38 
Psyche 
[March 
semi-erect scales that largely conceal the green surface beneath them 
and give to the plants their characteristic grey color. Large numbers 
of pollen grains are trapped in the spaces between these scales and 
the ants have no difficulty extracting them. The advantages of using 
Tillandsias as a pollen source for captive colonies of t exanus are out- 
standing. Since the epiphytes will stay fresh indefinitely, an ample 
supply can be kept on hand without difficulty and, when they are 
introduced into the terraria, there is no need to provide them with a 
water source. After a few weeks it became clear that the captive 
colonies could be kept in excellent condition with no food source 
other than the pollen grains trapped on the Tillandsia leaves. This 
was a most surprising discovery for, if captive colonies can subsist 
on nothing but Tillandsia-trapped pollen, free colonies should be able 
to do so as well. 
The significance of this will be obvious to anyone who has tried to 
arrive at an explanation for the diversity of the trees in which the 
nests of texanus have been found. The first data on this, published 
by Dr. M. R. Smith in 1947 (4), appeared to indicate that texanus 
has no preference for a particular kind of a tree as a nest site. Five 
of the six trees carried in this list belonged to unrelated genera. By 
1954 the number of records had risen to twenty-seven and with 
this increase there appeared the fact that some selective process 
must be involved in the nesting responses of texanus. In this second 
list seventeen of the records came from live-oaks and this dispro- 
portion (63%) was too great to permit the view that any tree is 
equally suitable as a nest site for texanus. At present the dispro- 
portion of records from live-oaks has risen to 71% and, in view of 
the fact that the remaining 29% of the records are spread over six 
different trees, it follows that the incidence of nests in live-oak trees 
is at least eight times greater than it is in any other tree. As soon as 
it was found that pollen grains are the principal food of texanus it 
became clear that the capacity of live-oak leaves to trap wind-blown 
pollen makes this tree an especially favorable nest site for texanus. 
This advantage is so striking that the difficulty is not to show why 
texanus usually nests in live-oak trees but to explain why it should 
ever nest anywhere else. There is a strong temptation to treat other 
records as accidents and sweep them under the rug, which was 
essentially what the writer did with them in 1963. If this is done 
one can then fall back upon the often cited but seldom proved 
explanation of host plant preference. 
But the fact remains that texanus occasionally nests in other trees 
than live-oaks and, with an ant whose responses are as rigid as those 
