W. A. Davis 



345 



laevulose follows more or less closely the general course of the curve of 

 total hexoses (calculated as invert sugar, from reduction values only). 

 But there is no real significance in this because with the dextrose values 

 apparently constant, the laevulose figures, which are also calculated 

 from the same reduction values, necessarily follow the figures for total 

 hexoses (at any moment D + L = total hexoses). 



Ssj 



Fig. 4. Mangold stalks, tops, Series I, Aug. 26-27, 1913. 



Exactly the same kind of relation can be traced between the 

 fluctuation of the apparent hexoses and the values of A in the bottom 

 halves of the stalks (Fig. 5). In this case, however, both the dextrose 

 and laevulose appear to undergo wide variations. From 6 a.m. to noon 

 the dextrose increases absolutely faster than the laevulose (from 8-55 

 to Il-O per cent, for dextrose compared with a change from 0-2 to 1-8 

 per cent, for laevulose) although relatively the dextrose does not increase 



so rapidly as the laevulose, as shown by the fall of y from 37-2 to 6-0. 



