466 ANNUAL EEPORT SMITHSONIAN INSTITUTION, 1912. 



cojninunity. They are worker-like forms with inflated humped female 

 mesonotum. The rearing of pseudogynes occurs most frequently 

 with our Formica sanguinea (fig. 17, a worker, b pseudogyne), which 

 rear the larvae of the single-hosted LomecJiusa strumosa. Here, too, I 

 was able to prove by the statistical method the connection between 

 the rearing of the adopted larvae and the formation of pseudogynes. 

 The rearing of the larvae of Atemeles and Xenodusa leads to the origin 

 of pseudogynes less frequently because these beetles, in consequence 

 of their double host relation, do not always get back into the same 

 individual Formica colonies in which they have themselves been 

 reared. Through the increase of pseudogynes within a nest the 

 destruction of the host colony is finally brought about. Therefore 

 the avowedly so ''intelligent" Formica in fact actually rear, in the 

 larvae of the Lomechusini, their worst enemies. 



When, however, we follow the phylogenetic development of sym- 

 phily, in which amical selection, that is, the instinctive selection prac- 

 ticed by the ants toward theii* guests, plays a large role, we must even 

 say: In the Lomechusini the ants have brought up for themselves 

 their worst enemies ! To enter more closely into the psychological 

 and phylogenetic phases of this interesting problem here the short 

 time unfortunately prohibits. 



h. The relations of the club-horned beetles (Clavigerinae) to the 

 ants are much more harmless. The beetles are eagerly licked by 

 their hosts and fed from their mouths, but do no harm to the ant 

 brood, although they sometimes gnaw at diseased or wounded larvae. 

 We already know, principally through Raffray's works, 40 genera of 

 Clavigerinae with far above 100 species. The habits of our little 

 yellow Claviger testaceus have become very well known since 1818, 

 and yet the larvae of all the Clavigerinae are still undiscovered. A 

 picture of the adaptational characters of these beetles is offered by 

 the gigantic club-horned beetle, 4 mjn. in length, froni Madagascar, 

 Miroclaviger cervicornis (fig. 18), which, besides a large abdominal 

 cavity, shows richly-developed yellow tufts of hairs on different parts 

 of the body. 



c. The beetle family Paussidae, which is so rich in diversity of form, 

 is very fruitful for the study of myrmecophilous adaptation, but 

 here can be treated only very briefly. Already in the Oligocene of 

 the Baltic amber we find six genera, of which three (Pleuropterus 

 Paussoides and Paussus) probabl}^ at that time already belonged with 

 the true ant-guests, while two others (Arthropterus and Cerapterus) 

 in their representatives of our time still show the primitive protective 

 type. Among the present-day genera we already find symphilous char- 

 acters in Pleuropterus (fig. 19), in spite of the still 10-jointed antennae, 

 in that the cavities of the pronotum and of the bases of the elytra serve 

 as exudatory organs. With the further development of symphily in 



