Lioyp: LEAF WATER IN GosSYPIUM 5 
it varies between the extremes of 318 and 220 per cent. The 
greatest difference between the two figures for any hour is 26, 
this greatest discrepancy being found at the 8 and at the 14 hour 
on September 17. If we assume an error of 13 units and apply a 
corresponding correction by addition to the smallest and by sub- 
traction from the largest percentage, a difference of 72 units is still 
afforded. If the correction is similarly applied to the percentages 
in either series, the difference may be reduced to 46 units, as 
between the 8 and 14 hour figures. It will appear from these 
figures that the error of observation is not great, and is probably 
within five per cent., a conclusion well sustained by the general 
correspondence of the figures for the two series. It is, further, 
probably entirely true that the differences are not due to a material 
error of measurement, but are actual differences in the leaf water 
of different parcels of material. Since these were not taken by 
myself, I have compared the differences with those displayed by 
my data for the ocotillo, upon which I based my conclusions 
embodied in my paper above referred to,* and I find that dis- 
crepancies in similar amounts are to be found in them. The 
same is true of the data which appear in the present account 
beyond. They may be due to individual differences, or to dif- 
ferences in groups of plants. under various external conditions, 
or in the positions of the leaves taken. The general correspond- 
ence between the two series (FIG. 1) is, however, sufficiently close 
to warrant drawing the conclusion that during the night there is 
a rise in the leaf water content (which probably had been pre- 
ceded by a greater increase previous to the 16 hour) and a decided 
decrease during the day till 2 P. M. The sharp rise and equally 
sharp fall indicated between the 18 and 22 hours is puzzling, 
since if we refer it to change in dry weight, the additional difficulty 
is introduced of accounting for marked changes in the same. 
Aside from this peculiarity of behavior the general rise in water 
content may be referred to the reduction of dry weight, but this 
involves the assumption that the dry weight was reduced much 
more rapidly in the period preceding the 18 hour (say between the 
14 and 18 hour) than subsequently. A somewhat similarly sharp 
* The relation of transpiration and stomatal movements to the water-content 
of the leaves in Fouquieria splendens. Plant World 15: 1-14. Ja 1912. 
