84 'WHAT IS A SPECIES?' 



this male was mated with a Blue Homer hen, which 

 produced healthy offspring. 1 



A comparison between the difficulty of producing such 

 a cross and that of obtaining hybrids between the Ring 

 Dove and the Rock Pigeon, the ancestor of the domestic 

 breeds, would probably throw much light on the Pallasian 

 hypothesis. 



If the view here proposed be sound — that Syngamy 

 lies behind, and is at least provisionally implied in the 

 transition which means so much to the systematist, and is 

 his only real evidence when the structural test breaks 

 down, the conclusion is suggested that the real inter- 

 specific barrier is not sterility but Asyngamy. Neverthe- 

 less, as argued on pp. 80, 81, Asyngamy will infallibly 

 lead to sterility, although the result may be long delayed. 

 This latter view, which was that of Darwin, is the exact 

 opposite of the ' physiological selection ' of Romanes, 

 in which sterility is supposed to arise spontaneously, 

 Asyngamy being not the cause, but the consequence. 



Asyngamy as a consequence of Asympatry. 



Asyngamy may be brought about in various ways, 

 of which the most obvious is geographical separation. 

 But Asyngamy is by no means the necessary result of 

 geographical discontinuity or Asympatry. Thus Darwin 

 considered that there is regular interbreeding between 

 Madeiran and continental birds of the same species. He 

 wrote to Hooker, August 8, [1866]. 'I do not think 

 it a mystery that birds have not been modified in Madeira. 

 Pray look at p. 422 of Origin [ed. hi]. You would not 

 think it a mystery if you had seen the long lists which 

 I have (somewhere) of the birds annually blown, even in 

 flocks, to Madeira. The crossed stock would be the 

 more vigorous.' 8 An even more striking case is that 



1 The Zoologist, November, 1903, p. 401. 



3 More Letters, vol. i, pp. 487, 488, Letter 370. Dr. Karl Jordan has 

 maintained (' Der Gegensatz zvvischen geographischer und nichtgeo- 

 graphischer Variation', Zeitschr. /. Wissenschaft. Zool. Bd. lxxxiii) that 

 this argument is valueless, inasmuch as more critical study of extensive 



