504 Conclusion. 



pollen, new lateral branches arise from the family ; the 

 members of such lateral branches are not henceforth 

 called mutants. The independence is, however, only ex- 

 ternal, the analogous mutants being related to one an- 

 other as sisters, or as nieces and aunts, etc. The identity 

 of their features is obviously due to the fact that they 

 originate from the same latent potentialities in the main 

 stem. 



But each mutation arises suddenly and directly from 

 the main stem without any preparation and with all its 

 characters.^ Every new dwarf that appears is as small 

 as the dwarfs of the fourth and fifth and later genera- 

 tions. Each /a/a-mutant is as purely female as the lata 

 of the present day which has been cultivated for ten 

 generations. The various riibrinerzns-plsinis which I have 

 cultivated during the course of many years for various 

 purposes are indistinguishable from the newly arisen 

 mutants of these forms. 0. gigas only arose three 

 times, O. scintillans fourteen times. But each time they 

 appeared with exactly the same characters. 



The occurrence of transitional and intermediate forms 

 is a very important point. These do certainly occur; 

 but they are phenomena of individual variability and not 

 of mutability. For, in the first place, these transitions 

 do not appear before the new species, but, at most, simul- 

 taneously with it; usually, however, not till well after it 

 has arisen. These transitions are not therefore the steps 

 which must be traced by a new form in its origination; 

 this origin, far from being reached by these steps, is ab- 

 solutely independent of them. The intermediate forms 



^ These characters are therefore to be regarded for each muta- 

 tion as expressions of one single internal change. See § 13. pp. 327- 

 330. For the relative frequency of the appearance of the various 

 mutations, the so-called mutation-coefficients, see § 14, p. 337. 



