14 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.64. 



Both of these new species convexinotatus and cardini (as well as 

 tenuis Hagen) have the ocelli variably present or absent in specimens 

 from the same colony and show variations in the size and shape of 

 antennal segments and in wing venation (branching of the veins and 

 area occupied by the cubitus and branches) ; the wing scale covers 

 the meso-notum in both new species and the meso- and metanota 

 are convex, while they are emarginate in tenuis. In convexinotatus 

 the median vein runs nearer to the middle (more intermediate between 

 the subcostal vein and cubitus) than in tenuis and cardini, where the 

 median is nearer to the cubitus. In convexinotatus the cubitus and 

 branches take up less area of the wing than in tenuis and cardini — 

 that is, the distance between the cubitus and lower margin of the 

 wing is less; there are also specific differences in the number of hairs 

 on the surface of the wings (pi. 3). The soldiers of my three new 

 species have the heads less hairy than in tenuis, but the soldier of 

 cardini has a few more hairs on the head and pronotum than that of 

 convexinotatus. The soldier of L. crinitus Emerson MS. has an ex- 

 tremely hairy head. 



To summarize, it is evident that there is either a single extremely 

 variable species involved, or a species in. the process of evolving 

 several nascent species, or a complex series of very closely related 

 species, there being two extremes tenuis Hagen and convexinotatus 

 Snyder with intergrading connecting species which display charac- 

 ters of either extreme. 



In 1902'" Wasmann mentions the soldier of Leucotermes Jlavipes, 

 subspecies paraensis from Para, Brazil; this is probably L. tenuis 

 Hagen or a related species of Leucotermes; he also describes (p. 140) 

 the species insularis from "Oceania," which may be my species 

 convexinotatus, since soldiers from the Galapagos Islands are the 

 same as convexinotatus. It is probable that the species insularis 

 from the Cocos Islands should be the same as that found in the 

 Galapagos and Central America. Wasmann's description is too 

 meager to enable or warrant precise identification. Mr. A. Emerson 

 has kindly loaned me specimens of soldiers collected by the Williams 

 Galapagos Expedition. 



A soldier in the Hagen collection" is lah'Bled Termes corticolaBates. 

 The head is very hairy (the hairs being long) . The anterior margin 

 of the pronotum is emarginate; the posterior margin is slightly con- 

 cave. This specimen is from Obispo, Panama, and is probably 

 Leucotermes tenuis Hagen. Hagen writes that the same soldier 

 occurs in Bates' termites from Santarem, Brazil. 



" 1902. Wasmann, E. Termiten, Termitophilen und Myrmekophilen Gesammelt auf Ceylon, von Dr 

 W. Horn, 1899, etc. Zool. Jahrbuch., vol. 17, Heft 1, pp. 99-164 (p. 119). 

 11 Hagen, Linn. Ent., vol. 12, p. 203. 



