W. A. Davis 339 



The closer the value , approximates to 1, the closer is the 



agreement between the cane sugar values by the two methods ; at 



4 p.m. particularly, when the value is practically 1, the divergence 



A is exceedingly small, viz. 1-5 per cent. only. It would appear that 

 at this time the amount of optically active impurities influencing the 

 results is practically 7iil. On the other hand, at 1 p.m., G p.m., and 



8 p.m., when the value of ^ departs considerably from unity, in either 



direction, there is a correspondingly large diii'erence in the results for 

 cane sugar by the two methods, the polarisation results being from 15 

 to 25 per cent. high. The figures at 6 p.m. and 8 p.m. are particularly 

 interesting because at these times there is apparently an excess of 

 laevulose^, pointing to the presence of a laevo-rotatonj, not a dextro- 

 rotatory, impurity such as was present in the earher part of the day 

 (e.g. at 1 p.m.). At the same time, however, the polarisation figures 

 are still higher than the reduction figures, showing that the change of 

 rotation which accompanies the inversion process involves probably 

 transformation of the laevo-rotatory substance into a compound with 

 a still greater laevo-rotation — such as would happen, for example, in 

 the transformation of asparagine into aspartic acid. 



It is interesting to consider the curves in Fig. 2 showing the variation 

 of the "apparent dextrose" and "apparent laevulose" during the 24 

 hours. Although these curves probably do not show the variation of 

 the true sugars so much as that of the optically active impurities, they 

 exhibit a regularity which points to the latter substances being formed 

 regularly and progressively. In general the apparent dextrose and 

 apparent laevulose increase when the cane sugar increases and fall 

 when this sugar falls. The dextrose curve rises and falls three times in 

 succession during the 24 hours, the rise and fall in the night (8 p.m. to 

 8 a.m.) being far more gradual and regular than the two abrupt changes 

 during the day. The laevulose curve is less regular, rising gradually 

 till 4 p.m., when D and L are practically ec^ual; from this point the 

 laevulose rises suddenly and follows the abrupt rise in the saccharose 

 curve, which takes place just before sunset. The latter rise is therefore 



' This apparent excess of laevulose would formerly have been explained by assuming 

 that at these times the respiratory changes (utdising dextrose) predominate over the 

 tissue-building changes. The polarisation data clearly point to abnormal quantities of 

 optically active impurities at these points, which falsify the real values of D and L. 



