348 The Dext rose- L(ie VI dose Ratio in the MuiKjold 



likely that the actual reducing sugars in the stalk reach the root as 

 invert sugar^ and that the apparent predominance of dextrose in the 

 stalks is due solely to dextro-rotatory impurities ; the existence of these 

 is clearly proved by the enormous differences found in the cane sugar 

 estimations by the polarisation method. Until really rehable methods 

 of determining the true proportions of dextrose and laevulose have 

 been devised it is impossible to draw any further conclusions on this 

 point. 



Series II. The considerations put forward above, correlating in 

 Series I the apparent dextrose and laevulose values with the divergence 

 between the cane sugar values determined by the reduction and polarisa- 

 tion methods, hold for the stalks and mid-ribs in Series II also-. 

 The following table gives the values of A, and Figs. 9 and 10 of the 

 preceding paper show the curves for dextrose and laevulose. It will 

 be seen that, as in Series I, when the apparent laevulose increases 

 rapidly as compared with dextrose the values of A become less positive; 

 when the apparent laevulose decreases, the values become more and 

 more negative. 



Fig. 9 of the preceding paper shows the apparent variation of dextrose 

 and laevulose in the stalks as compared with that of saccharose and the 



total hexoses and the variation of the ratio ^ . As in the tops of stalks 



in Series I the apparent dextrose rises slightly during the day (10 a.m. 

 to 4 p.m.) but then remains practically constant imtil 4 a.m. next 

 morning. The laevulose rises considerably more rapidly from 10 a.m. 



' It is quite jjossible that the ratio of dextrose to hievulo.se in the mixture of sugars 

 reacliing the root is not strictly 1, owing to one of the sugars being put more under 

 contribution for purposes of growth or respiration in the leaves or stalks than the other. 

 But it is probable that the ratio is very nearly unity as is the case in the leaves (September 

 and October), when the amount of optically active impurities interfering with the 

 determination is a minimum. 



- The same principle can bo applied to the leaves of Series II and III to explaiji the 

 fluctuations of A, i.e. the difference between the results fomid for cane sugar by double 

 polarisation and by reduction, which are far less marked in the case of the leaves than 

 with mid-ribs and stalks, because the proportion of opticall3' active impurities is relatively 

 less. In practically all cases when the apparent dextrose mcreases faster than the 

 apparent laevulose, the divergence becomes increasingly positive; when the laevulose 

 increases faster than dextrose the divergence becomes more negative. As pointed out 



on p. 339, when . is unity there is the closest agreement between the results for cane sugar 



obtained by the two methods and the departure of the ratio . from 1 is probably merely 

 apparent and not real. 



