W. Nkii-son .Ionks and M. riiKVKi.v Kaynkk -211 



Experimental. 



In actual expei'iment the nine $ plants resulting from this cross 

 bore the percentages of two-carpel lary flowers shown in Table III, the 

 hypothetical constitution of each group of plants being appended in 

 each case. 



TABLE III. 



Shoioing result u/ cross fr plant {GfifiJJ.,) x /•', phmt {0\!//r.j/.,). 



Name of plaut 

 19/11/ 

 19/11 b \ 

 17/11 k ■ 

 19/11 e ) 



J 19/11. "I 

 I 19/11 a { 



{ 17/iii ) 



- 17/11 / - 

 I. 17/11 rM 



Of these four types, the only one which has appeared previously is 

 (-U()iG.i<j-,, which is the constitution assigned to the two plants n{ F^. 



In the F^ ])lants the percentages of two-carpel lary flowers lay be- 

 tween 55 per cent, and 35 per cent. 



The percentages of two-carpellary flowers on the three plants of F.^ 

 of this constitution fall well within these limits. (Table III, plants 

 17/11 j, 17/in, and 17/11 (Z.) 



Fo (6). Theoretical. The heterozygote F^ selfed should give plants 

 having the constitution shown below in the proportions stated. 



Gi!liG-,!i2 (^'i plant) x GigiG.,().i(Fi plant) 



I I ! II I ■ I 



IG1G1G2G2 : 2Gi(?iG2.'/2 : iGnjiG^O-i : ( IG, Gifl.j;/-.; : 2G,£/i.r;.j,-/.j ; 2./i</i G^.'/- ; \<.l\i!i!h'j2 



(l.7i3if'2G., 



(See also Table IV, below, 4th column.) 



Expenmental. 



The percentages of twei-cari)ellary flowers borne by the tweiit}' plants 

 of the F., generation from the cross i^, x Fi are shown in Table IV, 



14-2 



