E. R. Saunders 163 



Evidence from further Crossbreeding. 



As has been Stated above it was already established that full hoari- 

 ness is dominant to half-hoariness, but the gametic relations of H and J, 

 the two differentiating factors, were not fully disclosed as no large F^ 

 generation from a cross between these two forms had then been raised. 

 We now have records of such F^ families and they show us at once that 

 in respect of H and J we have to deal with an allelomorphic pair. For 

 among a total of 319 plants composing 7 F2 families from matings of 

 semi-incana and incana 246 were fully and 73 half-hoary. (See Experi- 

 ment 1 in concluding Table.) From the fact that the F2 generation 

 contains only fully hoary and half-hoary individuals and that these two 

 forms occur in the proportion of about 3 : 1, it is evident that H and J 

 behave as allelomorphs, and that every gamete of the F^ individual 

 contains either one or other factor and that none contain both. We 

 have in fact the simple ratio that we expect when the parents differ in 

 respect of one factor only. Were H and / related as distinct factors 

 in the sense in which H and K are distinct we should expect the Fz 



generation to contain a proportion of quarter-hoary [ pnTT- ) ^^^ 



glabrous [pnj^] individuals in addition to the two parental types. 



Their absence is proof of the alternative distribution of H and / among 

 the gametes. Behaving as unequally active forms of the same factor 

 such a pair may be likened to the allotropic forms of the chemist. 

 Results in other matings confirm this conclusion. Where a coloured 

 glabrous strain containing H is bred with semi-incana (CRH x CRJK) F^, 

 as before, is fully hoary. In F2, in accordance with expectation, we find 

 fully hoary, glabrous, and partially hoary plants, but among the latter 

 there are no quarter-hoary, thus proving that here also every" gamete 

 must be carrying one or other of the two factors H and J; otherwise 

 gametic unions in the form CRJK x GRK would certainly occur and 

 give rise to the quarter-hoary grade. In order that this relation may 

 be plain to the eye an emendation of the original factorial formula is 

 desirable. By employing the symbol H^ in place of J we can display 

 the allelomorphic character of the relation between H and the factor 

 which replaces it in the several partially hoary grades and this sub- 

 stitution will henceforward be adopted, and has already been introduced 

 in anticipation in the scheme on p. 162. We shall therefore graphically 



