

 " there is, I believe, no important character of taxonomic value 

 presented by Lamarclciana .... that has not appeared in some of 

 my hybrids. I have, as it were, surrounded the group of biotypes 

 which we call the species LamarcMana with a circle of hybrids 

 that in various characters agree with the plants that have come 

 down to us through the cultures of de Vries." 



This, I think, is all that can be expected. To re-obtain a 

 particular form from the hetex'ogeueous progeny of a hybrid will 

 always be a question of luck. 



De A-^ries offers, in notes in his last book, some criticisms against 

 Davis's view ; one is that the Oenothera grandijiora used by Davis 

 is no pure species, but a hybrid with (E. Traceyi. This, of course, 

 does not better the case for (£'. Lamarchiana ; if true, it would 

 be the result of two crosses instead of one. Another is that 

 mutation is going on already in (E. biennis, and thus is transferred 

 to its hybrids with grandijiora. This objection would be of weight 

 if it were proven that mutation occurred in the descendance of 

 pure individuals of CE. biennis. Such proof unfortunately does not 

 exist *. 



A third objection is based on circumstantial evidence. De Vries 

 says that we know of no species hybrid behaving like (E. Lam- 

 archiana. This is true, but, probably owing only to lack of time 

 to find and then investigate througii many years such a hybrid, 

 at least I am now in possession of one which might easily be 

 mistaken for it. 



It is the hybrid between Nicotiana panicidata and N. rustica. 



* Neither is it furnisLiecl, to my way of thinking, by the rer^ent publication 

 of Stomps in the Bar. Deutsdi. bot. Ges. xxxi. (1913) 166-172. 



