222 
Ingvar Jorgensen and Walter Stiles. 
the second series of these workers, measured a fortnight later in 
the season, the proportion of pentosan is exactly the same as at the 
earlier period, while the proportion of matter insoluble in alcohol 
has actually decreased from about 60% to 50%. Nor is there any 
better evidence for the second alternative, for at this intermediate 
stage, where the measurements were taken before and after sunrise, 
the results obtained show actually a slight increase of these 
substances after sunrise. The most reasonable explanation is 
therefore that the differences recorded are simply within the range 
of experimental error. Similar considerations apply to the variations 
in the pentose content. 
6 am. IO 2rm. 6 Su.yv- to 2 an Su.-w 
Set Yi.se. 
Fig. 14. Variation in Content of Various Carbohydrates in the leaf of 
Mangold during 24 hours, Aug. 26—27, 1913 (After Davis, Daish and Sawyer). 
It is indeed regrettable that Davis, Daish and Sawyer give no 
data which enable one to judge the likely limits of error of their 
determinations. They do indeed, under the heading “ Probable 
Error of the Analyses and Methods of Sampling,” show that the 
reduction method and optical rotation method give values for sucrose 
which differ by about 20%, and they also give the analyses of hexoses 
