26 
Psyche 
[March 
The genus Myrmoteras appears to belong to an archaic 
branch of the formicine subfamily, and according to Emery 
represents the relicts of an otherwise extinct group. The 
structure of the gizzard is obscure, though it is apparently 
primitive and not developed into the complicated mechan- 
ism found in the higher genera of the Formicinae. From 
Forel’s figure (reproduced in the Genera Insectorum), it 
can be seen that the sepals of the proventriculus are only 
slightly reflected over the anterior end of the bulb, and ap- 
pear to be similar to the condition observable in other 
primitive formicine genera (such as Anoplolepis, Plagio- 
lepis, Melophorus and Myrmelachista) which are also near 
to the base of the stem from where these genera and other 
formicines have arisen. Nevertheless, Myrmoteras and an 
allied genus, Myrmecorhynchus, are considered by Emery 
sufficiently aberrant to be placed in a separate section of 
the subfamily, and his decision is based upon the nature of 
the gizzard. But he states that the organ in Myrmecorhyn- 
chus is more like those of the primitive genera mentioned 
above, and that the proventriculus of Myrmoteras is per- 
haps transitional to the condition seen in the more advanced 
genera. 
The anatomy of the gastric valves in Myrmoteras are too 
poorly understood to allow an unequivocal phylogenetic 
placement of this genus, but whatever its eventual position, 
it is certain that these ants form a very unusual group. 
The most obvious facts in this connection are the elongate 
and sharply dentate mandibles, with trigger hair mech- 
anisms, the modification of the cephalic structure to ac- 
commodate the mandibular muscles, and the enormous eyes 
fitted with very numerous facets. These structural charac- 
ters are clearly associated with feeding habits, as can be 
ascertained from similar and convergently evolved morpho- 
logical traits in other ant genera. These interpretations, 
together with a probably parallel but subsidiary accompani- 
ment of the phenomenon of retrosalience, as deduced by 
Creighton, show that Myrmoteras, though undoubtedly 
archaic, has produced some very specialized structures and 
behavior associated with its nutritional requirements and 
feeding methods. The mixture of primitive and highly ad- 
vanced adaptive features is not peculiar to these ants, but 
