22 



THE DIFPEEENTIATION OF SPECIES. 



Effects of 

 Geographi- 

 cal Isola- 

 tions. 



Operations 

 of Evolu- 

 tion. 



be produced. However much a species can be changed by Evolution, " the swamping 

 eflFects of intercrossing " prevent there being at any given time more tlian one species. The 

 fact of Isolation is, however, repeatedly implied, as, for example, in the phrase " when the 

 barrier to intercrossing, instead of being physiological, is geographical." Isolation forms 

 such a very important part of the theory of Natural Selection as propounded by Darwin, 

 notably so in Wallace's vindication of the theory in ' Island Life,' that the absence of the 

 term strikes the reader as unusual. It seems as if Mr. Romanes regarded Isolation and 

 Selection as interchangeable terms. He is continually comparing " geographical barriers " 

 with "physiological barriers," but he never speaks of Geographical Selection, though 

 constantly using the terra Physiological Selection. It seems to me that the part which 

 Geographical Isolation plays in the differentiation of species is to the minutest detail strictly 

 analogous to the part which is ascribed by Mr. Romanes to Physiological Selection. If it 

 exists at all, it is neither more nor less than an additional mode of Isolation ; and if it be 

 necessary to give it a name, that name certainly ought not to be Physiological Selection, 

 but might appropriately be Piiysiological Isolation. Now we know that Geographical 

 Isolation does form a barrier to intercrossing, and that by means of this barrier Evolution 

 is able, not only to further the development of a species, but to split it in two races, which 

 develop in different directions, and in process of time become so widely differentiated that 

 they no longer interbreed, even when the geographical barrier is broken down and the 

 areas of their distribution overlap. If, however, variation were accidental, the effect of 

 Isolation, whether geographical or physiological, would only be temporary. Under a 

 regime of accidental variation Evolution has to begin de novo at every generation, and after 

 two or three generations the swamping effects of intercrossing would surely be apparent, 

 even within the isolated family. It would be necessary for Physiological Selection to step 

 in again. The six conditions must again coincide ; but even then the progress would not 

 be in the same direction as before, unless we assume a seventh coincidence, namely, that the 

 variation protected by the new Isolation was of the same kind as that which had been 

 protected by the old one. It seems to me that no selection of any kind is able to perpetuate 

 a variation, however important to the welfare of the species it may be, except when a 

 variation occnrs simultaneously in a number of individuals inhabiting the same area. The 

 modus operandi of Evolution is probably as follows : — 



In every species there is a tendency to vary in definite directions : the variations are 

 hereditary and cumulative, so that evolution goes on steadily, though slowly, from genera- 

 tion to generation, but without partial or complete isolation no second species can be 

 evolved. Variation may change a species ; hut so long as free interbreeding is possible a 

 species may advance or retrograde, but it cannot split into two. If, however, a part of the 

 species be isolatsd from the rest differentiation commences, the evolution of the two colonies 

 does not proceed exactly in the same direction, and the rapidity with which differentiation 

 takes place is exactly m proportion to the difference in the circumstances in which the two 

 colonies are placed. Wherever there is a struggle for existence (and it can only be in very 



