W. A. Davis and G. C. Sawyer 379 



opposite direction. The fluctuations, whatever be their cause, show 

 throughout evidences of periodicity; this appears most clearly in the 

 shape of the curve of apparent laevulose. 



II. Stalks. 



The results for the potato stalks closely resemble those found for 

 the mangold stalks in the fact that the dextrose present appears always 



to be in large excess as compared with the laevulose ; the ratio y varies 



from 3-39 to 4-58 (pentose as xylose). But there is this striking differ- 

 ence : in the mangold, dextrose appeared to be the predominant sugar 

 in both leaf and stalks, but in the potato it is in excess only in the stalks, 

 whilst in the leaf, as pointed out above, laevulose predominates. It is also 

 very striking, that whereas in the mangold the greatest fluctuations and 

 the greatest divergences between the reduction and polarisation values 

 for saccharose were found in the stalks and mid-ribs ( A varied from + 40 

 per cent, to — 90 per cent., see Table VIII, preceding paper), caused 

 no doubt by large fluctuations in the optically active impurities present, 

 in the potato stalks the differences are as a rule relatively small, and, in 

 general, less than in the leaves. 



The following table (Table VII) shows this: 



Table VII. 



Divergence of Values of Saccharose by Polarisation and Reduction 

 Methods— Potato Stalks. July 16th-17th, 1914. 



2 a.m. 2-61 2-82 + 8-0 2-70 2-88 + 6-7 3-75 



The extreme divergence here is only 15 per cent., whilst in general 

 the divergence (A) is less than 10 per cent. There are no such abrupt 

 changes from positive values to negative values as were met with in 

 the mangold stalks (see p. 344) and had their counterpart in the sudden 

 variation in the values for apparent laevulose (see Figs. 7 and 8, Paper I). 

 One of the most striking differences between the potato stalks and the 



