164 MUTATION. 



thera hybrids are perpetuated if they are self-fertilized, which 

 process must be the same as self-fertilization in so-called pure 

 species, does not admit of segregation in heterozygotes. We 

 may say that in Oenothera the segregation which we see in all 

 other organisms if they are beterozygous, is suspended. We do 

 iot know what causes underly this process, but we must re- 

 member that anything which disturbs it, must set free this 

 Mendelian segregation which was only suspended by it. We saw 

 that in several pigeons, where all the females are heterozygous 

 for a gene indispensable for pigmentation, and where no albinos 

 are formed because some unexplained cause prevents the for- 

 mation of male-producing eggs without the gene, the sponta- 

 neous production of an albino in pure-bred coloured-stock 

 might not be called mutation. In the same way, now we know 

 that something suspends segregation in obviously heterozy- 

 gous Oenotheras, we may not call by the name of mutation the 

 result of a segregation, which was only suspended for genera- 

 tions by a cause of which we know very little. 



The assumption that Oenothera lamarckiana was of hybrid 

 origin, is in accord with its probable production as a hybrid in 

 the Paris botancal gardens. It does not grow in a wild state 

 anywhere, and for several years Botanists have in vain been 

 exploring the Southern States of America for it. The fact de 

 Vries finally sowed the seeds of Oenothera lamarckiana in the 

 United States in likely places, and thus made impossible fur- 

 ther search for possible wild habitats, sufficiently shows how 

 even he has given up all hope of proving it to be an original 

 wild species. 



Now we have to go one step further, and discuss the possi- 

 bility of the production of a new, dominant form within a pure- 

 breeding strain of Oenothera, without the intervention of a 

 mutation. 



It is very clear from all the work of de Vries and Shull and 

 Davis with reciprocal hybrids in Oenothera, that in some of 

 these plants there is some mechanism which links the ovule 

 furnished by the mother to the ovules produced by the daugh- 



