[SAUNDERS] INHERITANCE IN BARLEY 19 



as possible, with the knowledge available at the time, and the following 

 figures were obtained: 



Type I, 420 plants. 



Type II, 997 plants. 



Type III, 127 plants. 



Type IV, 477 plants. 



From a Mendelian point of view these figures seem inexplicable 

 at first sight. Careful study of several generations of the progeny of 

 these plants, however, gave some light on the problem. When seed 

 was sown from plants of each of the four types, as described above, 

 the following results were obtained : 



Type I was found to be fixed so far as the number of rows was 

 concerned, but, as was to be expected, it sometimes proved hetero- 

 zygous in regard to awns. Some plants produced uniform progeny, 

 while others gave groups in which representatives of type III occurred. 

 In many cases a few kernels were found in the four lateral rows but 

 not enough as a rule to cause confusion as to the type of the plant. 

 Yet in at least two instances plants of six-row type were found among 

 the progeny of a plant which had been assigned to this group, proving 

 the dilîficulty of accurate classification. 



Type II. Some plants bred true, others produced types II and 

 IV, others gave types II and I, and others produced all four types. 

 It is clear, therefore, that in the first classification we had grouped 

 under the one type not only the true six-row plants, but also those 

 which might be described as pseudo six-row — like the original plants 

 of the first generation. These very often show an almost perfect 

 development of the six rows. Evidently they are really intermediates. 



Type III bred true always. When extra kernels were present 

 in the lateral rows the progeny usually showed about the same number 

 of these. 



Type IV either bred true or else produced types IV and III. 

 Further study revealed the fact that this type, like type II, consisted 

 of true six-row plants (Arlington type) and pseudo six-row plants 

 which often very closely resembled the Arlington type but which 

 never showed full development of the kernels in the lateral rows. 



The difference between the perfectly developed six-row condition 

 and that which was somewhat incomplete (as shown by some of the 

 plants grouped under types II and IV) was often so slight that it 

 escaped notice when the original cla'ssification was made. That it 

 was a vital difference was, however, clearly demonstrated in succeeding 



