252 
Psyche 
[December 
grandidieri Forel of Madagascar, precava of the present study (Fig. 
7)], were primitive types within the genus, but now it seems to me 
that the opposite is true. S. loriae Emery (of the chyzeri group) and 
S. precava are viewed as derivative species with secondarily broadened 
prey specificity, and it is predicted that S. grandidieri will also event- 
ually be found to feed on a wide range of small arthropods instead of 
the usual Strumigenys diet consisting mainly of collembolans. The 
powerful head and mandibles of these species are probably an adapta- 
tion to prey less fragile than Collembola. 
Mandibular armament is probably the best character to use for 
determining direction of descent within Strumigenys. More primitive 
dacetine genera ( Acanthognathus , Orectognathus , Microdaceton) have 
strumigenite mandibles with three long teeth in the apical fork ; often 
the most dorsal of the three is also displaced slightly basad. In cases 
where such displacement has taken place, we have what in the genus 
Strumigenys would be called an apical fork (with two teeth) plus a 
preapical tooth. This is the condition found, with greater or lesser 
modification, in most Indo-Australian members of the genus as well 
as several New World species. In the African group, the species 
judged to be the more primitive ones have two preapical teeth, and 
derivative species mostly are smaller in size and tend to lose one or 
both distal preapical teeth. Quite a few of the New World forms, 
most notably those of the mandibularis group, have two well-developed 
preapical teeth on each mandible. In other New World forms, chiefly 
among smaller species, one or both of these teeth are present in greatly 
reduced form — in fact, in form so greatly reduced as to suggest that 
they serve no present function in holding struggling prey. It seems 
more likely to me that such feeble denticles represent vestiges of 
larger, functional teeth rather than the reverse, especially since so 
many of the species, and particularly the smaller species, have them. 
From this hint (which is no more than that) , I take it that in the New 
World fauna of Strumigenys the mandibularis groups two large pre- 
apical teeth represent the primitive condition. The extensive radiation 
of undoubted mandibularis group species also speaks for a relatively 
long-term existence of this armament pattern. I have accordingly 
placed the mandibularis group at the base of my phyletic scheme (Fig. 
30), despite the very good possibility that the earliest Strumigenys on 
a world basis may have had but a single preapical tooth. 
The mandibularis group shows what appears to be a clear double 
morphocline. Beginning with a more “normal” or “average” species 
such as S. smith'd, a string of species of increasing size and develop- 
ment (width) of occipital lobes, concurrent with a shortening and 
