120 
Psyche 
[March 
found to occupy the terminal six feet of the limb, ending approximate- 
ly 54 feet above the ground. 
Figure 2 shows the extent and character of what must have been 
several years’ work by this well-established and vigorous colony. The 
configuration of the galleries indicates that the colony had originally 
extended farther up the limb and probably weakened it sufficiently to 
result in the break during a windstorm. The very large and extensive 
central galleries in the upper half of the colony were almost completely 
filled with hardened masses of fecal pellets, while the same situations 
were filled approximately 50 percent to within 18 inches of the lower 
end. It is apparent that this species often takes advantage of the 
softer spring wood in driving new galleries (note cross sections in 
Fig. 2 ) , but through continued working these concentrically arranged 
chambers may be developed into cavities three or four inches in 
diameter and much more in length. The surface of the galleries has 
a coarse, velvety finish, characteristic of the other species of Zooter- 
mopsis. Cottonwood is fairly soft and light but very fibrous and 
tough. At the time the limb was cut, it was so wet from recent 
winter rains that water was squeezed out by the wedges used in 
splitting the sections. Although no other insects had penetrated the 
wood, the termites had followed shrinkage cracks in several places, 
particularly in the burls. The extremely tough and contorted grain 
of the burls seemed to be especially attractive. Four burls in the limb 
were extensively worked and each contained tunnels to the outside 
which were plugged with fecal material. One other tunnel, just 
below section C-C (Fig. 2) was driven directly to the outside. These 
openings measured about 3 X 5 to 6 mm. and were presumably used 
as exits for the alates and perhaps also for dumping fecal pellets. It 
is interesting to note that the galleries of Z. angusticollis are pre- 
dominantly peripheral in logs and stumps and that long, slit-like 
openings are made for the escape of the alates (Castle, 1934). 
Three other colonies were found in willows along the bank of the 
San Pedro River at Hereford. The second complete colony was 
taken from a living tree approximately 30 feet high and 12 inches in 
diameter at the base. Individuals of this colony were discovered in 
damp, fermenting wood at the base of a broken branch six inches in 
diameter and four feet above the ground. The tree was felled and 
handled in the same way as the limb from Patagonia. As shown in 
figure 3, the galleries extended about two feet below the crotch of the 
snag in a pattern similar to that in the cottonwood limb. Some large 
cerambycid larvae had attacked this region prior to the termites, for 
the latter had driven galleries through the damp and firmly packed 
