276 BOTANICAL GAZETTE [APRIL 
which apparently could: derive no advantage from an additional 
supply of water presented to their leaves. 
The record of these three species of Celtis, Covillea, and Lycium 
has been given at length, on account of the importance of estab- 
lishing beyond doubt the fact that in these plants, which have been 
taken to represent desert species that retain in structure and habits 
obvious indications of mesophytic origin, leaf absorption certainly 
takes place, and apparently as an entirely normal process. We 
have next to deal with a group of species genetically related, which 
deport themselves quite differently from members of the first bio- 
logical group in regard to leaf absorption. As representatives of 
this second group, species of Parkinsonia, Prosopis, and Acacia were 
selected, all belonging to the Leguminosae. The record of experi- 
ments and their results is such as to admit of statement in few words. 
GROUP II. 
Parkinsonia Torreyana. 
The specimens of palo verde employed in this work were seed- 
lings some two months old. One was cut about thirteen hours, 
the other (number 2) an hour and a half before weighing. After 
weighing an attempt was made to wet the leaves by repeatedly 
immersing the seedlings in water. The experiment might fairly 
have been dropped at this point, since, as it was found impossible 
to wet them, leaf absorption could hardly be thought of; but as 
there remained a possibility of some slight absorption where drops 
of water collected on the surface of the youngest parts, the attempt 
was continued with number 1, which was repeatedly immersed 
during a period of something over three hours. 
As seen from Table X this seedling, so far from gaining by absorp- 
tion of water presented to it, actually lost 1.8 per cent. of its weight 
in three hours and thirteen minutes, its surface having remained 
almost entirely unwetted, so that loss of water was possible during 
the whole, or nearly the whole, of this period. Seedling number 2 
was allowed to dry, after an unsuccessful series of attempts to wet 
its surface. Its loss of weight, as might be expected, was greater 
than that of number 1. 
If these results are compared with those of November 1 and 9, 
derived from similar experiments with shoots of Parkinsonia micro- 
