W. A. Davis 



535 



Table VII. 

 Mangold Stalks and Mid-ribs. Series III. October llth-12th. 1912. 



Discussion of Results. 



A. Leaves. 



Series I. In Series I, Table I, the actual percentages of total hexoses 

 are very small, especially at night, during which they range from 0-90 

 to 0-20 per cent. ; consequently, even the small differences in the 

 rotation of the pentoses, according as they are assumed to be arabinose 

 or xylose, lead to considerable differences in the proportions of dextrose 

 and laevulose apparently present. Thus, for example, at 8 a.m., if the 

 pentose is assumed to be arabinose, laevulose appears to be in excess 



of the dextrose, the ratio y- being 0-83 ; if the pentose be taken as xylose, 



the dextrose appears in excess, the ratio j becoming 1-68. In this 



particular case the rotation of the pentose represented, in a 200 mm. 

 tube at 20°, as arabinose a reading of + 0-049°, as xylose + 0-009°, 

 whilst, after allowing for the saccharose and pentoses present (see method 

 of calculation, p. 317), the rotation left for the hexoses was — 0-042° or 

 - 0-002° in the two cases. 



That dextro-rotatory impurities are present in this series to an extent 

 sufficient to invahdate the calculations and so to convey a false idea of 

 the true quantities of dextrose and laevulose, is shown by considering 

 the data obtained at 6 a.m. In this case, if the whole of the reducing 

 sugar, obtained from the reduction data, is calculated as dextrose, 



Joum. of Agrio. Sci. vii 23 



