98 PRESIDENTIAL ADDRESS SECTION D. 



(See Plate 8.) Thus, here we have a good example of conver- 

 gence arising through the action of a similar environment. 



In the Indo-Malay region, among the Staphylinid 

 beetles, we find a series leading to the maximum type of Tcnnito- 

 discus; the possible steps are indicated by such forms as Dory- 

 loxenus transfuga, Discoxcmis Icpisnia, and Hcciiiitopscnhis 

 caiidatiis. (Figs. A-D.) In the Ethiopian region the series is 

 less complete, but there is Pygostemits pubescens and a number 

 of species of Termitodiscus. (Figs. E-F.) In South America 

 the genus Eiipsenius is closely similar to Termitodiscus ; and 

 Termitonannus major and Termit op senilis limulus represent pos- 

 sible stages in the evolution. (Figs. G-I. ) 



The short, thickened antenna is a characteristic feature of 

 this defensive type, and intermediate stages between it and a 

 filiform antenna clearly occur. 



Although such a series can be formed, it does not neces- 

 sarily prove that evolution proceeded in the manner indicated, 

 but it is significant that similar series can be traced in the three 

 termite regions. With regard to the South American and the 

 ^Ethiopian regions the question arises whether such a beetle as 

 Eupscnins is genetically connected with Termitodiscus, or 

 whether the similarities observable are due to convergence. 

 With the undoubted convergence seen in the assumption of the 

 defensive shape in various kinds of unrelated arthropod termito- 

 philes (shown in the bottom row of PI. 8), and. bearing in mind 

 that the apparent gradations observed in South America are not 

 identical with those in the Indo-Malay region, it is probable 

 that the defensive forms have arisen inde])endently in the two 

 continents. 



The formation of fine-pointed bristles on the surface of 

 the body is sometimes associated with the defensive shape, and 

 these would doubtless assist in protecting the termitophile. 

 Among adult beetles which do not possess the characteristic 

 defensive shape, a good example of what would appear to be 

 protectivis spininess is seen in Termitopule.r natalensis Wasm.. a 

 termitophile from the nest of Tennes natalensis; here the body 

 is denselv covered with relatively long, very sharp spines. 



It has long been known that insects are very sensitive to 

 certain odours, some being repellent, and other attractive. An 

 interesting paper on the subject has been recently published by 

 N. E. Mclndoo.* The power of producing repellent odours at 

 will is a marked-feature among Staphylinids, and where the 

 termitophile is not otherwise modified, and the relationship with 

 the termite-community is not partcularly friendly, the safety of 

 the intruder doubtless depends on its power of producing an 

 offensive odour. 



2. The Seductive and Mimical Type of TermitophUe. — 

 The presence of certain termitophiles is undoubtedly pleas- 



* Smithsonian Miscellaneous Collections, 68 [2] (1917). 



