534 
ALBAN STEWART 
hand side of the chamber. By further growth of the cambium a 
condition comes about Hke that shown in fig. 5, in which the wood has 
been formed well around the base of the chamber. The final result 
of this process is shown in fig. 2, where the inner third of the chamber is 
surrounded by wood. The enclosing method here outlined is some- 
what similar to that described by Maule (15, p. 24), where woody 
bodies, formed in the bark of wounded stems of Salix phylicifolia, are 
gradually surrounded b}^ the activity of the cambium forming wood 
around them. 
After a larval chamber is vacated by an insect the opening to 
the outside remains for some time. It is possible to stick a dis- 
secting needle into the openings in some of the older galls for a distance 
of 1.5 cm. or more, and longitudinal sections through such chambers 
show that they may extend in nearly to the normal wood around the 
center which was evidently formed before the gall started. Sometimes 
the openings can still be seen several years after the cavity was 
vacated but after a time they may become closed and more or less 
obliterated by the growth of the bark around them. 
A longitudinal section through a chamber recently vacated by an 
insect is shown in fig. 2. The chamber is lined with several layers 
of stone cells which, however, do not extend to the periphery of the 
gall. The thin-walled parenchyma cells, which fill the outer portion 
of the chamber before the insect emerges, have been largely destroyed 
in the section figured, and can only be seen to the right and left just 
inside the opening. In earlier stages, before the insect comes out, the 
stone cells extend inward from the periphery toward the center oi 
the chamber and form a loosely arranged arch across it a short distance 
inside the opening. The loose arrangement of these cells offers but 
little resistance to the insect in leaving the chamber. Outside the 
protecting layer and running parallel with the sides there is a corky 
covering composed of several layers of cells, shown by a light line on 
each side of the chamber in fig. 2. It is probably somewhat similar to 
the corky covering around the protecting layer in the stem gall of the 
oak described by Lacaze-Duthiers (14, p. 338). It is not always 
the case that the protecting layer extends entirely around the chamber, 
as the inner end and sides of it may join directly onto the wood in 
these places. 
If the protecting layer is examined in a longitudinal section through 
a chamber, it will be noticed that there are in general three kinds of 
