.lamuu'y, 1921. 



SCIENTIFIC AGRICULTURE 



27 



between the ghimes as is normally the ca.se in grasses, 

 but i^emainecl inside. Secondly, an examination of 

 the anthers showed that they were very mudh smal- 

 ler than the anthers of some other species of Agro- 

 pynim w,hieh apparently were normally wind-polli- 

 nated. These peculiarities nat^irally caused a closer 

 investigation of the flowering process. Watching 

 the blossoming with tihe aid of hnis, it was observed 

 that when the flowering ghimes were in the act of 

 diverging from each other, the anthers wjiich mostly 

 were to be found surprisingly far down in the 

 "flower" suddenly exjjloded, bursting ojien from one 

 end to the otiier, and sjjrinkling the pollen all over 

 inside the flowering glumes. At the time of the ex- ' 

 plosion of the anthers the stigmas of the pistil were 

 enclosed wiUhin the flowering glumes and conse- 

 (luently, they caught a substantial portion of the 

 releaMcd pollen. When tIhe glumes closed again, after 

 tile pollination was over, two of tlie three exploded 

 anlliers in a flower were generally found hidden at 

 t/ie bottom of the glumes. Sometimes all three were 

 found in ithat pasition but more often one of them 

 was found squeezed in between the edges of the 

 glumes, protruding more or less but never hanging 

 outside as in most other gi-asses. 



I'''igure 3. — Western Rye Grass; liiKh-yicliliiiK varieties. 



It was further observed that tjie flowering process, 

 i.e., tihe opening up of ithe flowering glumes, the 

 .sprinkling of pollen on the stigmas of the pistil 

 throngli the explosion of the anthers, and the clos- 

 ing of the glumes again, took less time in plants 

 growing in the open' than in plants growing in the 

 shade of trees, and also that tIhe opening up of the 

 glumes was still more rapid in heads which had been 

 CAit off. The pollination process was studied on quite 

 a large number of the forms occurring in the Ed- 

 monton district and it did not take very long before the 

 writer was satisfied that, in the first place, self-pollina- 

 tion always took place, and, secondly, tjiiat the dhan- 

 ces for ithe pistil of a flower being fertilized by 

 pollen either from another flower on the same plant 

 or from a flower on another plant were extremely 

 small. 



After these observations it was of course only 

 natural to isuspeet ithat the various fonns of Western 

 Rye grass might normally breed true to type. In 

 order to test the possible truth of this a.ssumption, 

 seed was collected, at Calgary, Alta., in tllie fall of 



1913, from nine individiial plants representing nine 

 widely different types. The nine plants were all 

 growing toget^ier, as a matter of fact so close to- 

 gether that a pei-son could gather seed of three or 

 four of them without moving. The seed collected 

 from each j)lant was of coui-se kept separate and the 

 following spring sown in separate rows at the Cen- 

 tral Expei-imental Farm. Ottawa. The year after 

 seeding all doubt as to the breeding true to type of 

 tIhe different forms was removed. There was not t^e 

 slightest variation observed in the progeny from any 

 of the nine motJier plants, every row being perfectly 

 uniform in type (Fig. 1). Corroborating observa- 

 tions proving that the Western Rye grass forms 

 breed true to type have since been made on material 

 eolleeted in various parts of Canada. 



It is obvious that, with the numerous forms of 

 Western Rye grass breeding true to type, the devel- 

 ojHnent of new agricultural varieties passessing con- 

 stant characteristics is a very simple matter. All 

 one has to do, to start with at least, is to collect 

 promising forms, eitlher in cultivated fields or in the 

 wild state, and propagate them by seed. No isola- 

 tion is necessary and, this being the ease, a very 

 large number of types can be worked with simui- 

 taneously. At present there are over one hundred 

 different types, or varieties if you like, grown at the 

 Central Experimental Farm at Ottawa. While most 

 of these varieties are grown for the pnrpase of de- 

 mon.strating and testing tjie constancy of the dif- 

 ferent types and wlhile some of them are decidedly 

 inferior agriculturally (Fig. 2) tJiere are others 

 which are promising to become of great agrienltiiral 

 value (Figs. 3 and 4). 



3. Alfalfa.— In regard to fertilization, alfalfa pre- 

 sents some analogies to open-fertilized grasses in as 

 much as fertilization may be effected by pollen from 

 the same flower, by pollen from another flower of 

 the same plant, or by pollen from a flower of another 

 plant. This being the case the met,liods employed to 

 develop unifonn alfalfa varieties breeding true to 

 tyi)e are, in their principles, es.sentially the same as 

 those used for the development of con.stant varieties 

 of open-fertilized gras.ses. As, however, the agencies 

 which bring about the fertilization of alfalfa luider 

 natural conditions, viz: insects or mechanical impact, 

 are different, the breeder's pollination manipula- 

 tions must lie modified accordingly. As, furtju-r- 

 more, pollen can be applied to the stigma of the pistil 

 only once, on account of the weU-known tripping 

 peculiarity of the flowers, special care must be 

 exercised by the breeder in the application of pollen 

 to the pistil. 



Even if every possible measure has been taken to 

 secure pollination, the formation of seed may, how- 

 ever, often be found to be far from satisfactory. 



Two main factors may account for imi)erfect seed 

 formation, viz : 



1. — Defective sexual organs. 



2. — Unsuitable condition for seed development. 



Concerning defective .sexual organs it is well known 

 to all who have had occasion to observe different 

 alfalfa types that some types exist which under nat- 

 ural conditions regularly produce an abundance of 

 seed while others, which may grow alongside, have 

 practically no seed at all. This is particularly the case 

 among types that are segregations from crosses be- 

 tween the true alfalfa {Medkago sativa) and Yellow 



