The Termites of South Africa 15 



It is not to be inferred that I have any proof that the sub- 

 species recorded are fertile with the type. It is quite possible 

 they are not. On the other hand, the possibilities of *' forms " 

 interbreeding among themselves and with the type are so great 

 that such may reasonably be inferred. 



As I find no two authorities in agreement upon the funda- 

 mental points of a classification, all differing more oi less in their 

 view of family relationships, and even the status of genera and 

 subgenera, I have adopted a group system. This may seem to 

 be a retrograde step, but it is the only alternative. The arrange- 

 ment followed permits of any reshuffling subsequently found desir- 

 able, and will be found workable and non-contentious. More- 

 over, the grouping as here proposed has these immediate, if but 

 local, advantages: — 



(a) The characters of the soldiers will, almost always, give 

 a direct clue to the group to which they belong. 



(b) The group gives a key to the biology of the species in- 

 cluded in it; or, vice versa, the natural habit, when 

 known, will indicate what group the species belongs to. 



Although the termites of the African continent are estimated 

 to represent at least sixteen groups, it would serve no useful pur- 

 pose to notice all here and, for this reason, I have confined myself 

 to those of which we have representatives. On this basis there 

 are ten groups and twenty-one genera. These are: 



1. CALOTERMES Group. 



Calotermes. 

 Neotermes. 

 Cryptotermes. 

 Porotermes. 



2. HODOTERMES Group 



Hodotermes. 



3. PSAMMOTERMES Group 



Psammotermes. 



4. RHINOTERMES Group 



Schedorhinotermes. 



