AMERICAN JOURNAL OF BOTANY 
[Vol. 8 
is enclosed in a thin-walled sac. The hairs appear to be of two types, which 
apparently correspond to different stages in development : a densely granular 
and a sparsely granular form. This difference in granular density is 
interesting. In animal glands it has long been noticed that when a serous 
gland has been quiescent for several hours the secretory cells are granular 
throughout, and the outlines of the cells are only faintly marked as clear 
lines bounding the granular areas. When the gland secretes, many of the 
granules disappear and after prolonged secretion very few granules are left; 
i.e., during secretion the granules normally contained by the cells are in 
some way or other used up, probably to form a part of the secretion. Al- 
though the diminution of zymogen granules is a normal occurrence in the 
secretion of the salivary, infra-orbital, lachrymal, mucous, and pancreatic 
glands, yet in the case of the mammary glands the opposite is true, viz., 
that granules begin to form with the commencement of secretion and do 
not occur during rest. In the mammary gland, the active growth of proto- 
plasm, the formation of granules from the protoplasm, and the discharge of 
these granules in the secretion appear to go on at one and the same time. 
Investigation of the club-shaped hairs of Rhus diversiloba has not as yet 
revealed a positively glandular nature, and consequently a relation between 
differences in their granulation can not be definitely connected with secretion. 
From a morphological standpoint, however, as above pointed out, the 
club-shaped hairs seem to be glandular. 
Club-shaped hairs from leaves gathered in the morning before sunrise 
and from those secured in the heat of the day could not be differentiated. 
Hairs from rapidly growing leaves could not be distinguished from those 
of old leaves or stems. Hairs from leaves grown in sunny exposures ex- 
hibited no differences, although they were present in greater number than 
on leaves continuously in the shade. 
The Stem 
A transverse section of a green stem of Rhus diversiloba shows, beginning 
at the outside, the following tissues (PI. Ill, fig. 5): epidermis, with its 
trichomes and stomata; collenchyma; cortical parenchyma; pericycle, with 
bast fibers and thin-walled pericycle parenchyma; phloem, with resin ducts; 
cambium; xylem; medullary ray; pith. 
As the stem increases in diameter (fig. 3) the cortex develops a phellogen. 
The continuous activity of the phellogen results in an increasing thickness 
of the sheet of cork. The chloroplast-containing tissue beneath the cork 
layer maintains connection with the air by means of lenticels which have 
replaced the stomata. As may be anticipated, the dead cork cells are 
non-poisonous, i.e., they do not cause dermatitis when rubbed on the skin 
of a susceptible individual and therefore do not constitute a means of 
transference for the poison. 
No resin ducts have been found in the pith of this plant. Engler (5), 
