Pomona College, Claremont, California 159 
ing timbers; but, on the contrary, was always found working in the 
sound wood. This, I think, is a wise choice, for one of its dangerous 
foes is found abundantly, tunneling through decaying redwood. 
The surface chosen for making an entrance is generally vertical or 
slanting. When slanting or, as is sometimes the case, horizontal, 
the under surface is always chosen. For a short distance the bur- 
row takes a course nearly or quite at right angles to the surface 
entered, then gradually changes to a course parallel to that surface, 
and always (with very few exceptions) leads upward in the slanting 
or upright timber. These tunnels vary in length from one inch to 
twelve inches and are, as a rule, remarkably straight. I am at a 
loss to know certainly what guides these interesting little carpen- 
ters in the construction of so straight a tunnel. My first idea was 
that they followed the grain of the wood, but in one case, where a 
knot occasioned a decided curve in the grain of the wood, the tun- 
nels had been constructed straight as usual (Fig. 7). The most 
reasonable explanation seems to be that the vibrations of the wood 
serve to indicate the distance from either surface, for when boring 
in a plank only 54-inch thick they keep a line remarkably nearly 
equidistant from the two surfaces and never have I seen where they 
broke through to the surface. But a fact in the way of this theory 
is that they sometimes make tunnels just as straight in a 2-inch 
timber with the distance from one surface several times greater 
than the distance to the opposite surface. It is an interesting point 
which T have not yet solved to my satisfaction. My description 
fits the majority of tunnels. Ina very small percentage of the cases 
studied, the tunnels were short and seemed to be in almost any 
position. 
The excavation of these tunnels is evidently a laborious task, 
though the little creatures ply their trade with great avidity, and 
while at work they are not at all easily disturbed. The writer 
watched one of these patient workers for three hours, during which 
time she kept her mandibles working away continuously, leaving her 
work only twenty-five minutes, evidently for “lunch,” after which 
she returned to resume her task. By closest scrutiny I was unable to 
see that the two and a half hours of labor had lengthened her bur- 
row. I returned six days later to find her still vigorously pursuing 
