22 THE ROYAL SOCIETY OF CANADA 



ment of the hoods. At the one end of the series are found plants in 

 which the hoods on the median rows, though somewhat rudimentary, 

 are perfectly distinct. Next come plants with only a few rudimentary 

 hoods (perhaps half a dozen to a head) then others with only one or 

 two to a head, then others with only one on every second or third 

 head, then others where on the several heads of one plant only one 

 slight indication of an abortive hood is found (often merely a little 

 sideways bending of the tip of the very short awn) and, finally, 

 we have those where not a single indication of a hood can be dis- 

 covered, but where an occasional abortive hood can be found among 

 the progeny. In none of these cases is the normal, partial develop- 

 ment of the awns, characteristic of the median rows of the Arlington 

 barley, observed. These plants must be regarded as potentially 

 hooded, even though all trace of hoods be absent. 



One other interesting fact remains to be mentioned in this 

 connection. In some cases, where no indications of hoods could be 

 found on any of the rows of kernels, clearly defined, though rudi- 

 mentary, hoods appeared at the tips of some of the basal bracts. 

 In rare instances there were traces of hoods on the two median rows 

 and also, at the same time, on the basal bracts; but as a rule the 

 presence of hoods in the one place seemed to exclude them from the 

 other. Some two-row hooded plants (type I) were found which 

 carried traces of hoods on the basal bracts as well, but none of the 

 two-row plants ever lacked well developed hoods on the median 

 rows. The "potentially hooded" condition was observed in six-row 

 plants only. It is worthy of note that as a rule the plants of which 

 the progeny were studied transmitted their own particular degree of 

 development of awns or hoods to most of their descendants, though 

 occasional variations were observed. 



Plates C and D show typical heads of the four types produced 

 when seed from No. 908 and other plants of the first generation from 

 the same parentage was sown. The photographs also show a number 

 of exceptional heads, the progeny of which had to be studied before 

 trustworthy classification could be made. 



903G. A study of the progeny of this plant showed that it 

 belongs to type 11(C), as it is heterozygous for rows but homozygous 

 for hoods. It lacks many kernels in the lateral rows and must be 

 considered as really an intermediate. It shows rather more of a six- 

 row than of a two-row development. 



908P. This plant was shown by its progeny to belong to type 

 11(D). The four lateral rows are not well filled. This plant pro- 

 duced progeny of all four types. 



