NOTES ON THE ANATOMY OF THE PUNCTATUS GALL 535 
stone cells which compose it. There are stone cells which are iso- 
diametric or nearly so, and others which are much elongated in one 
direction, the last of which have their long axes arranged parallel 
with the chamber. There are still other stone cells which are more 
or less intermediate between the two kinds just mentioned, in which 
one diameter is somewhat longer than the other. Sections cut across 
a chamber sometimes show that these intermediate cells are arranged 
with their long axes perpendicular to it. It is evident that both 
isodiametric and elongated stone cells make up the protecting layer 
of this gall. There is a great difference in the thickening of the walls 
of the stone cells. Some have very thick walls with well-marked pits 
and small lumina, while others have thin walls and large lumina. 
Furthermore it is possible to find, among these, cells which have the 
walls unequally thickened on different sides, and which are very 
similar to those described by Weidel (21, pp. 329-330). Such cells 
do not occur in the normal parts of the oak, according to this author, 
but they do occur normally in certain members of the Rosaceae accord- 
ing to Kiister (12, p. 183). Weidel found still further that stone cells 
with one sided thickening of their walls were confined largely to leaf 
galls in the forms studied by him. He also makes some interesting 
comparisons between the stone cells in Cynipid galls and those which 
occur normally in the oak. He finds that isodiametric stone cells 
are the ones that commonly occur in the normal parts, those which are 
elongated in one direction being confined to the bud scales and repro- 
ductive parts. It might be well to mention in this connection that 
Kiister (12, p. 181) has found that the palisade-like stone cells with 
large lumina which occur in the gall of Hormomyia fagi call to mind 
similar cells that occur in the beech cupule. As elongated stone cells 
occur in the reproductive parts of the oak and related forms, one is led 
to suspect that possibly this kind of stone cell may be more primitive 
in the oak than the isodiametric form, since primitive characters are 
likely to persist in these parts. This supposition is strengthened 
by the fact that this kind of stone cell reappears in many oak galls 
so that the ability to form them must be latent and is only awakened 
to activity through the action of the gall stimulus. Such may be 
true of the ray structure to be considered further along in this article. 
In young normal stems of this species of oak the stone cells of the 
bark are arranged in one or more concentric layers. Groups of stone 
cells may also occur opposite the broad rays, which are connected 
