A. St Clair Caporn 243 



Group C. Forms throwing tight-containers, hardbacks, and penulti- 

 looses. 



Group D. Forms throwing tight-containers and hardbacks. 



Group E. Forms throwing hardbacks, penulti-looses, and pure 

 looses. 



Group F. Forms throwing penulti-looses and pure looses. 



Very likely more exist. A pure line of pure looses could probably 

 be isolated from No. 5/8/1, and Nos. 5/23 and 5^/19 of group D may 

 conceivably serve as a basis for obtaining tight-containers which breed 

 pure. Very likely, too, the lack of uniformity in the large groups may 

 signify a mixture of different lesser groups. But so far as the experi- 

 ment has at present been carried, and with a limited amount of data 

 available, nothing beyond this one important empirical result has come 

 into prominence ; namely, that irrespective of the actual proportions of 

 the offspring-constituents in individual members, certain groups broadly 

 similar in the nature of their progeny can be established. 



Sjnkelets. 



One of the main objects for which the crosses were made was to 

 ascertain whether it would be possible to transfer the many-flowered 

 habit of the Avena nuda spikelet to the tight-grained forms. A type 

 with a combination of these two characters would plainly be of pre- 

 eminent agricultural value; for not only would it yield a very heavy 

 crop, but the grain would also not be liable to shake out. Occasionally 

 seedsmen have put on the market new varieties which have been 

 extracted from Avena nuda crosses, and for which it has been claimed 

 that the slightly higher yield they gave was due to this combination ; 

 but the observations set forth in this paper, dealing with some thousands 

 of plants of all sorts, do not suj)port these contentions. Far from doing 

 so : they tend rather to foster the idea that the presence of these two 

 features fully developed in one and the same plant is physiologically 

 impossible. 



The many-flowered spikelet appears to be a function of the mem- 

 branous palea. In "nuda" forms the spikelet with 6 — 10 flowers is 

 found right through the panicle. As soon as tight grains occur in it, 

 however, as in the Fi generation, the multiflority is partly suppressed, 

 the reduction always taking place in those spikelets bearing the tight 

 grains. When one comes to the pure tight type extracted from the 

 F^ generation, the maximum number of grains per spikelet ever found 

 is four, and that very uncommonly. Only a few of the spikelets on the 



