26 



Journal of Agricultural Research voi. xviii, No. i 



The latter is probably due to at least three causes: (i) Respiration, (2) 

 conversion into sucrose, and (3) conversion to a slight extent into starch. 

 The changes in the dextrose-levulose ratio in the juice during growth and 

 in the various joints of the cane (fig. 16) are in the reverse order. 

 Thus, from the standpoint of maturity, the lower joints contain the most 

 dextrose while the older plants contain the least. No explanation is 

 offered for this apparent anomaly. 



Another significant fact brought out by the curves is the undiminished 

 upward trend of the sucrose curve clear to the last stage analyzed. In 

 this stage the plant was mature, judged by the condition of the seeds, 



Fig. 18.— Comparison of Early Amber sorghum grown in Minnesota with that grown in the District ol 



Columbia. 



since the latter were hard and starchy and showed fair germination. It 

 is apparent, however, that the changes in composition of the juice had 

 not ceased. If the taking of later samples had not been prevented by 

 frost, there is no doubt that the sucrose content would have undergone 

 still further increases. In figure 1 8 there is clearly shown a continuation of 

 increases in total sugars, dry matter, and purity in CoUier's Virginia-grown 

 cane during a period of 10 to 20 days after apparent maturity. Sorghum 

 cane grown as far north as Minnesota probably never reaches the maxi- 

 mum possible content of sugars. The data in the curves are for 1914, 

 which was a dry and hot season during the period of maturation of sor- 

 ghum. Although this plant can withstand these cUmatic conditions 



