[SAUNDERS] INHERITANCE IN BARLEY 21 



of the doubtful cases were assigned to type I. Further, it was often 

 extremely difîfîcult to differentiate type II from type IV, because 

 frequently the hoods of the latter were not actually developed, 

 though perhaps one or two slight indications of hoods might have 

 been found by examining all the heads on the plant. When the 

 classification was made it was supposed that any six-row plant which 

 did not plainly show hoods on the median rows must be of the Arling- 

 ton type; but this supposition was later shown to be incorrect. No 

 doubt many plants were put down as belonging to type IV which 

 should have been classed as type II. 



We have, therefore, a satisfactory explanation of the abnormally 

 large numbers obtained for types I and IV. If we subtract from these 

 figures the excess over the theoretical number 381 we have 135 plants 

 to transfer to type II, making the total of this latter 1132 plants 

 instead of 1143 which would be predicted. This is a sufïïciently close 

 agreement to be acceptable. 



A few additional observations should be made in regard to the 

 inheritance of hoods in the group of plants we are now considering. 

 Plants belonging to type I were very seldom abnormal. They usually 

 carried well developed, unmistakable hoods, though sometimes these 

 hoods were borne on awns of varying length. Plants of type II, 

 however, showed great variations. As the Arlington barley has no 

 awns on the lateral rows, and as the awns on the median rows are 

 not fully developed, one naturally looked for a corresponding con- 

 dition when a hooded type, otherwise similar to the Arlington, was 

 produced. Many of the plants fell in exactly with the preconceived 

 ideas, that is to say, they had no hoods at all on the lateral rows but 

 carried imperfectly developed hoods on the median rows. Very 

 rarely, plants were found which showed traces of hoods on the lateral 

 rows in addition to very evident hoods on the median rows. But 

 many very curious, exceptional plants were observed in which, 

 though they did not show the normal development of awns for the 

 Arlington type, the expected traces of hoods were lacking. At first 

 these practically awnless plants were classified as of the Arlington 

 type — ^for the degree of development of the awns in that type varies 

 considerably. When their progeny were studied, however, occasional 

 plants were discovered which carried a few abortive hoods, while 

 others showed a moderate development of awns — quite different 

 from the condition of the parent. Clearly there was a splitting up 

 into types II and IV. The parent plants in these cases must, there- 

 fore, have been really of type 11(B), in spite of the absence of hoods. 

 Long study of these plants revealed many different degrees of develop- 



