20 Lioyp: LEAF WATER IN GOoSSYPIUM 
tions prevailing when the observations were made. Among these 
conditions it may be mentioned incidentally that the soil was well 
drained and rich in moisture at the time. Severer circumstances © 
would no doubt effect a still greater loss. The observed loss, how- 
ever, may be taken as indicative of a usual phenomenon, the reality 
of which is made evident in an observed daily wilting of the leaves 
beginning at about the 9 hour, detectable not alone by change in 
position (since this may occur as a phototropic response) but by 
flaccidity. The case may be otherwise stated by saying that 
under usual day conditions, with sunshine, the roots are unable 
to supply loss of water from the leaves. Balls’* view that the 
water supply is the limiting factor of growth, and his observation 
that no growth takes place under the Egyptian sun appear to be 
quite applicable to Alabama. With regard to the amount of 
growth in Alabama, preliminary measurements, prompted by the 
results obtained from the determinations of leaf water, indicate 
that even under the presumably more favorable humidity condi- 
tions obtaining here growth does not take place for the major 
portion of the day, since during the latter part of the growing 
season an actual daily shrinkage in stem and leaf length has been 
observed. I can hardly concur, however, with Balls in his view 
that because growth does not take place in sunshine this is to be 
interpreted as unfavorable. Comparative measurements on the 
same variety of cotton obtained in Arizona betray a no more un- 
favorable reduction of leaf water than in Alabama, when there is 
sufficient water in the soil. There is evidence that variations in 
soil moisture are registered in both the absolute leaf water content 
and in the rate of recovery after the minimum quantum for the 
day has been reached, while the loss during the first part of the 
day appears to be less affected. It would seem that the real test 
is the growth integer for the season, and it is not evident that, 
with irrigation, the conditions in the semiarid Arizona desert at 
all events are unfavorable from this point of view. A hot sunshiny 
day after all may be good for cotton, but this good may not be 
apparent in growth at the time. This is indicated by the amount 
of photosynthates formed (measured with small error due to well 
* Balls, W. Lawrence. The physiology of the cotton plant. Cairo Sci. Jour. 4: 
i-9. Ji to910; Cotton investigations in 1909 and 1910. Cairo Sci. Jour. 5: 221- 
234- S$ 1911. 
