482 [June, 1912. 



"The plots with their yields were as follows :— 



Plot Distance. Number of Plants Yield in Bushels 



per Acre. per Acre. 



Bunches 6x6... — ... 82 



A 4 x 4 ... 392,040 ... 37 



B 6 x 6 ... 174,240 ... 60 



O 8x8 ... 98,010 .. 60 



D 10 x 10 ... 62,726 ... 52 



E 12 x 12 ... 43,560 ... 18 



" From each of the transplanted plots 100 plants were separately gathered, 

 the number of tillers or fruiting stalks sent up by each plant was separately count- 

 ed, and tne grain from each plant was separately weighed. I had therefore 500 

 definite observations of weight to select from. The results of these operations 

 were as follows :- 



Number of Plants having different Weights of Grain in Grammes. 

 Plut ... 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 



A 4 x 4 ... 21 24 25 15 10 4 1 



B 6 x 6 ... 5 18 22 20 24 5 5 1 



C 8 x 8 .... 3 9 10 12 20 16 15 7 4 1 1 . 1 



D 10 x 10 2 9 5 9 14 16 14 7 6 5 6 2 4. . . . . 1 



E 12 x 12 ... 3 16 16 21 8 10 8 10 6 . . 2 . 



Total ... 29 61 74 75 59 48 44 41 28 11 7 8 6 3 4 1 



"The weights in the above table are given, as they were recorded, in grammes, 

 28 of which go to an ounce. 



Attention is specially directed to the single plant of plot D which yielded 42 

 grammes of grain as against an average of 16 grammes from the remainder of the 

 plot. The hope was entertained that a specially good yield might be obtained from 

 the progeny of this plant, and it will be found that this was not disappointed. 



Number of Tillers. 



Plot ... ... 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 



A 4 x 4 ... ... 2 19 33 18 20 6 1 . 1 



B (5 x 6 ... ... 6 19 35 24 11 3 2 



C 8 x 8 ... .... 4 9 28 21 14 18 5 . 1 



D 10 x 10 ... .... 1 8 6 14 13 18 20 7 3 5 3 1 . .1 



E 12 x 12 ... ... 2 13 15 20 21 6 9 9 2 2 .1 



Total ... 10 56 100 96 87 42 48 38 10 6 5 4 1 . . 1 



" It will be seen at once that both the number of tillers and the weight of 

 grain depend a good deal on the distance of transplanting. Each plot, however, 

 generally contains one or two plants which are notably superior to the remain- 

 der, and by sowing separately the seed gathered from these we may hope to see 

 some definite improvement in future generations." 



The seed from the two best plants in each plot was selected for further 

 sowing, that is to say, from the two plants which produced the greatest weight of 

 seed. The seed from each plant was broadcasted in a separate isolated patch in 

 October, 1911, and the seedlings were transplanted a month later into the same five 

 plots as before. In order to minimize error the sequence of plots was reversed, that 

 is to say the offspring of the best plants of plot A were transplanted in plot E and 

 those from plot B in plot D. The distance of transplanting was one foot apart in 



