146 



These analyses show very decided differences between the two samples. 

 The conditions were in all respects similar, except as to the distance 

 apart of the canes in the row, and the large samples taken diminished 

 the possible error of sampling, so that considerable reliance may be 

 placed upon the results. It will be seen that in both experiments the 

 canes which were thin in the row were much better in quality than those 

 which stood closely together; the content of sucrose is higher, of glu- 

 cose lower, and the purity is greater. It is evident that close planting, 

 while it increases the tonnage, diminishes the yield of sugar per ton. 

 Of course there is a proper mean between too close planting on the 

 one band and too thin planting on the other, and this subject is worthy 

 of more attention arid discussion than has previously been given it. 



It is probable that the distances at which canes should be planted 

 vary to some extent with the varieties. For instance, it would seem 

 that the small canes of the Early Amber do not require so much space 

 as the much larger canes of the Honduras, and it also seems that soils 

 and climate may require the distances between the canes to vary. For 

 instance, it is well known that corn is planted much closer in the 'North, 

 than in the South. 



VI. MISCELLANEOUS EXPERIMENTS AND RESULTS OF OBSERVATIONS. 



EFFECTS OF FROST. 



The effect of a light frost upon sorghum cane has always been a 

 mooted question, some holding that it is not injured by a frost which 

 only kills the leaves, but rather has the effect of ripening the cane. It 

 seems reasonable to suppose that it does affect it unfavorably, how- 

 ever, as it kills the leaf and stops further growth and vitality in the 

 plant. The question is an important one, for it is quite common to 

 have a slight frost quite early in the season. A few observations 

 were made on this point at this station the past season. The first 

 frost occurred on the night of September 27. On October 5, about a 

 week afterwards, when the effects of the frost were plainly perceptible, 

 the different plots were examined to see if any observations of impor- 

 tance could be made. The more immature varieties seemed to have 

 resisted the action of the frost better than those which were more ma- 

 tured; the Honduras, for instance, holding the bright green of its leaves, 

 almost without exception. Some varieties appeared to have resisted 

 the action of the cold much better than others, giving some ground for 

 the hypothesis that this might prove a constant characteristic. Other 

 plots, again, showed some spots that were almost entirely untouched 

 by the frost, while in other spots the leaves were quite dead, the differ- 

 ences being doubtless due to different conditions of evaporation from 

 the soil. These plots seemed to offer an opportunity for comparative 

 analyses of frosted and unfrosted canes. Large samples were taken of 

 both kinds, taking all the care possible to have them comparable in all 



