151 



the task of improving the plant. The new factories have so much to 

 contend with that they can not possibly devote the necessary time and 

 expense to it. The agricultural experiment stations, in whose province 

 it would seem to fall, have been but recently established in thesorghum- 

 growing States, and are not fully equipped for such work, besides hav- 

 ing their attention taken up by other agricultural products. Yet sev- 

 eral of them have already done something in the line of sorghum im- 

 provement, and others have announced their intention of doing so. It 

 would seem to be essentially fitting and proper if the Department of 

 Agriculture were provided with authority and means for its continuance. 



Whoever it may be that undertakes the work, it is important that 

 they should have the benefit of whatever the experience of the past 

 season at this station has taught ; we think it advisable, therefore, even 

 at the risk of some repetition, to outline in a general way the principles 

 and methods to be pursued in the future conduct of such work. It must 

 be remembered, of course, that we have only the experience of one short 

 season to draw upon, and while many of our ideas are based upon that, 

 and upon analogies in beet culture, some have only the foundation of 

 our own judgment to rest upon. 



In selecting sorghum seed the following may be outlined as the gen- 

 eral course of procedure : 



1. Seed should be selected from the varieties which have proved to be 

 the best adapted to the locality. Those which are defective in any 

 respect should either be thrown out or their faults removed by such 

 crossing or selection as will have that tendency. 



2. The seed of these varieties should be selected from the individuals 

 which show the fewest faults of form, the highest content of sugar, and 

 the least content of other substances. 



3. The seed from the best individuals should receive such cultivation 

 and fertilization as may be shown by experiment to give the best results 

 in yield of sugar, in proportion to the area of soil covered. 



It may seem impossible to carry on these several lines of selection at 

 once; to select seed from the individual canes which yield most sugar, 

 and at the same time to select seed with reference to the physical char- 

 acters of the canes. But more than one point is always necessarily con- 

 sidered in all plant selections. The faults of form in the beet have been 

 bred out, merely to obtain a form to admit of ready cleansing. 



The faulty forms of the sorghum cane have already been pointed out. 

 Seed should never be saved from "tillers," or secondary canes, or sup- 

 plementary heads, as they would tend to reproduce canes which would 

 produce a second crop of seed. 



Photographs of some of the canes selected for future propagation 

 at this station will show ho\v faulty forms inherent to certain varieties 

 may be eliminated. The canes are from a plot of Link's Hybrid. This 

 variety has nearly always proved to be a good sugar producing variety, 

 and its greatest fault is one of Conn. The top joint is apt to be very long, 



