TRUE GUESTS, ECTO- AND ENTOPARASITES. 



403 



general these tufts are the better developed the greater the reduction 

 in the number of the antennal joints, but Protopaussus is an exception 

 to this rule. The Paussidse are also peculiar in possessing crepitating 

 repugnatorial glands like the bombardier beetles (Brachynus), but dis- 

 tinct from the alimentary tract though opening at the tip of the 

 abdomen. The secretion of these glands in Cerapterus j-maculatus, 

 according to Loman (1887), contains pure iodine! The Paussidae 

 usually live in nests of Pheidole, more rarely in those of Acautholepis, 

 Cremastogaster, Aphcsnogaster and Tctramorium. All these ants are 

 much smaller than the beetles. According to Escherich (19070), who 

 is one of the few who have observed these insects in a living condition, 

 some of the species are treated by the ants as synechthrans, some as 

 synoeketes and some as symphiles. He says : " Paussus turcicus, which 

 I met with in Asia Minor, is extravagantly loved by its hosts (Pheidole 

 pallidula), is continually licked, caressed with the antennae and not 

 infrequently carried about in the nest. Much less affection is lavished 

 by its hosts on the decidedly smaller Paussits favieri, which I studied 

 in North Africa (Oran). In this case it is hardly possible to speak 

 of a friendly relationship, for the ants usually ignore their guest com- 

 pletely, and only occasionally lick it in a per- 

 functory manner. Cooler still, or rather 

 inimical, is the behavior of ants towards 

 Paussus arable us Raff, which I had an op- 

 portunity to study in northern Abyssinia 

 (Erythrea)." Feringuey (1906) has made 

 similar observations on the south African P. 

 lineatus and linncci, the former of which is 

 treated with indifference, the latter with 

 hostility. Both Peringuey and Escherich 

 saw the beetles devouring the ant brood. 

 Since even P. turcicus is not fed by the ants, 

 the species of Paussus cannot be said to have 

 reached the highest stage of symphily. 

 Escherich believes that in this genus synceky 

 and synechthry have been secondarily 

 derived from symphily by a kind of degener- 

 ation, but it seems more probable that 

 these conditions are primitive, especially as they prevail, in all proba- 

 bility, among many of the less specialized genera of the family. The 

 repugnatorial glands, whose very existence suggests the primitively 

 synechthran or synceketic character of the group, are said to function 

 less frequently in the symphilic than in the other species. At any rate, 



FIG. 239. Atcmelcs />- 

 bicollis. (Original.) 



