79 



weather cane grows best on clay soil; in dry seasons it succeeds best on 

 sandy soil, as has been shown by work at this station in 1888, '89, 

 ? 90, and '91. In this year canes grew much better on clay soil in the 

 mouth of June, which gave 4.83 inches of rain, and in July, which gave 

 6 inches of rain, than the canes on sandy soil. But the canes on sandy 

 soil succeeded best in August, which gave but .75 inch of rain, and in 

 September, which was also dry. Many varieties on clay soil which had 

 made unusual growth during the wet weather were injured in the dry 

 Avcjither, while the varieties on sandy soil Avere fresh and vigorous in 

 the driest weather. It would have been much more convenient if all 

 the selections of one variety had been planted in one part of one field, 

 but by so doing the general analysis of that variety would have been 

 higher or lower, as the conditions in that place were more favorable or 

 unfavorable than average conditions. 



Many selections from the best varieties were planted. For instance, 

 forty-five lots of the Collier variety were planted, including two 1-acre 

 lots. The object of planting so many lots of one variety with seeds 

 from selected canes, whose juices had high percentages of sugar, was 

 to secure seed from the most successful lot. All canes which have rich 

 juice have not the power of producing canes of similar quality. When 

 one such cane has been found an improvement is made at once in that 

 variety. For instance, Folger's Early, in 1888 and 1889, was a 10 per 

 cent variety. Its juice contained on an average 10 per cent of sugar. 

 By a single selection of seed from a superior cane, made in 1888 and 

 planted in 1889, that variety became a 12 per cent variety, and in the 

 three years succeeding the selection its descendants have retained its 

 quality. In 1888 Planter's Friend was a 12 per cent variety. Thirty se- 

 lections of canes having 14 to 15 per cent of sugar were made in that 

 year. But one of these selections produced canes which retained that 

 percentage of sugar. The descendants of the successful selection have 

 retained its quality in three years succeeding the selection. By plant- 

 ing the seeds from many superior canes of one variety in as many sep- 

 arate lets, and by selecting seed from the best canes which grow in the 

 best one of these lots, it is believed that a variety may be quickly and 

 permanently improved. The planting oi the experimental field was 

 planned on this basis. 



The lots planted with seeds from a single cane were planted by hand. 

 The acre lots were planted with a drill. As was observed in 1888, '90, 

 and '91 machine planting was best. The seeds germinated sooner and 

 more evenly. In a dry spring hand planting is unsatisfactory. There 

 seems, however, to be no better way to plant small lots with seeds of 

 varying size. None of the lots planted by hand required thinning, for 

 the right number of seeds were dropped. On the acre lots the machine 

 was regulated to plant 2 pounds of seed per acre. Nearly all of these 

 lots required to be thinned. Two pounds of sorghum seed is a rather in- 

 4efinite quantity, so f ar as number of seeds is concerned, for there are from 



