398 THE WHEAT PLANT 



by Backhouse with a dark pubescent glumed variety of T. turgidum. 

 The average glume-length of the polonicum parent was 28-29 mm -> that 

 of the turgidum 1 1 mm. 



The F x was intermediate in length of chaff, varying from 14 to 17 mm. 

 The F 2 generation consisted of a series of plants impossible to separate 

 by eye into long-, intermediate-, and short-glumed classes. The frequency 

 distribution of measurements of the glume-lengths, however, when 

 plotted appeared to give a ratio of 514 long and intermediate : 178 short. 



The Fj generation had felted white or faintly coloured glumes. 



In F 2 were pubescent and smooth-glumed segregates, and white, 

 ringed, and deeply coloured ears, the most highly felted and deepest 

 colour being associated with short glumes, the factor for long glume 

 apparently inhibiting colour and pubescence. 



e. T. polonicum x T. vulgare. The F 1 of this cross obtained by 

 Tschermak had empty glumes intermediate in length between those of 

 the two parents. 



In the F 2 generation polonicums, intermediates and vulgare forms with 

 glumes somewhat longer than those of the vulgare parent were obtained, 

 and in addition a small number (about 6 in 100) of new durum forms 

 appeared. 



Neither the polonicum nor vulgare forms of this nor subsequent genera- 

 tions were exactly like the parents of the cross. 



The proportion of the several forms were : vulgar es, i ; durum, 3 ; 

 intermediates , 8 ; polonicums , 4 ; the ratio of the vulgare forms to the 

 rest being i : 15, suggesting the presence of two pairs of factors. 



Some of the durum forms were found to be constant, others segregated 

 in the ratio 3 durum : i vulgare. 



From the foregoing records of the hybridisation of T. polonicum with 

 other wheats it is seen that the F x is always intermediate between the 

 two parents in length of ear, form, size, and texture of glume, and form 

 and size of grain. 



In the majority of cases segregation into polonicum, intermediate and 

 the second grandparental types occurs in the F 2 generation, the ratio 

 being i : 2 : i . 



In the natural cross between T. polonicum and T. dicoccum I found 

 the homozygotes readily distinguishable from the intermediate hetero- 

 zygotes, but where the second parent is a durum or turgidum it is practi- 

 cally impossible to classify the F 2 generation into three groups by eye 

 inspection alone, since the ears obviously long-glumed and short-glumed 

 are always connected by a closely graduated series of intermediates. 



In the latter cases, however, the curve obtained by plotting the fre- 

 quency distributions of the measurements of the glumes is tri-modal, 

 corresponding to the three types of segregates. 



