548 Mr. H. Seebohm on the 



Crow at least, the two forms had been long enough sep'arated 

 for the intermediate forms to have been absorbed by inter- 

 breeding, or eliminated by sexual or natural selection, and 

 afterwards reproduced by the interbreeding of the extreme 

 forms wherever the geographical areas of their respective 

 distributions again met, not only in the valley of the Yenesay, 

 but also in the valley of the Elbe and in the highlands of 

 Scotland. 



We must, however, look upon the example of the Crows as 

 an exceptional case. Interbreeding seldom takes place be- 

 tween the extreme forms, because they are too widely sepa- 

 rated geographically. The intermediate forms occupy the 

 intermediate localities, and were probably the original race, 

 which has spread in different localities and has had to struggle 

 with different difficulties, and has consequently developed in 

 different directions, but not to such a degree as to prevent 

 the individuals of each valley breeding with their immediate 

 neighbours ; so that a complete series from one extreme to 

 the other is obtainable, though, as in the case of the Shrikes, 

 the two extremes have become so widely separated that when 

 they have subsequently remigrated into the same locality they 

 remain distinct, having lost the power, or at least the will, to 

 interbreed. 



The case of the Shrikes may be given as a typical example 

 of incipient species, of imperfectly segregated species, of 

 species in the process of formation, of conspecies, of sub- 

 species, or by whatever name ornithologists may agree in 

 future to call the phenomenon the great fact lying at the 

 bottom of it all, and explaining it all, being that inter- 

 breeding takes place. We must, however, bear in mind that 

 there is no hard and fast line between a specific difference 

 and a difference which is only subspecific. Two forms may 

 have become so widely separated that interbreeding between 

 them has become physically impossible, or they may have 

 become sufficiently separated to cause the cross to be a barren 

 hybrid, or the produce may only be less fertile than usual, or 

 no perceptible decrease of fertility may be observable. The 

 practical result is that slight subspecific variations are con- 



