PRESIDENTIAL ADDRESS SECTION D. Ill 



The above facts suggest the following iiypothesis : — 



Since in the Tertiary Epoch the genera Termes and Euicrmcs 

 were fully established, we must supix)se that in Pretertiary times, 

 when the contour of the great land-masses was not the same as 

 at the present day, a fauna of unspecialised termite genera 

 extended over considerable portions of Australia, Africa, America, 

 r^urope, and Asia. Failing any other direct communication 

 between the Old and New World, we must suppose that the 

 fauna was continued to the north, even as far as Behrings Straits. 

 To explain the distribution of termites it is necessary, as in the 

 case of mammals, at least to assume a land-connection in this 

 region. As the climate became colder, and the land-masses 

 gradually approached their present contour, evolution of new 

 species occurred, especially along the central warmer portions of 

 the world (South America, Africa, India, and Malay Region), 

 while north (North America, Europe and Middle Asia) and 

 south (Australia) the remnant of the old termite fauna persisted 

 and shrank towards the equator. 



It is thus possible to explain the present distribution of 

 termites without the assumption of a land-connection by an 

 Antarctic continent. 



According to this view, the genus Porotcrnnes, which occurs 

 in both Chili and in Australia, is to be regarded as a oortion of 

 the old cosmopolitan fauna which has persisted along the southern 

 border of the termite area. 



The relatively primitive character of the Madagascar termite 

 fauna may be due simply to the fact that a forest country is 

 favourable to Calotermes, but the condition is paralleled by the 

 persistence in Madagascar of the ancient lemur fauna, which was 

 largely replaced on the main-land. 



VI. Conclusions. 



We have seen above that, broadly speaking, the termitophile 

 faunas of the different continents do not appear to be more 

 closely allied to one another than the general free-living . 

 arthropod faunas. We have also observed that termitophiles 

 are apparently scarce among the primitive, unspecialised genera 

 of termites. It would thus appear that the termitophile faunas 

 have mostly arisen since the present geographical separation of 

 the termite regions has taken place. It is significant that the 

 .Ethiopian and the Indo-Malay faunas are more allied than the 

 others, and these regions are more closely connected geographi- 

 cally. 



When once a termitophile has become specialised and 

 adapted to its termite host there would seem little reason why 

 it should become further modified or disappear, and accordingly 

 if a considerable termitophile fauna had developed prior to the- 

 present geographical separation of the regions, we should expect 



