sept.i,i92o Fixed Intermediate, Hordeum intermedium haxtoni 585 



obviously larger than in the 2-rowed segregates, but as a rule they 

 are not so large as in group 5. Groups 6 and 7 can be accurately sep- 

 arated only by the breeding test. 



It will be seen, then, that although fertility is the basic distinction, 

 it is expressed in other ways than in the production or nonproduction 

 of kernels. Indeed, the percentage of fertility probably varies more 

 with environment than do the morphological differences. Such varia- 

 tion, however, does not imply any lack of reliabihty of the fertility 

 factors. 



According to the factor hypothesis, the first four groups have the 

 inherent possibility of producing 6-rowed segregates. Only plants 

 belonging to these four groups have lateral florets the lemmas of which 

 are awned or awn-pointed. The A factor, either as AA or Aa, is found 

 only in these groups. This factor, then, must be associated with the 

 possibility of awns. When A is homozygous, fully fertile, long-awned 

 florets which develop kernels of normal size are produced. When A is 

 heterozygous, all lateral florets are short-awned or awn-pointed. When 

 homozygous, it is epistatic to B, the spike being normal 6-rowed 

 irrespective of the condition B. When heterozygous, A has little effect 

 on fertility, and the amount of fertility present in this case is in direct 

 relation to B. The lateral florets of AaBB are quite fertile, those of 

 AaBb occasionally so, while those of Aabb as a class are sterile. Aa 

 may have a slight effect on fertility, for AaBB has a higher percentage 

 of fertility than aaBB. The factor BB, then, is a fertility factor which 

 at its highest expression produces a lower percentage of fertility than 

 AA. It is not accompanied by the presence of awns. The lateral 

 florets under its stimulation produce undersized kernels, the largest 

 form of floret seeming to be associated with AA. 



INDICATION OF A THIRD FACTOR 



The presence of a third factor of fertility was first suggested by a 

 study of the data in Table I. Column 2 of that table shows the per- 

 centage of fertility of the Fg parents. In group 2, three plants exhibited 

 a much lower fertility than the other four. vSince one of these was 

 abnormal, there were, excluding this plant, two low-fertility and four 

 high-fertility plants. It is difficult to measure the inheritance of these 

 variations in this group. The percentage of fertility of the intermedium 

 segregates in the F3 generation of four lines was studied. The F3 inter- 

 medium segregates of the others were missing. Three high-fertility 

 lines gave almost identical fertility of 37+ per cent. One low-fertilfty 

 line gave 29.5 per cent of fertility. In group 3, five out of twenty-five 

 plants showed some fertility. The F3 plants were not studied for differ- 

 ences in percentage of fertility. 



In group 5, two out of seven plants had fertile lateral florets. In 

 this group the progeny were studied carefully. This is the only group 



