292 Morgan, Are the Germ-Cells of Mendelian Hybrid? "Pure"? 



cases recorded, where the pure dominants äre not pure, but show 

 fcraces of the recessive character. This is paraticularly the case for 

 poultry, as shown by the results of Bateson, Hurst, and Daven- 

 port. We can see how cases of this sort may shade off into cases 

 of blended inheritance, so that jt may become extremly difficult 

 in practice to pick out the three Mendelian types. In fact, on my 

 view that neither of the two extracted forms are really pure, the 

 results depend largely on the relative strengths of the two con- 

 trasted characters, and the Mendelian cases do not in principle 

 differ from cases of blended inheritance, except in so far as the 

 contrast is more or less sharp. On general grounds this seems to 

 me to be a distinet advantage in favour of my view. On the cur- 

 rent view of pure gametes there is in principle a hard and fast line 

 drawn between blended inheritance and Mendelian inheritance, but 

 even the Mendelian literature does not always bear out this Inter- 

 pretation. 



Let us next examine certain theoretical possibilities for the 

 extracted recessives, (G)W. If a white mouse of this kind is 

 crossed with a gray, the young will all be gray. They will be 

 (G)WG, and their germ-cells G(WG) and W(GG) ; and since in each 

 there is a free character, they will, if inbred, give the Mendelian 

 prnportion again, unless the addition of the latent G produces some 

 influence on the result, so that more gray germ-cells are likely to 

 be formed. If this oecurred it would increase the proportion of 

 gray offspring in the next generation giving more grays thari is 

 expected on the Mendelian scheine. If this should prove to be the 

 case, could we by sucessive back crosses between the extracted 

 recessive and dominant parent alter the Mendelian proportion for 

 this groüp? Improbable as the result may seem it may be worth 

 while testing; for, as yet we do not know how a latent quality 

 like that of the gray in an extracted recessive behaves when more 

 of the same color is added; whether, for instance, it still remaius 

 in its latent condition, or simply unites with the added color of 

 the same kind. Whether union of a free and a latent color means 

 addition of these colors we do not at present know. These and 

 other questions must be determined by further experiment. 



While it has been shown experimentally that the extracted 

 recessive contains the dominant character in a latent form, it has 

 not been shown, except in Darbishire's experiment, that the ex- 

 tracted dominant contains the recessive character in the latent form, 

 which must be the case if my view is correct. It is evident why 

 this difference between the two cases should exist, because the latent 

 color of the extracted recessive can be demonstrated in one 

 generation, while it requires two generations to demonstrate the 

 Latent color of the extracted dominant. 



