202 BOTANICAL GAZETTE [MARCH 
In case the chrom-acetic fluid might have altered the hygro- 
scopic character of the cell walls, the same test was repeated with 
corrosive sublimate as the killing agent, and the same kind of 
result obtained. The leaves of Hymenophyllum polyanthos were 
also killed and carried through the moist chamber, with closely 
similar results. 
An attempt was made to determine whether the hairy covering 
on the leaves of Hymenophyllum sericeum and related species per- 
forms any function in connection with water supply or conserva- 
tion. The hairs cover both sides of the leaf densely, are multi- 
cellular, and made up of a basal stalk upon which are jointed (4-7) 
hairs about o.5-1 mm. in length, diverging from each other and 
parallel to the surface of the leaf. Both the basal stalk and the 
radiating hairs are hollow and devoid of living contents. In sur- 
face-dried leaves and those commonly gathered in the field, the 
hairs are filled with air. Leaves which were submerged, with all 
of the occluded air squeezed out from among the hairs, were found 
after 24 hours to have their hairs filled with water. On observing 
the access of water to empty hairs beneath the microscope, it was 
seen to enter them at once and rapidly, until only a small bubble 
of air remained, which persisted for several hours. When leaves 
in which the hairs had been completely filled with water by long 
submergence were surface-dried with filter paper and hung in the 
open air at a humidity of 80 per cent, they were found at the end of 
t hour and 20 minutes to have lost approximately half the water 
contained in the hairs. Of the amount lost it is more than probable 
that as much escaped by evaporation from the hairs as became 
available to the leaf cells. In other leaves which. had empty hairs 
and which were placed in a very moist atmosphere the gains . 
weight were not accompanied by a visible accumulation of water m 
the hairs. It appears, therefore, that the leaf hairs do not serve 45 
reservoirs of water either in sufficient quantity or for a sufficient 
length of time to be of importance in that réle. The coating of 
hairs as a whole is capable of retaining externally an amount of 
water equal to twice the weight of the surface-dried leaf, and 
under the conditions of evaporation of the rain forest this external 
water might be retained for 5-10 hours. The writer has placed 
