83 



lected. From these 7,827 were selected whose juices had a mean value 

 of 15.98 sucrose, 73.78 purity. From these were selected 1,658 whose 

 juices had a mean, value of 16.41 sucrose, 76.84 purity. 



In the 26,635 selections, 5,905 contained between 15 and 16 per cent 

 of sugar, 5,296 contained between 16 and 17 per cent, 2,550 contained 

 between 17 and 18 per cent, 172 contained between 18 and 19 per cent, 

 and 23 had 19 per cent and more. 



These figures serve to show the capabilities of the sorghum plant. 

 The general averages given for the varieties in four years serve to show 

 what the sorghum plant now is. 



There is a general belief that seeds from canes which have an extraor- 

 dinary percentage of sugar are most desirable for planting, with a view 

 to producing rich canes. It is perhaps well to quote the opinion of the 

 veteran seedsmen, Yilmorin, Andrieux et Cie., of Paris : 



As far as our experience goes it does not seem to be the safest way to discard merely 

 superior plants, and to give preference only to those having extraordinary quality. 

 On the contrary, it is often the case that merely superior plants aiford the surest 

 means of rapidly improving the variety. This applies especially to the most impor- 

 tant quality the capability of transmitting their good qualities to their descendants ; 

 and as this quality can not be ascertained at the outset, it is necessary not to be over- 

 severe in the hrst selection. 



If this opinion is correct, as it probably is, when endeavoring to im- 

 prove a 14 per cent variety of sorghum, it is safer to plant seeds from 

 15 per cent canes than from 19 per cent canes j that is, the lower class 

 of selections is worth more than the higher. 



In making selection of seed, canes were taken only from those varie- 

 ties which gave the best results in average analyses. The largest canes 

 were selected in the field, and the smaller canes were thrown out by a 

 second selection at the mill, so that selections of seed were made first 

 from the best varieties; second, from the largest can^s; third, from 

 canes whose juice showed in analysis an unusually high percentage of 

 sugar; fourth, from canes whose juice had exceptionally high purity. 

 Another selection, based upon early ripening canes, was also made. In 

 many of the central States there are frequent inquiries for " varieties 

 which mature not much later than Amber." It is believed that the 

 time of maturing of the varieties can be hastened by selecting seed from 

 the canes which show a good percentage sooner than average canes, 

 and that superior varieties may thus become well suited to more north- 

 ern latitudes. For instance, many single canes in the Folger variety 

 showed 15 per cent of sugar two weeks sooner than the majority of the 

 canes of that variety. In many, perhaps in most, of these cases, the 

 difference in time of maturing was only accidental, but if a single seed 

 head can be obtained from a cane which naturally matures two weeks 

 earlier than the Folger, a valuable subvariety can soon be established, 

 and a point will be made in favor of that variety. In this selection 

 2,800 seed heads were reserved from five varieties for future planting. 



In 1890 455 lots were planted in crosses in the hope of finding a new 



