SPECIES AS TO VARIATION ETC. 185 



the important announcement that, in the oak genus, 

 the best known species are just those which present the 

 greatest number of spontaneous varieties and sub-vari- 

 eties. The maximum is found in Q. Robur, with 

 twenty-eight varieties, all spontaneous. Of Q. Lusi- 

 taniea eleven varieties are enumerated, of Q. Calli- 

 prinos ten, of Q. coccifera eight, etc. And he sig- 

 nificantly adds that "these very species which offer 

 such numerous modifications are themselves ordinarily 

 surrounded by other forms, provisionally called spe- 

 cies, because of the absence of known transitions or 

 variations, but to which some of these will probably 

 have to be joined hereafter. " The inference is natu- 

 ral, if not inevitable, that the difference between such 

 species and such varieties is only one of degree, either 

 as to amount of divergence, or of hereditary fixity, or 

 as to the frequency or rarity at the present time of 

 intermediate forms. 



This brings us to the second section of De Can- 

 dolle's article, in which he passes on, from the obser- 

 vation of the present forms and affinities of cupulifer- 

 ous plants, to the consideration of their probable his- 

 tory and origin. Suffice it to say, that he frankly ac- 

 cepts the inferences derived from the whole course 

 of observation, and contemplates a probable historical 

 connection between congeneric species. He accepts 

 and, by various considerations drawn from the geo- 

 graphical distribution of European Cupuliferce, forti- 

 fies the conclusion — long ago arrived at by Edward 

 Forbes— that the present species, and even some of 

 their varieties, date back to about the close of the Ter- 

 tiary epoch, since which time they have been subject 



