A. H. Trow 



253 



sixteenth only of the individuals constituting the whole colony will be 

 of the type xxcc. These ocxcc plants will alone be pure creams. 

 The progeny of this sixth plant is shewn below. 



1909. Fi. 



1910. Fj. Exp. 41. 



No. 368 yellow, not selfed. 



37 plants, of which 4 were NR and 2 others did 



not flower 

 6 Deducting these, there were left 



31 plants 



eUc 



24 yellows 



7 creams 



No. 13, selfed 



No. 30, selfed 



1911. Fi. Exps. 43 and 43. 90 yellows, CC. 



No. 1 selfed. Exp. 44. 



98 creams, {eexx) 



88 plants, which seemed at first all yellow. 

 I No. 96 was a typical cream, and 



close examination led to the dis- 

 I covery of three types of plant 



7 Creams. 

 (ccxx) 



Cream to Yellow, 

 the colour diflicolt 

 to estimate, but 

 certainly with some 

 cream. Xxcc 



77 Yellows, the colour 

 rather faint and some- 

 times shewing a creamy 

 tinge on the under side 

 of the corolla 



As one-sixteenth of 88 may be taken as 6, the result corresponds 

 fairly with the view that the recessive cream only appears as a definite 

 character as the result of the association of the cream factor with some 

 other recessive factor. At first, I suspected foliage-leaf colour to be 

 concerned in producing this result, but the observations so far made 

 seem to shew that cream-coloured flowers may appear on plants with 

 either dark-green or yellow-green foliage leaves. For the present we 

 must be content to assign to this unknown pair of factors the signs X 

 and X. X, when present, more or less prevents the development of the 

 cream-coloured character. 



Mutation. A Fimbriate Type. 



The 461 plants of the ^2 generation of multicaulis x erectus, radiatus 

 {Exp. 15) included one plant — No. 43 — which was not only markedly 

 unlike all the others, but unlike any other radiate groundsel that I had 

 ever seen. The ray florets, in this case, were of the usual number, but 



Journ. of Gen. 11 18 



