1912] Long: Living Eggs of Rats and Mice nl is} 
three hand holes are large enough to permit the hand to pass 
easily. They are closed by small sliding doors; and the middle 
and right hand ones are cut in 
= sliding doors, which in turn 
cover larger holes in the wooden 
=) 
glyc. back (see fig. D and pl. 16). This 
arrangement permits enough lat- 
Reeurator eral and vertical motion of the 
mer. wrists to enable the operator to 
reach all parts of the interior 
jee without appreciably affecting 
wood 
the temperature within. Several 
folds of cloth (omitted from the 
hae photographs for the sake of 
clearness), fastened along the 
top above the holes (a, fig. D) 
and hanging down over them, 
Fig. E. Longitudinal vertieal are very effective aids in keeping 
sectional view of one end of the 
box shown in figure D. Letters . 
as in figure D. X \, In figure D it can be seen 
the temperature constant. 
that the microscope tubes pass 
through the glass where the edges of the roof plates meet. The 
holes in the glass are bored so that the glass fits snugly about 
the tubes of the microscope. Plate 17 shows how the wooden 
back, a part of the wooden end walls, and all of the glass back 
of the microscopes, can be removed in order to allow the micro- 
scopes to be put in place or to introduce large objects. 
The floor on whien the microscopes and apparatus rest con- 
sists of wire mosquito netting supported at the edges and by two 
longitudinal beams (bm., figs. D and E) running through the 
middle. The microscopes rest on strips of heavy glass which lie 
on top of the netting, but are supported directly by the beams 
and at the edges. Smaller objects are most conveniently placed 
on pieces of glass large enough to reach from one wooden support 
to another. 
The object of having a floor of netting is to allow the ascent 
of the air which is warmed in the lower part of the box below 
the floor. This lower part of double, wooden walls is shown in 
