122 



The following table shows the results: 



The last test was made with a plat of Indian cane. The topping was done August 

 23, three days after the heads began to appear. 



The comparative analyses were made October G. At this date the seeds were per- 

 fectly ripe, and would drop from the head when shaken. 



The results are given in the following table: 



These results show an increase of over 3 per cent, of cane sugar in favor of the 

 topped cane. 



Dr. Collier also was led to investigate tbe same subject. 1 

 Two sets of analyses were made. In tbe first set of ninety- six pairs 

 of analyses the results are as follows : 



In the, juice. 



In tbe second set of forty-two pairs of analyses tbe results were as 

 follows : 



Dr. Collier makes tbe following observations on tbe results of tbese 

 analyses : 2 



Tbe practical conclusions from these results are, that there is no incompatibility 

 between the maximum crop of ripe seed possible, and the maximum content of sugar 

 in the juice of the stalks ; and that, owing to the more rapid development of the cane 



'Collier's Sorghum, pp. 1IW and 241; Special Report, pp. 18 ct seq. 

 *0p. cf. r p. 140. 



