1937] 
Habits of Strumigenys 
107 
however, convinced me that the mandibles of Strumigenys, 
unlike those of Odontomachus, do not act as shearing organs. 
I have already pointed out that each time a Strumigenys 
worker strikes at a victim the closure of the mandibles is 
immediately followed by a backward jerk of the insect. At 
first this used to annoy me considerably since it seemed to 
defeat the whole plan of attack by moving the Strumigenys 
out of range. I later learned to appreciate the vital part that 
this sudden backward motion plays in the amputation of 
appendages on which the mandibles close. It may be readily 
observed that when the Strumigenys workers close in for the 
final attack the damages to the antennal funiculi and tarsi 
of the victim involve mangled joints. The portion of the 
appendage beyond the injured joint will often bend in a 
fashion which plainly indicates that the joint has been badly 
crushed. As a rule the part of the appendage beyond the 
damaged joint is quickly removed as the result of further 
attacks. As its final removal is accomplished with the speed 
of a conjuring trick it was some time before I realized exactly 
how this is done. It is obvious that the initial damage to the 
appendage is caused by the flattened inner faces of the man- 
dibles of Strumigenys which crush the joint on which they 
close but lack the power to completely sever it. I have 
already pointed out that in repose the mandibles of Strumi- 
genys do not quite meet and this may account for their 
deficiency as pincers. It might be supposed that the back- 
ward jerk which follows this initial damage would sever the 
appendage but this is not usually the case. I believe that this 
can be explained if we consider that the crushed joint is free 
to move along the approximated inner faces of the mandibles 
which are smooth and without teeth except for the three at 
the apex. As I have already stated the final removal of the 
distal portion of a damaged appendage is a very speedy 
process and I am by no means certain that the method which 
I am about to describe is invariably employed. There has 
been ample opportunity, however, to observe that it is fre- 
quently used. It involves a relation between the appendage 
and the mandibles such that the four apical teeth close 
around the appendage. Because of the shape of the space 
included between these apical teeth (vide supra) they can 
lock against the appendage. This is particularly true if the 
