AN ANATOMICAL STUDY OF GYMNOSPORANGIUM GALLS 4I I 
cork cells. Except near the central cylinder, the parenchyma cells 
for the most part are not strikingly larger in the older gall than they 
are in the younger gall (shown in text fig. i). Starch as well as 
droplets of oil occurs abundantly in many of the cells. 
Since this article was sent away for publication a paper has ap- 
peared by Reed and Crabill (12) in which the anatomy of the gall of 
Gymnosporangium juniperi-virginianae is briefly considered. These 
authors regard this gall as arising from nothing more than an hyper- 
trophied leaf, and that its fibro vascular system is a modified con- 
tinuation of the fibrovascular system of the leaf. 
From the descriptions and figures which have been given in the 
preceding pages, it seems evident that these authors have been mis- 
taken in their interpretations. The diagrammatic drawing of a 
longitudinal section of a young gall and of a branch, which they 
show in their fig. 4, is probably of a stage intermediate between the 
galls I have figured in fig. i, and text fig. i. These authors have 
seemingly failed to find a leaf-trace bundle in addition to what they 
have shown, and have interpreted the fibrovascular system as a. 
modified leaf trace. A section cut at right angles to the one shown in 
their figure would probably reveal the fact that this system has the 
structure of a stem and is in no way a leaf bundle. 
Gymnosporangium globosum on Juniperus virginiana. 
The relation of the younger galls of G. globosum Farl. to stem and 
leaf is very similar to that of G. juniperi-virginianae, considered 
earlier in this article. In fact the similarity is so great that it is 
practically impossible to distinguish the two with any degree of 
certainty until the sori have formed. Cross sections of the younger 
galls have essentially the same structure as galls of similar size which 
are caused by G. juniperi-virginianae (text fig. i). There is this 
exception, however, that the strands of tracheids given off from the 
stem in the gall, which correspond to the two strands at the right of 
the resin cavity ir' in the above mentioned figure), are separated by a 
very broad band of parenchyma. This condition is to be expected 
from what is usually found in the older galls caused by this species of 
Gymnosporangium . 
The woody cylinder in older galls presents two entirely different 
types as far as was shown by the material studied. The more common 
of these types will be considered first. A cross section of an olde- 
