176 THE JOURNAL OP BOTANY 



By carefully watching therefore a heterozygous individual we 

 see the origin of genotypes. 



The next step is thus to produce at will these genotypes — 

 originating heterozygotes. This we can do by crossing two 

 individuals belonging to different genotypes. 



The next question is : Do all heterozygotes obtained by crossing 

 segregate and thus give rise to different genotypes ? 



Until very recently it was believed that only heterozygotes 

 obtained by crossing so-called varieties did segregate, while crosses 

 between so-called species were said to give a stable offspring. 

 Evidence is rapidly accumulating that this is not true as far as the 

 latter is concerned, that species-hybrids also segregate. 



I therefore claim that the origin of genotypes by crossing is of 

 much wider application than was formerly supposed, that perhaps 

 it is of universal application. I further claim that the genotypes 

 are the real, long sought-for units of the natural system, and I 

 propose that in future the term species be limited to them, in 

 accordance with the view held by systematists for ages that 

 " species " is the proper term for the units of the natural system. 

 [The question of Progressive Mutations, based upon Prof. H. 

 de Vries's work upon (Enothera Laviarckiana, was then discussed.] 



The chief question is whether (2/. Lamarckiana is a pure geno- 

 type, because only if the purity of type is beyond any possibility 

 of doubt is there good reason to explain the throwing of deviating 

 types as due to new formation of factors ; in any other case these 

 can be explained as the result of new combinations of factors 

 already present in the plant throwing them. Heribert Nilsson 

 has, in the opinion of the speaker, undoubtedly shown that 

 CE. Lamarckiana is no pure genotype, and consequently cannot 

 serve as a reliable basis for the study of the origin of mutants. 

 Nor did the speaker know of any other case in which progressive 

 mutation from a pure genotype has been proved. 



[Mutation by loss of Factors was next considered. It was 

 submitted that but one thing is proved, viz. ; that the real units 

 of the living kingdom are genotypes ; that such genotypes can, 

 under proper precautions, be kept pure for an indefinite time ; and 

 that there is no certain evidence that they can be changed in any 

 other way than by crossing.] 



What then is the reason of the apiparent variability of a species 

 in the Linnean sense ? In the first place, the fact is that a Linnean 

 species is a collection of independent stable Jordanian species. 

 Indeed, as Bateson says ; " Between Jordan with his 200 odd 

 species for Erophila and Grenier and Godron with one, there is 

 no hesitation possible : Jordan's view .... is at least a view of 

 natural facts, whereas the collective species is a mere abstrac- 

 tion." The Linnean species, indeed, has been a snare, and if we, 

 as Darwin did, consider it as a unit, the small species contained 

 in it must naturally appear to be deviations of the type — in other 

 words, varieties. If, then, one further sees that between Linnean 

 species which one considers to be units, transitional forms exist, 

 it is perfectly logical to build up a theory that one species can 



