OBSERVATIONS ON SOME SOUTH AFRICAN TERMITES. 



851 



nest and then rise again (text-fig. 9). Here they become 

 greatly reduced in diameter (PI. XXV, fig. 7) and then travel 

 in a horizontal plane, but somewhat erratically, away from 

 the nest. Owing to the labour involved and the indirect 

 route taken, it was found impossible to trace a gallery the 

 whole way either from the nest to its exit at the surface or 

 from a surface burrow to the 



Text-fig. 9. 



iiliiiiiiAi4[rh 



nest. The greatest length 

 of straight gallery exposed 

 was about 30 ft. This rose 

 from a depth of 4 ft. to 

 witliin 2 ft. of the surface, 

 runnino- in horizontal lenyths 

 of 6 to 8 ft. rising by short 

 inclines from one level to the 

 next. Because of the few- 

 ness of pouches in this par- 

 ticular gallery, and its re- 

 lationship to one of the hives, 

 it was regarded as a middle 

 length in a main passage- 

 way. 



Upon reaching to within 

 a foot or so of the surface 

 the galleries branch and in- 

 terlace, some naturally con- 

 necting up with others; fi-om 

 here galleries are driven to 

 the surface, and as many as 



twenty-six openings were counted on one feeding ground 

 10 ft. by 5 ft. Near to the surface of the soil, cavities are 

 made for the temporary storage of the material harvested. 

 These ai'e usually to one side of the gallery and in the form 

 of Avide, flat pockets. Deeper in and also Avhere the tunnels 

 approach the hive there are much lai'ger cavities (6 to 8 in. 

 in diameter). The floors of these gallery extensions are all 

 broadly convex ; they are quite shallow, however, and the 



H o d u t e r m e s 1 1" a ii s a' a a 1 e ii s i s . 

 Section of subteiTaneaii grauaiy 

 shcjwing entrance galleries and 

 descending blind shafts. X o^r;. 



