18 



THE DIFFERENTIATION OF SPECIES. 



Morpho- 

 logical 

 species. 



Steinheirn 

 fossils. 



Transmuta- 

 tion of 

 species. 



Siberian 

 species of 

 Willow- 

 Wren. 



without the necessity of passing through the intermediate stages of childhood. There is 

 no hard-and-fast line between so-called (one might say pedantically so-called) morphological 

 species and physiological species. The degree of sterility between species varies indefinitely 

 from absolute sterility to nearly absolute fertility. There are cases in which the fertility 

 increases for some stages of the differentiation, as if Nature was making a desperate struggle 

 to prevent the formation of the inevitable new species. There is not a shred of evidence in 

 support of the theory that any physiological species exists that has not passed through a 

 series of stages in which it was, to use the scientific cant of the day, only a morphological 

 species. 



The history of the Steinheim fossil shells (Planorbis multiformis) is a most convincing 

 proof that differentiation cannot take place without isolation, though evolution may go on 

 to an indefinite extent. It also proves that variation is not fortuitous, but only occurs to 

 an extent sufficient to produce evolution when some adequate cause presents itself, and on 

 such occasions affects a large proportion of individuals, so that it is the exceptional indi- 

 viduals which do not vary that are swamped by interbreeding. For anything we know to 

 the contrary, the " variation-periods " may have been glacial epochs ; but, be that as it may, 

 we may be absolutely certain that they had a cause, and cannot possibly be ascribed to 

 accident. In this little, lake-basin we have the most interesting record of, 1st, a species 

 remaining constant until the need of variation arose ; 2nd, variation occurring simultaneously 

 in a majority of the individuals composing it ; 3rd, the "swamping effects" of interbreeding 

 upon the minority; 4th, the evolution of a new species, which remained constant as before, 

 until a fresh need of variation arose ; and, 5th, the impossibility of differentiating a species 

 into two species without the aid of isolation. 



Transmutation may go on to any extent, without geographical isolation ; for example, 

 a preglacial species may differ specifically from its postglacial descendants. But even 

 transmutation cannot take place without isolation of some kind. The specific differences 

 which are. produced by the accumulation of a long series of successive slight variations 

 protected by isolation of space, we call the result of differentiation ; those produced, in the 

 same way by isolation of time, we call the result of transmutation. 



If we could see the preglacial ancestors of the Willow-Wrens which breed in the 

 Yenesay valley and winter in India, we should probably find that they were specifically 

 distinct from all their descendants. But there are four other species of Willow-Wren 

 breeding in the same valley. How can we account for their differentiation? Where were 

 they isolated ? And when ? The answer appears to be easy enough. They were isolated 

 during the last glacial epoch in their present, winter-quarters, where they were compelled 

 for the time being to breed in a climate which was made, for the time being, suitable for 

 their purpose, by the same causes which made their present breeding-grounds for the time 

 being unsuitable. P/ijjlloscopus trochilus was possibly isolated in the valley of the Nile ; 

 P. tristis in Turkestan and Baloochistan ; P.fuscatus in India ; P. superciliosus in Burma ; 

 and P. borealis in the Malay Archipelago, 



