I _N LOCK: STUDIES IN PLANT BREEDING 



pei cent, as eras the case with the offspring of black grains in 

 V .). black being then recessive in the case of certain grains. 

 Most usually the behaviour of black heterozygote grains 

 when crossed with a pure white strain was of the former kind : 

 whereas the offspring of white heterozygote grains, similarly 

 pollinated, showed, the reduced proportion of black grains, 

 but this was by no means always the case. 



lu Kj , it will be remembered, full dominance of black 

 was found as the result of the cross black x white, whereas, 

 in the case of the reciprocal cross (white x black), white was 

 sometimes dominant. 



In F 2 — from black F, grains (white x black) — the evi- 

 dence showed that dominance of black failed in nearly the 

 same number of cases, no matter whether the white strain 

 was the pollen or the seed parent. 



En later generations there occurred sometimes the one and 

 sometimes the other kind of behaviour, although the part of 

 pollen parent was now always taken by a pure white 

 strain. 



It did not appear that particular white strains as a whole 

 behaved in a particular way. On the contrary, it seems as 

 if the result can be better described as being due to a high 

 degree of variability in the relative dominance or " Rntfal- 

 tungsstarke " of the blue " Anlagen" in different individual 

 grains, the degree of dominance being inherited to some 

 extent. 



The phenomenon may be compared to some of those which 

 have been described by \)e Vries under the title of " ever-sport- 

 ing varieties," and in particular the cases have this much in 

 common thai the facts are derived from cultures to which 

 ewhat wholesale methods of pollination have been applied. 

 doubtful whether they can be discussed to advantage 

 at any greater length until the experiments have been repeated 

 using the more | ii method of exact pollination between 



particular individual 



