THE EVOLUTION OF LIVING BEINGS. 53 



inexact as the conception species used to be, and such 

 inexactness is unconsciously carried forward even 

 when, in itself exact, methods gradually evolve. 



So it has even tainted Mendelism, for strictly spea- 

 king, mendelian behaviour has nothing to do with 

 heredity and to speak of Mendelian heredity is, au fond 

 — I beg Mr. Bateson's pardon — nonsense. 



If an organism shows mendelian segregation, it 

 shows no heredity of the character-complex it posses- 

 ses, but just the contrary: disintegration of this com- 

 plex and distribution of the vestiges of it over diffe- 

 rent individuals e. g. non-transmittability of this 

 complex. 



The problem of the origin of species is not the pro- 

 blem how a species is perpetuated — how its charac- 

 ters are inherited by its offspring — but is the problem 

 how a species (or a pair of them) can give rise to a species 

 (or to more species) differing from it (or from them). 

 This problem has been solved by Mendel who showed, 

 that by mating gametes of different constitution — 

 brought together by crossing different species — zygo- 

 tes are formed from which individuals arise, able to 

 form a number of gametes of different constitution, 

 from whose matings new species can arise. 



By crossing two monogametic individuals of diffe- 

 rent constitution, one consequently obtains a poly- 

 gametic hybrid which is the source — and up to the 

 present the only known source — of the origin of new 

 forms, some of which are heterozygotes, others of 

 which are homozygotes e. g. new species. 



By isolation of such homozygotes in the experiment 



