W. O. Backhouse 129 



longs, there existed some which, though they could not show it, were, 

 in fact, homozygous for colour, five long glumed individuals were crossed 

 with the short glumed Kubanka and the first generation plants grown 

 this year (1916). There is no need to wait for the second generation 

 for the results. Two individuals gave all tinged, one gave all white, 

 and two gave a mixture of tinged and whites. There can be no doubt, 

 therefore, that when the second generation is groAvn and true shorts 

 appear, there will also appear fully coloured individuals. 



Only two families of long x short have been described but they are 

 typical of no less than seven second generations grown, all of which 

 show that the long glume in- wheat behaves as an inhibitor which, in 

 extreme cases, is as complete as though it were a case of genetic repul- 

 sion. There remains only one thing more to note — that from such long 

 and short glumed crosses it is possible to isolate a number of pure lines, 

 each with a different average length and breeding perfectly true to its 

 particular length. 



It is possible that the gi-eater variation in length of glume among 

 the polonicums is merely due to the effect of magnified small differences, 

 these escaping observation in the ordinary short glumed wheats. If a 

 curve be plotted of the variation in short glumed wheats, this is always 

 steep and acute ; long glumed varieties, on the other hand, however often 

 reduced to single plant cultures and so purified, always give a long low 

 curve. The heterozygote curve is more or less intermediate in shape 

 between the two parents'. 



Should it be the case that the apparent greater variation in glume 

 length is only due to a magnifying effect of the extra long glume, then 

 it seems possible to regard T. polonicum, which has always been con- 

 sidered a good sub-species in wheat, as merely a number of variations of 

 T. durum, differing from the short glumed type in one single unit 

 character which makes the long glume. Thus, had the existence of 

 black and fully pubescent polonicums been possible, a separate sub- 

 species would possibly never have been created. 



Further experiments with T. polonicum throw some light on the 

 strange behaviour of Polish Rivet crosses 2. 



I have been able to examine some of Professor R. H. Biflfen's material 

 and am familiar with the behaviour of this cross in England. When 

 leaving for this country, I took with me, in the form of grain, hybrids 



1 R. H. BiffeD, Journal of Agricultural Science, Vol. i. Part 1. 



^ " Suppression of Characters on Crossing," R. H. Biffen, Journal of Genetics, Vol. v. 

 Part 4. 



