STUDIES OF THE DEVELOPMENT OF PEPEROMIA HISPIDULA 329 
these stomata are also like those mentioned in being unable to close 
at all completely. It seems evident that such a stoma would be no 
serious disadvantage to a plant of very humid habitats. It is possible 
that Haberlandt's suggestion, that the sharp-edged guard cells of the 
plants mentioned prevent closing of the stomata by water drops, 
may also be applicable to this Peperomia. The only portions of the 
lower side of the leaf where stomata are more abundant than is indi- 
cated above are the areas underlying the glandular swellings of the 
veins near the margin of the leaf {G, in fig. i). Here the stomata 
become relatively numerous, though apparently not differing in size 
or structure from the stomata occurring on the thinner portions of the 
lamina (figs. 17, 18, 21). It is probable that these stomata serve as 
water stomata, as we shall see later, when speaking of the glands with 
which they are associated. 
The mesophyll portion of the lamina, between the veins, consists 
of a single layer of slightly specialized, nearly isodimensional palisade 
cells, and of one or, near the veins, of two layers of sponge parenchyma 
(figs. 16, 17). Near the margin of the lamina the tissue between the 
upper and lower epidermis may consist solely in addition to the single 
layer of palisade, of one layer of very scattered parenchyma cells 
(figs. 17,21). The palisade cells are 40 to 45 jjl in vertical length and vary 
from 20 to 45 fjL in diameter. Each contains a single large nucleus and 
from 15 to 25 chromatophores. These cells are closely in contact with 
other cells except at their rounded inner ends, which are surrounded 
by considerable air-spaces and rest upon one or more of the underlying 
parenchyma cells. The latter are often 10 to 15 thick and 40 or 50 jj, 
long tangentially, and may contain 12 or 15 chromatophores, which are, 
however, only about half as bulky as the chromatophores of the pali- 
sade cells. It is evident, from their contents and from their relation 
to other tissues, that these sponge-parenchyma cells must serve not 
only for photosynthesis but likewise for the conduction of the products 
of this process from the palisade cells to the vascular bundles. They 
must evidently also carry water in the reverse direction. The water- 
storing hypodermis, characteristic of the leaves of all other Peperomias 
studied except those of P. pellucida and P. tenera (Jaderholm, 1898) 
is entirely wanting in the leaf of P. hispidula. Moreover, there is no 
evidence from the structure of the epidermis, that this latter tissue 
serves for water-storage in our species, as Jaderholm says it does in 
P. pellucida. In the leaf of P. hispidula then it is evident that we 
