3oO The DextroHe-Laevuluse Ratio in the Ma/ii/old 



In any particular case, one is entirely ignorant as to the nature of these; 

 if it be assumed that they consist of arabinosc and xylose, whilst it is 

 possible to introduce a fairly accurate correction for the reducing power, 

 owing to the fact that the reducing power of arabinose is nearly identical 

 with that of xylose (Daish [1914]), this is not the case for the rotatory 

 power as [a]^ has widely different values for the two pentoses (for 

 arabinose [a]f = + 122°, for xylose [a]f = + 18-78°). This large 

 disparity in the specific rotatory powers may, in certain cases, involve 

 a diil'erencc of 0-1° or more in the rotation from which the dextrose and 

 laevulose are -calculated, according as the pentose is assumed to be 

 arabinose or xylose respectively; in the example given, showing our 

 method of calculation (see p. 317), the difference is only 0-041 , but it 

 is frequently much greater and then represents quite a large proportion 

 of the actual angle used in calculating the reducing sugars. On this 

 account, in default of information as to the exact nature of the pentoses 

 present, we have always calculated the dextrose and laevulose on the 

 two assumptions: (1) that the pentose is arabinose, (2) that the pentose 

 is xylose. But it is quite possible that the pentose really present may 

 consist to a greater or less extent of one of the less known pentoses, 

 e.g. (^ribose, and if this is the case the results for dextrose and laevulose 

 will be correspondingly at fault. 



Dextrose and laevulose are, moreover, calculated from values 

 obtained after allowing for all the other substances present — cane sugar, 

 pentoses, maltose (if present). The degree of accuracy obtained will 

 naturally depend on the accuracy with which the other constituents 

 have been estimated. Even the difference caused by calculating the 

 small proportion of pentose as arabinose or as xylose may, as for example 



in the mangold leaf, 5 p.m., October 11th, make the ratio ,, which 



ajjpears to be strictly 1-00 when the pentose is taken as xylose, have 

 a very different value (0-844) when the pentose is assumed to be 

 arabinose. 



In putting forward the results given in this paper, I am conscious 

 that the values given as dextrose and laevulose probably do not, 

 in most cases, represent real values; they are therefore designated 

 "apparent dextrose" and "apparent laevulose." Although little con- 

 fidence can be placed in them as absolute values for these sugars, they 

 show a regular variation which is sufficiently striking to justify detailed 

 consideration. This variation may be due either to a real variation of 

 the dextrose and laevulose or, what is more probable, to a regular 



