TERMITES (WHITE ANTS). 129 



in diameter, and were traced by Dr. Haviland for a distance of twenty feet, and to 

 a depth of five and a half feet, but without coming to a main nest which probably 

 exists. It would seem to be by no means improbable that a more searching investi- 

 gation in tropical Australia will reveal the existence of some allied eyed, day-working, 

 Termitidce also in that Continent. Such a possibility is, at all events, well worth 

 the quest. 



The recent researches of Dr. Haviland, anticipated partly by Bates' exploration 

 in the Amazons Valley, have established the fact that several species of Termites may 

 be found inhabiting the same termitarium. Smeathman, in fact, has attested to 

 two species inhabiting the characteristic turret nests of Senegal. In South Africa, 

 Dr. Haviland reports having found as many as five species of Termitidse and three 

 species of true ants in a single mound. In this instance, while in close proximity, 

 there appeared to be no absolute intimacy between the species. Other instances, 

 however, are reported in which a second species lives as a guest on intimate terms 

 with the original constructors of the mound. This tendency of the Termites to live 

 in mixed communities considerably enhances the difficulty of accurately determining 

 their specific range and identity, more especially having regard to the circumstance 

 of the plurality of individual modifications in the same species. Supposing, for 

 example, that the five consorting species above referred to presented from twelve 

 to fifteen individual modifications ; this would signify that from sixty to as many 

 as seventy-five distinct types might be yielded for discriminative classification from 

 a single termitarium. 



As the result of the investigations of the morphological structure of the 

 Termitidte that have been conducted up to the present date, it has been found that 

 the variations presented by the singularly modified heads and jaws of the soldier 

 individuals yield the most reliable and readily recognisable characters for specific 

 diagnosis. Consequently, in all collections made of these insects for classificatory 

 purposes, special care should be taken to include samples of the soldier types. As an 

 indication of the structural modifications that may obtain in this direction, a few of 

 the more notable ones, as originally indicated by Hagen, are reproduced in Figs. 9 

 to 13, on page 103. The most remarkable deviation from the customary 

 scissors-blade contour of the opposing jaws is presented by Fig. 13, that of Termes 

 armiger, in which the formidable mandibles are replaced by a tubular pike-like 

 prolongation of the front of the head, out of which a fluid secretion can be 

 ejected for the purposes of either assisting in the nest construction, or for the 

 R 



