ON VAIUETIEP, KACEB, SUIj-SI-KCUlS, ASIJ 61>liClES. ^5 



cultivate the white and yellow rooted specimens, as their taste 

 was less aromatic. 



We see then how tlie season of sowing influences plants 

 obtained from tlie seeds of the wild carrot, and how the modifica- 

 tions resulting therefrom are not common to all such plants , this 

 must be attributed to the fact that the circumstances under which 

 vegetation took place were different, or that the seeds themselves 

 were different, or to both these facts combined; and if the latter, 

 the second of the above facts would have a greater infiueuce than 

 the first, as we shall hereafter see. 



M. Vilmorin's researches on the modifications of the wild 

 carrot aflbrd a good example of the great influence a scientific 

 method of cultivation has on the researches into the proximate 

 causes of vegetable modifications, and they at the same time show 

 the possibility of solving the questions relating to the type to 

 which modified individuals belong, and which would always have 

 remained in obscurity had it not been for the powerful aid of 

 experiment. 



La fine, in the species to which the distinction Gamma is 

 applicable, there are individual types and varieties capable of 

 propagation by seed, and that with sufficient certainty to form 

 tvell-dejined varieties and even races. 



Distinction Betainlus Gamma, 



This applies to species in wliich there are types, and in 

 addition — 



1. Varieties which can be propagated by division, but not by 

 seed. 



2. Varieties which can be propagated by seed with sufficient 

 certainty to be defined and even to form races. 



If varieties of Robinia pseud-acacia had been really capable 

 of being reproduced by sowing, then the species Robinia w^ould 

 have afforded an example of the distinction Beta plus Gamm,a. 



Distinction Delta. 



It is very far from true that the origin of the different groups 

 of individuals belonging to the same species is as well known in 

 the greater number of cases as they are with respect to the 

 species to which tlie distinctions Beta and Gamma apply. There 

 may then be species which present two or more groups of indi- 

 viduals, constituting two or more varieties, witiiout it being 

 possible to consider one of these groups as formed of individual 

 types of the species. We may then correctly say that there is 

 no known individual type of tliis species, inasmucli as the notion 

 of the latter includes characters which are common to distinct 



