242 Inheritance of Tight and Loose Paleae in Oats 



but in 1917, when whole plants were used, the exact type was always 

 noted. Hence the indication in the table of only some of the parents. 



Groupings have been made ; but they are not meant to show the 

 identity of the members in more than one broad, fundamental character. 

 Thus in group A, where some parents Avere tight-containers and others 

 hardbacks, the different progenies have one common feature in that 

 they cover the whole range of the four " nuda " forms. The total for 

 the group is 87 tight-containers, 144 hardbacks, 85 penulti-looses, and 41 

 pure looses. The fact that the tight-containers make up almost exactly 

 one-quarter (expectation is 89) is of interest. On the other hand it 

 must be recognised that there is some lack of uniformity in this first 

 section of the table. This is even more marked in the next section, in 

 which two families each lacking penulti-looses have been placed. It is 

 very possible that No. 5^4/1 may really belong to group A, and that 

 No. ^I^j'2,, with its huge preponderance of tight-containers, should go to 

 group D, especially as the five pure looses were in this case wretchedly 

 small plants all under two feet high, and with only a few spikelets each 

 on most. 



In the following set of seven, however, a moderate degree of uni- 

 formity exists, though the two known parents were of distinctly unlike 

 kinds. Here pure looses are not found, but on a rough approximation 

 the tight-containers again form a quarter of the total (expectation 

 47). 



The plants in Group D threw neither penulti-looses nor pure looses. 

 The proportion of hardbacks in the individual families fluctuated con- 

 siderably ; yet it is noteworthy that all the classified parents were of 

 the same sort. It was also remarked that among the offspring tight- 

 containers the tendency to produce a large relative number of pure 

 tight grains was much greater than in any other group. Numerous 

 plants had from 35 — 45 "/^ of pure tight grains. Taken as a whole the 

 group exhibits a 3 : 1 ratio. 



Tight-containers were entirely absent from the last two parts of 

 the table. In E the hardbacks form a very uniform quarter of the 

 progeny, while in F, which has even less variety, the same is true of 

 the penulti-looses. Group E consists of 1916 sowings only and un- 

 fortunately nothing definite can be said about the parents. 



Five distinct breeding systems, therefore, have been recognised 

 among the " nuda " types : 



Group A. Forms throwing tight-containers, hardbacks, penulti- 

 looses, and pure looses. 



