E. R. Saunders 149 



In L'VL'iy Ciise such iiiutings gave a propurtiuii uf hoaiy and smooth 

 in F., approximating to 911 : TG. From these results it is clear that 

 this commercial material consisted of pure (homozygous) forms and, 

 further, that in the glabrous strains one was dealing solely with 3- factor 

 forms (CRH, CRK, CHK and RHK). Starting with this material 

 however it should be possible, according to the view here taken and in 

 the absence of any indication of further complications due to coupling 

 by appropriate matings, to eliminate successively all the different 

 factors and to obtain the several simpler forms containing only 2 factors, 

 1 factor, or even none. From the new forms thus obtained a new set of 

 ratios would of course result. We may state the expectation for the 

 various cases in general terms thus : 



When a hoary plant, however produced, is heterozygous in regard to 

 any of the four factors 0, R, H, K, it will yield a mixed offspring in 

 which hoary individuals will predominate, or glabrous, according as the 

 parent is heterozygous in one or two factors only, or in more than two. 

 The ratio in the several cases will be as follows : 



3H : IG if the parent is heterozygous in 1 of the i factors. Excess of hoary 



^H: IG „ „ 2 



21 H : dlG „ ,, 3 ,, Excess of glabrous 



8m:17SG „ „ 4 



In the earlier accounts of this work abundant evidence has been 

 given of the occurrence of the ratios ^H : IG and 9H :7GK It remained 

 to obtain similar proof that by a proper sequence of matings the simpler 

 forms, presupposed by theory, could be obtained from the more 

 complex : in other words to procure evidence of the actual occurrence 

 of the ratios 27H :37G and 81^: l75(r. 



This has now been done. F.^ families bred from gi-andparents 

 containing the required combination of factors have given excess of 

 glabrous individuals in the expected ratios, as shown by the results 

 given below. 



A. Cases where the expectation is 27 H : BIG. 



' See Reports I, II and III. 



