1962] 
Brown — Strumigenys 
253 
thickening of the mandibles, leads through S. planeti and S. godmani 
to S. mandibularis. In the other direction, we find a trend toward 
lengthening of the mandibles through the series S. biolleyi, S. saliens, 
S. sanctipauli and so on to the species near S. cordovensis , climaxed by 
the remarkable S. dolichognatha , the mandibles of which are relatively 
longer than in any other ant known to me. Side offshoots of the 
mandibularis group are species such as S. borgmeieri and S. trinidaden- 
sis; the greatly weakened proximal preapical tooth of the last species 
shows the first stages of a trend that apparently led to groups such as 
the hindenburgi and emeryi assemblages, and beyond these to the 
elongata and silvestrii groups respectively. Species such as S. perparva 
and S. ogloblini , both of which have a single preapical tooth on each 
mandible, were previously grouped together, but now I think it more 
likely that their similarities are due to convergence. Such highly 
reduced species are doubtfully placed at best. 
The emeryi group, especially S. emeryi itself, is linked to the lo'uisi- 
anae group by the virtually perfect intermediate S. mixta. The louisi- 
anae group leads to the connectens group and through this to the 
gundlachi group. These last three groups all have two (or rarely 
more) intercalary denticles between the main teeth of the apical fork. 
The genus N eostruma represents a further development of the louisi- 
anae groups connectens group-> gundlachi group trend or morpho- 
cline. 
The three remaining species groups, all small, appear to be derivable 
directly from the mandibularis group: the tococae group by addition 
of a second intercalary denticle in the apical fork, the cultriger group 
by development of a mandibular lamella, and the ludia group by the 
serial loss of mandibular teeth. 
Identification of Species 
This section is intended to provide materials with which any compe- 
tent entomologist can hope to identify quickly and surely the Strumi- 
genys species at present known from the New World. Of course, 
there are certainly species remaining to be discovered in this hemi- 
sphere, but I believe that we now know all or nearly all of the species 
that are both widespread and reasonably common, and many of the 
rare or local species as well. 
Before discussing the species, though, it is necessary that we charac- 
terize the genus Strumigenys well enough to recognize it in this hemi- 
sphere. It will be enough to say that any New World ant with the 
following combination of characters is a Strumigenys : Worker and 
