424 CLAUDE FULLER. 



as Hodotermes mossambicus Hagen. In this I think I 

 am right, although it must be admitted that the characters 

 are ehisive. I have considered it safer to regard them, for 

 the time being, as different species, because the terms sub- 

 species, variety, and race are either quite opinionative, and 

 therefore involve the personal equation, or they imply a 

 biological knowledge, and are only correctly used Avhen 

 founded upon a comprehensive study of an extensive series 

 with full details of range, local conditions, and life-history. 

 In the case of each, however, I have been able to examine one 

 or more good series from separate colonies and localities, and 

 have found that, when compared side by side, the soldier caste 

 of each would seem to represent different species. 



The insect which I have specified as H. mossambicus is 

 that to which the descriptions of Hagen^s species, which are 

 available to me, apply best. Further, it is that which, in 

 the collections of the Division of Entomology and the Natal 

 Museum, has been obtained from localities approximating the 

 range of mossambicus. It is represented by material from 

 the moist coast region of Zululand and the dry table-lands 

 of the interior (Kimbei'ley, Marico). In dealing with the 

 series the description of the soldier of H. transvaalensis 

 sp. n. is given as a standard for comparison. Photo-micro- 

 graphs (PI. XXXY) are given of the labrum of each species 

 recorded, and, in order to fix the elusive variations of man- 

 dibles of similar dentation, a series of camera-lucida outlines 

 of these is also furnished (PI. XXXIV). This series shows 

 the mandibles selected from the largest soldiers equally 

 niao"nified. In the illustrations it is seen that the different 

 sets have been variously orientated, but all are arranged 

 so that the basal processes are level. Fui-ther to this the 

 appearance is largely due to the fact that mandibles are 

 articulated in different species in slightly different attitudes. 



I have not regarded any of the Hodotermes as having 

 major and minor castes, either among the soldiers or workers, 

 because in any extensive series of South African forms all 

 o-rades are to be met with between two extremes — the smallest 



