1912. 



PUBLIC DOCUMENT — No. 31. 



23 



ditions possible as to space for the plants and soil, and given 

 good care as before. The grower should study them carefully 

 during the growing period and make a note of the rows showing 

 desirable characters. As yield is a most important character, 

 it may be well to provide for a careful record of the yield of 

 equal length of row of the different selections. In the fall the 

 seed from the most successful rows should be saved as before. 

 If one cares to take the trouble to select from the Letter plants 

 within the row it will do no harm, but it seems doubtful if it 

 will do any good, and it will delay the increase of the stock to 

 a point when it is sufficient for the general crop. The superior 

 rows chosen will each yield seed enough, after rejecting the 

 imperfect seeds by sifting or other convenient method, to plant 

 a longer row or a small plot, thus providing for the growing of 

 the selected strains in greater quantity and ena])ling one to 

 judge their comparative value better. This operation may be 

 repeated a third time, and this should result in the final selec- 

 tion of the best that the original selection contained. 



Fig. 1 will serve to make this method of selection perfectly 

 clear. Twelve plants are chosen from a field plot and sown the 



Selection First Year 



Second Year 



Third Year 





•mtci'^^^t'^-^ ^'0<^S^ — 



« «o i» s ^ 



Fig. 1. 



first year in 12 rows. The ones with the heavy -faced figures 

 prove the better and are sown the second year in small plots. 

 One of these is rejected and 5 are retained for a third year. 

 Of these number 7 proves the most desirable and is retained 

 for field sowing. If possible a larger number of plants than 



