154 HaJf-hoariness in MattJiiola 



always in the earliest flowers. The siliqua is distinctly hairy though 

 less so than in the half-hoary type. The stem is glabrous in the 

 flowering as well as in the vegetative region. The appearance of the 

 siliqua furnishes the readiest means of recognition ; doubles are conse- 

 quently less rapidly identified, the leaves and sepals often requiring 

 close examination. Exceptionally, one may meet with quarter-hoary 

 individuals showing in some limited area or member a tetal absence of 

 hairs ; in one plant, e.g., the second siliqua in the raceme was quite 

 glabrous although all the others showed the quarter-hoary condition; 

 in another the two surfaces of one siliqua were sharply differentiated, 

 the one component carpel being characteristically hairy, the other quite 

 glabrous. Such ' mosaic ' cases are however distinctly rare. As regards 

 behiiviour and constitution the quarter-hoary individual is a heterozygote 

 incapable of breeding true since it forms no gametes corresponding to 

 the appearance of the zygote. It is comparable with the classic case 

 of the Blue Andalusian among fowls, and like its animal counter- 

 type yields always a proportion of one of each of the parental forms 

 from which it is derived to two of itself, the parental types breeding 

 true, the_ heterozygous form yielding again this same proportion in 

 succeeding generations. This behaviour is clearly proved by the results 

 recorded above in Fo (Experiments 9 and 10) and in F^ (Experi- 

 ments 12, 13, 14). In Experiment 9 where the total obtained from 

 the quarter-hoary Fi was 711 the expectation based on a ratio of 

 \G:2 1H:\^H is 



178 glabrous, 



356 quarter-hoary, 



178 half-hoary. 



The record actually made showed 



163 undoubtedly glabrous. 



70 glabrous or quarter-hoary. From these 70 plants no record was 

 obtained fnjm the calyx and ovary, it is therefore uncertain how many 

 of these originally glabrous plants really belonged to the smooth class 

 and how many should be added to the quarter-hoary. 



297 quarter-hoary. 



181 half-hoary. 

 There is little doubt that, had the 70 plants which were classified only 

 on the early vegetative character reached the flowering stage, the 

 agreement between the observed and the expected results would have 

 been extremely close, as it is in Experiment 10, despite the very small 

 number recorded. The same may be said in regard to the numbers 



