236 Inheritance of Ti<jht ami Loose Paleae in Oeits 



85 times in every 100, the possibility of the number 144 being an 

 abnormally augmented figure is not difficult to conceive. 



The likelihood of a heterozygous panicle being taken to indicate a 

 homozygous plant is well illustrated in Table II, plant D, panicle 1. 

 If such a panicle were only slightly damaged, the chances are all in 

 favour of the two loose grains with membranous paleae being the first 

 to break off, while it is also possible that the four paleae classified as 

 " intermediate " might be such near approaches to the pure tight form 

 as to be separable from the grains only by extra hard rubbing. 



A random selection of 26 taken out of the 144 apparently " pure 

 tights " yielded an F^ generation fairly representative of the whole 

 group. The plants of 22 rows were uniforndy tight grained on every 

 panicle. The remaining 4 rows showed distinct splitting. Twenty- 

 two out of twenty-six is equivalent to 122 "pure tight" out of 144 

 tentatively considered pure. This is a much better agreement with 

 the 1295 expected on a 1 to 3 basis, the corrected result for the F„ 

 generation reading : — 



Pure loose and 

 heterozygotes Pure tiglit 



Observed 396 122 



Expected 388-3 129-5 



More substantial evidence of this agreement will now be offered in 

 connection with the detailed classification of other F,. results. 



A large number of sowings was made from F.2 heads having a 

 mixed character. Double rows were planted, so that the tight grains 

 in each sowing might be placed in one section, and the loose in the 

 other. This was done in case the nature of the husk should in any 

 way be related to the zygotic constitution of the embryo, and hence 

 have some bearing on the segi-egation, though it was hardly expected 

 that maternal characters would so affect it. As a matter of fact they 

 did not. The double rows, however, if for no other purpose, were very 

 useful in that they enabled one to obtain a few more F^ plants. 



In order to determine just which among these F^ i-ows were showing 

 splitting after the manner of offspring of F^ plants, and could therefore 

 be regarded as an exact repetition' of an F« generation derived from 

 actual Fi individuals, the seed from half a dozen F^ panicles which had 

 been saved over was sown at the same time. These F^ panicles had 

 shown on rubbing out a proportion of " pure tight " grains varying 

 from one-fifth to three-fifths of the total. Allowing for a reasonable 



' In regard to the throwing of one-quarter pure tights, tliat is to say. 



