82 PROBLEMS OF EVOLUTION 



a rule: if it were a matter of vital need to the species, advantage 

 might be taken of these variations, and we should see an adapta- 

 tion. In this connection we may recall the willows, the brambles 

 and the dog-roses and the many problems they present. In the 

 Flora of Herefordshire we find thirty-one species of brier 

 (Rubus), besides many varieties. I believe that these and also 

 the many species (or they may be, more properly, mere varieties) 

 of willow and dog-rose are passing through a period of variation 

 during which they present a number of points from which natural 

 selection may some day single out one or two as serviceable 

 under circumstances that have happened to arise. Most of them 

 are very trifling points that cause the utmost perplexity to the 

 systematist. To prove this I need only tell the story how a 

 very competent botanist in his long solitary rambles elaborated a 

 practical joke : he cut two specimens from the same brier and 

 sent one, and then, after a week's interval, the other, to a pro- 

 fessor of botany who was, perhaps, the greatest living authority 

 on the Rubi. In this case, however, the oracle was at fault, for 

 it assigned them to two different species. 



The chaotic state of the brambles is, probably, largely due to 

 intercrossing. We may look upon many of the so-called species 

 as mere varieties, that have not become separated from one 

 another by inter-sterility. The fact that the differences have 

 not become more marked may be due to want of isolation and 

 consequent intercrossing. If we could take forty varieties and 

 put them each on a separate island, it is probable that the dis- 

 tinction would become more noticeable and definite, even 

 though they never assumed an adaptive character. A feature 

 common to all the individuals separated off in a particular 

 island would tend to be accentuated and, if it happened to be 

 harmless, natural selection would leave it unmolested. It is to 

 such isolation that I attribute the enormous number of species 

 representing one family of molluscs in the Sandwich Islands. 1 

 There has been through isolation a stereotyping of what were 

 before mere shifting varieties. In the same way the small lady- 



1 In the chapter on isolation I shall return to the subject of these molluscs. See 

 p. 177. 



