SOUTH AFIMCAN TKK'M ITKS. 31 



supplied liy the venatiDii he unravelled and its siii'iiilieanee 

 appreciated. 



For the transverse suture that cuts aci-oss the base of the 

 wino" and shar])ly defines the \viny--stnin)) IVoni the lamina, 

 the terni " h'ne of fracture" s<»eins most a ppi-opi-iate, a-< it is 

 here that the lamina l)reaks away from t lu' slum]) when the 

 wing is bout, whetluu- artitieially, accidentally, oi- by the 

 muscular effort of the termite. 



The ultimate fuuctiou of the \vino--stum]) sihmus to be that 

 of protectiug the uiembranous wing-socket throughout the 

 long life of the imago. According to its exteut, or accoi-diug 

 to the degree to which it is chitiuised, it i-eveals or couceals 

 the basal ai"rangement of the veuatiou. In some species, 

 therefoi'e, its characters are helpful, in others confusing. 



In the majority of species the stumps of all four wings are 

 relatively small aud sub-equal ; in others, those of the fore- 

 wings are proportiouately quite large. The first condition is 

 characteristic of the more specialised tertnites, the secoiul is 

 only met with among the more generalised ; but, as certain 

 of the latter (Hodotermes, etc.), exhibit the first, there is 

 really little to indicate which of the two conditions is actually 

 the more primitive. The length of the stump is regulated 

 by the distance from the hinge of the wing at which the 

 fracture line develops, but what governs this distance cannot 

 be stated. Where the stump of the fore wing is considerablv 

 the larger — as in Calorermes and Cryptotermes — the 

 atrophied anal field is within the stump, \vhei"eas it is 

 partitioned in the himl wing, the fracture crossing the curved 

 or oblique furrow which demarcates the anal field. 



The fracture is due to an abscission forming quite late in 

 nymphal life. If its ontogeny is not strictly analogous to the 

 cell wastage leading to the falling of a leaf, it ma}' be compared 

 to what takes place with those plants in which the petioles 

 are cut across at some distance awa}' from the base so that, 

 on the leaves separating, the base of the petiole is left 

 attached to the stem. The development is not connected 

 with any tracheal deviation, as has been suggested. 



