1951] 
Creighton — Arizona Ants 
61 
agers. But the ants rarely ascended the fruit stalk. In- 
stead they confined themselves largely to the upper surface 
of the leaves. The areas around the bases of the lateral 
spines and particularly the area at the base of the long 
terminal spine seemed especially attractive. A worker would 
often stand for many minutes at the base of the terminal 
spine, apparently licking some substance from the surface 
of the leaf. The explanation which best fits this behavior 
is the assumption that, during the fruiting period, the leaves 
of the agave give off a sugary substance which attracts the 
ants. This secretion cannot be produced by the younger 
plants as otherwise they would also be visited by the ulcero - 
sus workers. 
Since feeding methods had failed to reveal a nest, it 
seemed advisable to examine the soil in the immediate vi- 
cinity of the agave plants in the hope of finding colonies 
there. The difficulty was to get at the soil immediately 
beneath the plant. Any one who attempts to uproot a full- 
grown plant of Agave palmeri will soon have a healthy re- 
spect for the needle-sharp terminal spines. To get close to 
the roots of the plant it was necessary to ‘dehorn' all the 
leaves on one side with a pocket knife. The plant could then 
be wrenched out of the ground with a pick. But the vio- 
lence of this operation defeated the purpose for which it 
was intended. Tearing the roots out of the ground dis- 
turbed a considerable area of soil and badly displaced the 
stones in it. The result was to obliterate any nest passages 
present. The one fragment of an ulcer osus colony which 
was exposed in this fashion was so badly scattered through 
the soil that, for reasons already explained, I doubted that 
the female was that of ulcerosus. It then occurred to me, 
as should have been obvious sooner, that if this species 
customarily nests under agave plants, the nest should be 
present for at least a while after the plant has died. There 
were many dozens of dead agave plants on the side of the 
canyon. These consisted of a rosette of dried and shrivelled 
leaves with a central hole where the fruit stalk had been. 
The whole thing was comparatively light and easily handled. 
With a little care it could be lifted entire and tossed to one 
side without disturbing the ground beneath. I removed a 
