278 BOTANICAL GAZETTE [SEPTEMBER 
of the dates given. Chemical changes subsequent to collection were prevented 
by dropping the material immediately into a large volume of boiling 95 per cent 
alcohol containing 1 per cent concentrated ammonia, which instantly destroys 
the enzymes present. In the subsequent analyses, the methods outlined by the 
authors in their papers on the estimation of carbohydrates in plant material 
were employed; the chief new features of these methods are the employment of 
ro per cent citric acid for the inversion of saccharose, the estimation of maltose 
by fermentation with maltase-free yeasts, and the determination of starch by 
the use of taka-diastase. 
Maltose and starch were entirely absent from the leaves at all times, day 
and night, in all three series. Starch is present in very young leaves, but 
disappears as soon as the roots have grown sufficiently to be capable of storing 
sugar. In the first series, hexoses seca to i as eee after sunrise, 
attained a maximum of 2.16 cent between A.M. and noon, declined 
sharply until 4 P.M., eee dectssd steadily sbseaahont the night to start 
upward again at 4 Aw. The curve representing saccharose rose more slowly 
from sunrise, maintaining a abies of 3.11 to 3.06 per cent from noon to 
4 P.M., then dropped in an almost straight line through the night to start up 
at 4A.M. Both curves roughly paralleled the temperature curve. 
In the second series (that of September 10) both curves were complicated; 
that for hexose shows a minimum at 8 A.M., with a rapid rise to 7.5 per cent at 
1 P.M., followed by a slight decline for 3 hours which is succeeded by a rise to 
8.9 per cent at 6 p.m. Two hours later this has fallen to 6.75 per cent, but 
there is again a rise to a new but lower maximum of 7.81 per cent at 2 A-M., 
after which there is a sharp decline, continuing until sunrise. The curve for 
saccharose is similar, in that it shows two maxima at 6 P.M. and 2 A.M., but 
differs in that the second is much the largest, the amounts being 6.39 and 8. 27 
per cent. The curves for hexose and saccharose in the third series are alike 
‘in that each presents three maxima; for hexose these occur at 1 P.M., 9 P-M.; 
and 3 A.M., the last being greatest, while those for saccharose occur at 3 P-M., 
P.M., and 3 A.M., the second being considerably higher than the others. In 
neither the second nor the third series is there any resemblance to the tempera- 
ture curve. In the first series, the amount of saccharose present is at all times 
much greater than that of hexose, becoming 7 times as great at 4 A.M., and the 
fluctuations in amount of hexoses are much greater than those of saccharose. 
In the second series, hexoses vary between 8.9 and 5.4 per cent and are larger 
in amount than saccharose, which varies from 8.27 to 4.24 per cent. To this 
statement there is one exception at 2 A.M., at which hour saccharose is slightly 
in excess. In the third series, hexoses are again in excess, varying from 12.4! 
to 9.39 per cent, while saccharose ranges between 9.52 and 4.98 per cent. 
The variations in saccharose are greater than those in dextrose in both second 
and third series, and are greater in the third than in the second. Consequently, 
while the ratio of invert sugar to cane sugar varies, in the first series, between 
0.133 (at 4 A.M.) and 0.710 (at 10 a.M.), and is expressed by a curve closely 
