ANTS AND THEIR GUESTS WASMANN. 463 



opment of social parasitism in the northern species; we can only 

 supplement both conjecturally, the former tlii'oiigh comparison with 

 Polyergus among the camponotines, the latter through comparison 

 with the workorless parasitic ants, to which we will now pass on. 



Wlien with a formerly dulotic species like Strongylognathus testaceus, 

 parasitic degeneration proceeds further, its own worker form will 

 finally become completely extinct and will be replaced by that of the 

 auxiliary ant, so that the one-time master species contmues to exist 

 onl}" as males and females. We know a considerable number of 

 such workorless parasitic ants from the palearctic and nearctic regions, 

 recently also one from the East Indies ( WheelerieUa Wroughtoni For.). 

 One of the palearctic species, WheelerieUa SantscMi (fig. 11), has been 

 discovered in Tunisia by Santschi withm the colonies of Monomorium 

 salomonis and possesses, as also the North American genera Epoecus, 

 Sympheidole, and Epipheidole, wmged, still fairly normal, sexual 

 forms. Upon the lowest level of degeneration, however, stands our 

 little, black parasitic ant Anergates atratulus, living with Tetramoriiim, 

 and whose males (fig. 12) are pupa-like, and the fertilized females 

 of which, to unpede the extinction of the species by their fertility, 

 have developed an enormous physogastry. But for none of these 

 workerless parasitic ants can we prove with certamty that their para- 

 sitism has sprung from a former dulosis. There are still three other 

 ways which theoretically lead to the same goal, nameh^: The further 

 development of a former temporary parasitism, the parasitic degen- 

 eration of a former guest relation, and finally the relatively sudden 

 (mutation-like) appearance of a new dimorphism in the female 

 (and later also in the male) of the former parent form and present 

 auxiliary species. In those cases where, for example as in Sym- 

 pheidole, Epipheidole, and Epixenus, the parasitic genus is very 

 similar to the sexual forms of the host genus, the last explanation 

 should even be the most probable. This is also verified through 

 the discovery in Portugal of a new parasitic Pheidole species, 

 Ph. symbiotica, whose males and ergatoid females live in the nests 

 of Pheidole pallidula. In the latter cases we have to assume a 

 relatively rapid origin of the workerless parasitic species, as this 

 has probably never (since its separation from the parent species) 

 possessed a worker form of its own, and therefore also needed no 

 time to "lose" it. In other cases, however, where the parasitic ant 

 departs very -widely from its present auxiliary species and pre- 

 sumably former parent species, there was probably a longer course 

 of development necessary, connected with a real dymg out of its 

 own worker form. This applies, for example, to Anergates atratulus, 

 for which genus we can only conjecturaUy assume Tetramorium as 

 parent form, and for which it is not at all so unlikely that it sprang 

 from a former dulotic form by an intermediate stage similar to that of 



