Fruwirth: Selection in Pure Lines 



91 



one direction. Not one of these at- 

 tempts yielded the result hoped for. 

 The pure lines seemed to be unchange- 

 able by selection. 



EXPERIMENTS WITH OATS 



2. In oats the palea inferior of the 

 kernel at the base of the spikelet — 

 what we, following Atterberg, call the 

 aitssenkorn, or in English the outer- 

 kernel — may bear a bristle and the 

 basis of this kernel, the callus, may bear 

 hairs. Some years ago I demonstrated'' 

 that the length of the hairs (see Fig. 13) 

 and to a less extent the number of the 

 hairs, are useful characters in the classi- 

 fication of the various varieties of oats. 

 Pure lines of oats are not only charac- 

 terized by a relatively constant length 

 and nimiber of these hairs but often by 

 a relatively constant percentage of 

 haired outer-kernels per plant. 



Now the mean per cent per plant of 

 outer-kernels bearing bristles or hairs 

 is easily modified by external conditions. 

 Never have I seen a variety of oats called 

 bristled where every outer-kernel bore 

 a bristle ; nor have I ever seen a variety 

 advertised as free, from bristles where 

 a bristle could not be found on one or 

 more kernels. The same is true of the 

 hairs on the basis. The characters 

 "bristled" and "haired" can only be 

 described by giving percentages per 

 plant, the mean of a large number of 

 plants of a variety or breed or line 

 being determined. 



Just because these two characters 

 are so easily modified, I chose them for 

 a test of selection in pure lines, making 

 at the same time a test of another 

 character: the per cent of 2-kerneled 

 spikelets per plant, a so-called internal 

 character. I determined to test these 

 three characters not only in one and 

 the same year, but also by taking the 

 means for one line for a number of 

 succeeding years, thereby measuring 

 what Tower has called the place- 

 variation. The method of selection 

 which I emploj^ed is in part the same 

 as that used by Hopkins and Smith in 

 their classic selections of corn for high 

 and low protein and oil content.^ I 



' Fruwirth: Fuhling's landw. Zeitung, 56th yr., p. 298. 



^ Smith: ///. Agric. Exp. Station Bull. 128; American Breeders' .Assn. Proc, Vol. v\ (1910), p. 5. 



spotted seeds whatever failed to produce 

 a strain with unspotted seeds. 



(b) In nine generations of selection 

 with a vetch {Vicia sativa) having green 

 and creamy seeds, both, also, on the 

 same plant, I failed to obtain either full 

 heredity of the green color, or even a 

 higher percentage of green-seeded plants. 



(c) In nine generations of selection 

 with snap beans (Phaseolus vidgaris) of 

 the variety called green Chevrier, which 

 bears also on the same plant solid green 

 seeds, solid white seeds, and seeds which 

 are part green and part white, / Jailed 

 to bring the green color to full heredity 

 or even to intensify its inheritance. 



{d) In ten generations of selection of 

 a fodder pea {Pisum arvense) called in 

 the United States the Canada pea; 

 the variety being that called with us 

 the black-hulled pea — a variety with 

 }'ellomsh-green seeds with sometimes 

 a tinge of violet; I failed to get the 

 \'iolet tinge expressed over a greater 

 area of the seeds of the selected plants 

 or to be produced on all the seeds of 

 the selected plants 



(e) In six, seven and eight genera- 

 tions of selection of another lentil (Lens 

 esculenta) called the Puy lentil, bearing 

 green seeds marbled with black, the 

 marbling being sometimes so intensified 

 as to produce an almost solid black 

 seed ; / failed to get plants which would 

 transmit the full black color of seeds or 

 even an intensified marbling. On the 

 other hand, selection of plants with only 

 normally marbled seeds failed to prevent 

 seeds with intensified marking from 

 appearing from time to time. 



(/) In six generations of selection of 

 white mustard (Sinapis alba) where the 

 seeds are normally white (strictly speak- 

 ing, the color is yellow, despite the 

 name), but where some plants produce 

 brown seeds, some others yellow seeds, 

 and again others produce seeds of both 

 colors, / failed to establish full heredity 

 of either of the colors by selecting 

 plants bearing only seeds of the one 

 color. 



Here were six careful and persistent 

 attempts to change the character of a 

 pure line by selection of variations in 



