108 
Psyche 
[December 
latter happens to be an antennal funiculus which increases 
in diameter toward the tip. The backward jerk of the 
Strumigenys is then exerted in the direction of the long axis 
of the already injured appendage which may break in con- 
sequence at the damaged joint. It may be added that unless 
the appendage has been previously crushed it cannot be 
pulled off when caught by the apical teeth. Under such cir- 
cumstances there ensues a sort of a tug-of-war with the 
Strumigenys holding grimly to the appendage and the victim 
struggling frantically to get free. 
Aside from their vital role in attack the mandibles of 
Strumigenys are little used. They occasionally employ them 
in carrying nest mates about but this was seldom observed. 
I never saw any indications that the mandibles play a part in 
the trophic reactions of these insects. Feeding is entirely 
cared for by the palps and other mouthparts. When Strumi- 
genys is feeding the mandibles are kept closed and thrust 
upward over the food until the stubby palps come in contact 
with its surface. The palps apparently act both as rasps and 
spoons for both liquids and semisolids are ingested by their 
help. The feeding reactions of Strumigenys are exceedingly 
difficult to follow because all the more delicate mouth parts 
are reduced in size and closely packed into the small buccal 
cavity. It is interesting to note that the mandibles are kept 
closed during regurgitation. This results in a rather awk- 
ward situation for both regurgitant and recipient. The 
closed mandibles prevent the usual approximation of the two 
heads and it is only after considerable preliminary fencing 
that the correct posture is assumed. Each ant turns its head 
sidewise so that ventral surfaces are parallel although not 
opposite. The closed mandibles are then crossed and slid 
over each other until the palps are close enough for the 
transfer. It seems a remarkably clumsy way of securing a 
result which could be easily reached by simply opening the 
mandibles. 
There remains the matter of retrosalience. I saw this 
phenomenon only twice during the many hours that the 
captive colony was under observation. In one case it resulted 
when the mandibles of an attacking Strumigenys worker 
closed on the convex surface of a glass tube which led to one 
of the feeding chambers. The resulting leap threw the insect 
