Lioyp: LEAF WATER IN GOSSYPIUM 7 
rise in water content is indicated after the 6 hour, September 17; 
continuing till 8. The tables following show such behavior in 
only one instance (series 12), but it must be understood that the 
data therein contained were based on material collected on bright 
days, while those now being considered were taken on a morning 
(of September 17) during the first few hours of which a general 
haze prevailed. The single exception, offered by the old leaves 
collected on October 13, I91I (series 12, TABLE viit), differs 
quantitatively, there being here a difference of 15 units (251 to 
266 per cent.) as compared with Ig units (284 to 303 per cent.) 
in the case before us of September 17. 
The value of the comparison is slightly compromised by the 
fact that the data of TABLE vill indicate a decrease of dry weight, 
in an amount that may, however, fall within the limits of error. 
If no change in dry weight is assumed, the difference in water 
content in percentage of dry weight would be somewhat reduced. 
In view of the further fact that the data for both TABLES v 
and VIII were obtained after an increase in soil moisture, it seems 
entirely possible that under certain conditions the first two hours 
of daylight may be accompanied by an increase in leaf water. 
This involves as a corollary that the leaf, at the hour of sunrise, 
may not be completely plethoric of water. 
In order to evaluate with some degree of exactness the meaning 
of the graph embodying the data of TABLE I, I have assumed that 
on September 17 the dry weight of the leaves increased in the 
amounts of 9 per cent. on the original weight in the first four 
hours (6-10) and 17 per cent. in the whole period of eight hours, 
namely 6-14. These amounts are taken from TABLE VII, series 
8, and represent fairly closely the average performance for young 
leaves. The weighings for the 6 hour are taken as a basis for the 
calculation, and the leaf water content is assumed to remain 
constant, the dry weight only increasing. By comparing the 
ratios thus obtained with those actually observed, as displayed 
in TABLE II, we see that the calculated value of the ratio at the 10 
hour is less than the observed, and at the 14 hour greater (graph 
abc, FIG. 1), the latter indicating a net loss of leaf water. If 
however, an increase a 30 per Soar on the original dry weight 
be assumed, an a ly justified in view of the data 
ca aa os 
