584 Journal of Agricultural Research voi. xix. No. n 



Group I is the only group offering any complications, and these are of 

 no particular complexity. If the 2-factor hypothesis is tenable, group i 

 must include three different classes of 6-rowed forms, two of which are 

 homozygous. If the factor A A is considered epistatic to the factor BB, 

 the three classes of group i become phenotypic for 6-rowed and may 

 be considered as a single class. In genetic constitution, however, the 

 homozygous 6-rowed segregate AAbb differs from the parent 6-rowed 

 AABB; and, inasmuch as the BB factor has been lost, the form AAbb 

 is referred to here as a regressive 6-rowed. If this regressive 6-rowed 

 AAbb were crossed on the Svanhals parent aabb, there would be no 

 possibility of securing the homozygous intermediate aaBB. Heterozy- 

 gous types corresponding to such a cross probably are found in group 4, 

 which is supposed to be heterozygous for this regressive 6-rowed AAbb 

 X the 2 -rowed aabb. On the other hand, group 3 is heterozygous for 

 the parent 6-rowed AABB X the 2 -rowed aabb. 



It will be seen in Table II that the plants obtained in each group 

 came very close to the expectancy. The numbers of 6-rowed and 

 2 -rowed forms produced coincide exactly with the expected numbers,, 

 while the number of fixed intermediates is only two greater than in the 

 calculated ratio. 



The nature of the factors is better discussed in connection with the 

 description of the genetic classes which is given in Table II. It will 

 be seen from the table that except in group i these various classes differ 

 in appearance as well as in genetic constitution. In some instances the 

 separations are easily made on appearance alone, while in others only 

 a part of the plants belonging to a group can be easily identified. The 

 first four groups have lateral florets, the lemmas of which bear awn 

 points or awns of varying lengths. Group i, of course, presents no 

 difficulties, for 6-rowed segregates are unmistakable. Group 2 is almost 

 as definite, for there are no other high-fertility segregates with small- 

 kerneled, awn-pointed lateral florets. Group 3 is readily separated from 

 group 2 by the lower fertility of the individuals in group 3. The sterile 

 individuals of group 3, however, are easily confused with the individ- 

 uals of group 4. Probably 80 per cent of such plants in group 3 can 

 be identified by the more obtuse lemma tip on the lateral florets. Groups 

 5, 6, and 7 differ from the other groups in the absence of awns on the 

 rounded tips of the lateral florets. These conditions are not absolute, 

 in that an occasional floret may possess an awn. In this case, however, 

 other florets on the same spike have the characteristic rounded obtuse 

 tips. Group 5 is characterized by the possession of lateral florets which, 

 even when sterile, are larger than those of the 2-rowed. Under field 

 conditions higher fertility probably would have been present in the 

 members of this group. When the Fg material was being studied, it 

 was possible, however, to isolate the sterile spikes of intermedium forms 

 by inspection. In group 6 the lateral florets of many individuals are 



