AN ANATOMICAL STUDY OF GYMNOSPORANGIUM GALLS 403 
of this gall seemed to offer an opportunity for further investigation. 
As the juniper trees in the vicinity of Madison, Wisconsin, furnished 
abundant material of this gall in various stages of development I 
concluded to determine, at least for my own satisfaction, which of 
these statements was correct. I soon discovered, from external exami- 
nation, that some of the older galls had remains of more than one leaf 
attached to them, a fact which in itself would indicate that stem 
tissue was involved to a greater or less extent. I found still further, 
that normal branches are attached to the gall, apparently growing 
out from it at times (fig. 7), a condition that would be rather unusual 
if the gall were composed entirely of modified leaf tissue. Rather 
thick hand-sections showed that when a gall was located on the end 
of a stem, the stem continued into it for some distance before it 
broke up. This hasty examination seemed to justify a more careful 
study, the results of which are given in the following pages. 
Before entering into the description of the anatomy of the gall it 
seems desirable to describe briefly the anatomy of both the leaf and 
young stem, as both of these are involved in gall formation. 
The normal leaf is triangular in cross section except towards the 
base where it is rather four-sided. The epidermis is heavily cutinized 
and forms a smooth layer on the outer (morphologically lower) side 
of the leaf. Towards the edges and inner side, however, the indi- 
vidual cells are somewhat raised towards the outside causing them to 
appear more or less papilliform when seen under the microscope. 
The stomata occur mostly in the inner (morphologically upper) 
epidermis, but they may appear occasionally in the outer epidermis 
near the edges of the leaf. 
In addition to the epidermis, the outer side of the leaf is protected 
by a hypodermis which may extend entirely or only part way across 
this side. The hypodermal cells are usually arranged in one or two 
layers, although three layers may be present at times over a space. 
Sanford (8) describes a double epidermal layer for this side of the 
leaf but he evidently mistook the hypoderm for an inner epidermal 
layer. The interior of the leaf is filled with rather thin-walled paren- 
chyma cells (text fig. i) which are loosely arranged towards the inner 
side near where the stomata are borne, and resemble very much 
ordinary spongy leaf tissue. Towards the outer side of the leaf the 
parenchyma cells are elongated perpendicular to the surface and form 
a palisade tissue. The base of the leaf is united with the stem for some 
