W. A. Davis 329 



mid-ribs and stalks (where the ratio , is often extremely high), whereas 



in the fotato the reverse holds, the laevulose apparently predominating 

 as in the cases studied by Brown and Morris, and by Parkin. At the 

 same time it can be shown that the apparent excess of dextrose or 

 laevulose is correlated with certain abnormahties in the cane sugar 

 estimations, caused by the presence of optically active impurities. 

 The apparent excess of dextrose in the tissues of certain plants (sugar 

 beet and mangold) is indeed due to the presence of a dextro-rotatory 

 impurity (possibly glutamine), whilst the predominance of laevulose in 

 other plants (e.g. tropseoluni, snowdrop, potato) is to be attributed 

 to a laevo-rotatory impurity (e.g. asparagine). 



In the mangold the difference between the results obtained for 

 saccharose by the reduction method and by the double polarisation 

 method, which we have referred to in the preceding paper (p. 273), is 

 always far greater in the stalks and mid-ribs than in the leaves, a fact 

 which we attribute to the accumulation of optically active impurities 

 in these parts. Side by side with this we have the fact that, whilst in 



the leaves the ratio of dextrose to laevulose [ ^ j is in general not very 



far removed from unity, in the stalks and mid-ribs the ratio j is very 



much greater, generally varying from 2-5 to 10. This ratio, too, is far 

 higher in the bottom halves of the stalks than in the top halves, pointing 

 to an accumulation in the lower part of the stalks of the dextro-rotatory 

 impurity. Striking differences are also found between the results for 

 cane sugar in the top and bottom halves, according as they are calculated 

 from the reduction data or from the polarimetric values. Thus in the 

 top halves the results obtained by polarisation may be 80 to 90 per cent. 

 lower than the values obtained by reduction, whilst in the bottom halves 

 they are high by 40 per cent. As the day proceeds, the relation of tops 

 and bottoms may be reversed, the impurity which was in the top half 

 passing down to the lower part of the stalks (compare the values at 

 noon, 6 p.m. and midnight given on p. 344, Table VIII). 



Independently of any error which may be caused by the improper 

 use of basic lead acetate (see p. 270), a difficulty which makes it impos- 

 sible to obtain really accurate values of the proportion of dextrose and 

 laevulose hes in the fact that allowance has to be made in the calculation 

 for the reducing power and rotation of the pentoses which are invariably 

 present in the alcoholic extracts prepared by our method of working. 



