W. A. Davis and G. C. Sawyer 377 



It would appear, therefore, that in the potato as in the mangold 

 leaf two oppositely active impurities are present at different times of 

 the day. During the greater part of the day laevulose ajDpears to be 

 in excess owing to a laevo-rotatory impurity predominating, but at 

 night the amount of this impurity diminishes until it is replaced just 

 before sunrise (4 a.m.) by an excess of rfea'/ro-impurity. The variation 

 is well seen by considering the data obtained by assuming the pentoses 

 to be arabinose ; the amount of laevo-rotation left unaccounted for when 

 the whole of the hexose is assumed to be laevulose gradually drops from 

 10 p.m. to 2 a.m., whilst at 4 a.m. dextrose appears to be present. 



At 10 p.m. negative rotation unaccounted for = -0-062° (200 mm. tube) 



12 raidniglit negative rotation unaccounted for = -0-010° 

 2 a.m. negative rotation unaccounted for = -0-002° 



4 a.m. positive rotation unaccounted for = +0-01.5° 



Between 6 a.m. and 8 a.m., that is just after sunrise, the qiiantity 

 of negative impurity suddenly increases very largely, the negative 

 reading unaccounted for at 6 a.m. being greater than at any other 

 period of the 24 hours (— 0'1.38° if pentose is arabinose, — 0-083° if 

 xylose). 



The following table (Table VI) shows how the presence of the 

 optically active impurities causes abnormally large differences in the 

 results found for saccharose by the reduction and by the polarisation 

 methods. This table should be compared with the similar table 

 obtained in the case of the mangold leaf (see p. 338, preceding paper). 



Table VI. 

 Divergence of Results for Saccharose by the Reduction and Polarisation 

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



