402 PROCEEDINGS OF THE NATIONAL MUSEUM vol.94 



to be closely related to certain Old World genera, as will be discussed 

 later, but otherwise those of the two faunas seem to be very distinct. 

 The other families, Oligembiidae, Anisembiidae, and Teratembiidae, 

 are peculiarly American. 



Conclusions regarding the geographic distribution and phylogeny 

 of the New World species must await more adequate data. It is in- 

 teresting to note, however, that in the Clothodidae we have species 

 exhibiting the most generalized structural features of the order, while 

 in the genera Oligemhia Davis and Ghelicerca Koss some of the highest 

 specialization occurs. Most of the genera are well defined and are 

 often difficult to relate to one another, but in the Anisembiidae it 

 is possible to trace a serial specialization of generic and specific char- 

 acters that seems to correlate with a distributional pattern — the more 

 generalized forms being found in tropical South America and the 

 most highly specialized in environments bordering the Sonoran deserts 

 of North America. As evidenced by the fossil record {Clothoda 

 ■florissantensis Cockerell), the order ranged beyond its present limits 

 during the warm periods of the Tertiary. Most of this migration 

 appears to have come from the south, but there is a possibility that 

 during one of these warm periods one species {Gynemhia tarsalis Ross) 

 came to North America from the Old World, as did so much of its 

 Pacific coast biota, by means of a land bridge in the vicinity of the 

 Bering Strait. This will be more fully discussed hereinafter (p. 497) . 



References to the South and Central American Embioptera are 

 widely scattered in the literature. Navas (1918) made the only 

 attempt to treat in one paper the South American species, but his 

 concepts have been considerably altered by later workers and by the 

 discovery of many additional species. Davis (1939-40) in his "Tax- 

 onomic Notes on the Order Embioptera," has added more than any 

 other one worker to our knowledge of the American Embioptera. His 

 work is made particularly valuable by the fact that he had the oppor- 

 tunity to rcdescribe and figure the type specimens of many American 

 species. 



Before passing to the present treatment it may be well to repeat 

 that the systematics of the order, as they probably always will be, are 

 based almost entirely upon the characters of the mature male. The 

 females are neotenic to a high degree and exhibit few characters. 

 There are as yet no available clear-cut characters that can be used to 

 determine the genus, or even family, of the females or immature 

 specimens. The best means of identifying these is by their definite 

 association with males, although it is possible at times to make deter- 

 minations by a process of elimination based upon geographic distribu- 

 tion, color, size, and number of hind basitarsal sole-bladders. 



