On the Hybridising of Ducks. 257 



true, yet the "power to vary " does not seem to have been in 

 any way overcome, so that variations, showing either con- 

 siderable departures from the normal, or only slight aberra- 

 tions, are continually making their appearance even among 

 wild species, and are only seeking for favourable conditions 

 in order to become permanent. 1 



The fixity of a species is, therefore, largely, if not entirely, 

 dependent on the power of natural selection to restrict, for 

 it cannot annihilate, variation. 



If then variation can make itself felt in the face of natural 

 selection, how much more is it likely to do so when the fixity 

 ■of a species is artificially disturbed by hybridisation. And 

 since it is generally allowed, as has been stated above, that 

 species represent suitable varieties, which have become fixed 

 and perfected through many generations, we should expect, 

 when the restrictions of natural selection have been partially 

 removed by hybridisation, to find varieties appearing which 

 in some characters will resemble, or tend to resemble, species 

 other than their parents. But at the same time, if my 

 argument holds good, we should also expect to find that 

 varieties would appear that resemble no known species, these 

 last representing the unsuitable variations that have been 

 eliminated by natural selection. In point of fact my experi- 

 ments show that this is precisely what we do find. The birds 

 noted under " Eeversion " (ante p. 255) are instances of 

 successful varieties that are now found perfected as species in 

 various parts of the globe ; while, as instances of unsuccessful 

 varieties which have reappeared, we may note the rumps of 

 the hybrid Mallard-Pintail, the bill of the tetragen M.P.S.Z., 

 and perhaps, the best example, the third generation, light 

 trigen (PI. xi.). 



Consequently, it seems to me that there is considerable 

 probability that these generally accepted instances of 



1 It would be out of place to dilate here on these varieties that are continually 

 cropping up in Nature, but I may mention one or two examples. I have in my 

 collection a Teal, showing a tendency to the white wing of the Mallard ; Eider 

 Drakes occasionally show the black V characteristic of the Pacific species ; in 

 the pattern of its feathers '"Sabine's Snipe" shows certain affinities with the 

 Great or Solitary Snipe. These are all instances of varieties tending towards 

 other species, while as an instance of a variety resembling no known species, 

 I may mention the case of Athene cMadriœ, a variety of Athene noctua, recently 

 described from Italy. 



