282 
Psyche 
[December 
is slightly more advanced than Archotermopsis. The various char- 
acters of JJlmeriella seem to be derived directly from primitive 
T ermopsinae. 
The Stolotermitinae has a very primitive left mandible of the 
imago (Ahmad, 1950; Emerson, 1942, 1947), but is derivative in 
numerous other characters, particularly in the sharply slanting 
humeral suture of the forewing. L. R. Cleveland discovered that 
Stolotermes is the only genus of termites that shares a genus of wood- 
ingesting intestinal flagellates with the wood-feeding cockroach, 
Cryptocercus (see Emerson, 1961, for a discussion of the evolutionary 
implications of this discovery). The Stolotermitinae could have 
branched off from primitive Termopsinae or possibly from primitive 
Hodotermitinae, but could not have the Porotermitinae in its an- 
cestry as that subfamily is now conceived, primarily because the Poro- 
termitinae has a much reduced second marginal tooth of the left 
mandible. The Porotermitinae could not be derived from the Stolo- 
termitinae because of the advanced type of humeral suture in the 
latter subfamily, but it could have branched off from the less primitive 
Termopsinae or possibly from the Hodotermitinae although only 
before the specialized harvester behavior had evolved. 
Because of the less advanced type of humeral suture, the fore- 
wing of the Cretatermitinae could not have arisen from the Stolo- 
termitinae which it resembles in some characters. The wing of the 
Cretatermitinae might be related remotely to the Porotermitinae, 
but the data are insufficient to enable us clearly to separate homolo- 
gous from analogous characters, and parallel reduction from homolo- 
gous reduction. With more characters of the species available, it 
may be possible to find consistent patterns of phylogenetic impor- 
tance, but at present it is necessary to state that several alternative 
hypotheses of phylogenetic derivation could account for the meager 
facts. The small size of Cretatermes is reasonably certain to be 
derivative, but the small size of Stolotermes africanus could have 
arisen independently, each from more primitive groups of larger size. 
With our present information, the Hodotermitidae consist of fossil 
and living relicts that are spotty and often discontinuous in their 
distribution. The Porotermitinae and Stolotermitinae, neither rep- 
resented by known fossils, are found only in the southern hemisphere, 
and the genera are probably Mesozoic in origin and dispersal because 
of their existence on different continental masses long isolated from 
one another. The living and fossil Termopsinae are all found in 
the north warm temperate zones with a gradation from the most 
