104 
Psyche 
[December 
in g reactions involving the antennal funiculi, the fore legs 
and sometimes the tip of the gaster. A single injury such as 
might have been received from the mandibles of the attacker 
would scarcely be expected to evoke such a general response. 
Moreover, as I shall presently show, injuries are rarely sus- 
tained by the Brachymyrmex during the initial attack. Yet 
this cleaning reaction after the attack is an exceedingly con- 
stant one. I observed it so often that I am led to suspect that 
some irritating substance is ejected by the Strumigenys at 
the moment of striking. I much regret that I cannot prove 
this point which is very interesting if true. 
While the Brachymyrmex was busy cleaning itself other 
Strumigenys workers would arrive and encircle it. Although 
the advantage of a concerted attack would seem obvious I 
never saw this happen. Each Strumigenys attacks separ- 
ately moving in with great deliberation and ascertaining the 
position of the Brachymyrmex with the antennae before 
striking. It is interesting to note that this particular re- 
action prevents the Strumigenys from striking each other. 
Quite frequently the positions of the attackers would be such 
that they would have struck one another had they relied 
solely upon the trigger hairs to release the mandibular mech- 
anism. By bringing the antennae into play they were enabled 
to recognize their nest mates and would alter their position 
accordingly. In the hundreds of times which I witnessed 
the Strumigenys attack I never saw one strike a nest mate. 
During this phase of the attack the Brachymyrmex would, 
of course, be struck at repeatedly. Not infrequently it would 
break out of the circle of attackers and run to another part 
of the nest. This, however, merely delayed the final result 
for sooner or later it was again surrounded. After the 
Brachymyrmex had been struck at perhaps twelve or fifteen 
times it usually became much dejected. It ceased the clean- 
ing movements which it had continued up to this time and 
made no further attempts to avoid attack. I wish to stress 
the fact that this dejection, if I may use that term, was not 
due to injuries. Except in rare cases the Brachymyrmex 
would reach this stage of the attack without any visible sign 
that it had sustained damage from the mandibles of its at- 
tackers. The change in its reactions was, nevertheless, most 
pronounced and this again leads me to the conclusion that its 
