ROSE-GALLS. 113 
amber-coloured, hard, shining, and appearing to the un- 
aided eye to be nearly spherical. They are about as 
large as dust-shot. or a long time I could not satisfy 
myself about them, not being able clearly to ascertain 
whether they were deceased insects or merely hardened 
sap. That they were probably of insect origin was 
evident from the fact that they were always found in 
cells which had no opening, and from which the insect 
had not escaped. 
At last, however, one of them happened to lie on the 
paper so that it could be well illuminated, and then the 
whole mystery was unfolded. These strange little objects 
were the pupe of the insects, which had died in the cells, 
and shrivelled up into the singular forms which have 
been described. ; 
The cells are of different sizes, some being more than 
ten times as large as others. The superior dimensions 
of the cell seem to be obtained at the expense of the 
walls, so that the large cells can be broken by the finger 
and thumb, while the small cells cannot be opened with- 
out the knife. 
The insects themselves are equally variable, some being 
mere dots of shining blue and green, while others are 
about as large as the common red ant of the gardens, 
but with plumper bodies. In consequence of these two 
facts, the large, strong-jawed insect can easily make its 
way through the comparatively thin walls of the large 
cell in which it was enclosed, while the small and neces- 
sarily weak-jawed specimens are utterly unable to pierce 
the walls of their cells, which are so thick that they 
must bore a hole equal in length to that of their whole 
body before they can escape into the air. Consequently 
the great mass of the insects that are found in the 
cells are the small specimens, the larger having made their 
-P 
