ANTS AND THEIR GUESTS WASMANN". 465 



their giiests. The exudations of the true c^uosts, as, for example, 

 the saccharin containing secretions of the aphids, do not appear to 

 be a food for the nuts, but only an agrooablc stiniidant. 



Among the mynnecophilous Coleoptera there are three principal 

 groups which are prominent on account of their true guest relation 

 to the ants: The Lomechusini among the Staphylinid^e, the Clavi- 

 germse among the Pselapliidse, and finally ajnong the PaussidaB by 

 far the majority of tlie genera from Pleuropterus to Paussus. The 

 remainmg symphiles among the Coleoptera I do not mention here. 



a. The true guest relation is most highly developed with the Lome- 

 chusini, in so far as these beetles are not only licked by their hosts 

 (first step), but also are fed regularly from their mouths (second step), 

 and finally also the larvae of these beetles are reared by the ants like 

 their own brood (third step). The largest representative of the Lome- 

 chusini is the European Lomecliusa strumosa (fig. 13), which lives 

 with Formica sanguinea as its single host and also has its larvae 

 (fig. 14) reared there. These latter, although they possess six legs, 

 imitate in their attitude the immovable larvae of the ants and are 

 fed by their hosts like the ant larvae, indeed even far more eagerly 

 than these. Beyond this, however, they feed themselves, particu- 

 larly in earliest youth, frojn the eggs and young larvae of the ants 

 and devour them in large numbers; on this account they are in fact 

 the worst enemies of their hosts. The species of the genus Atemeles 

 are not, like Lomecliusa, restricted to a single host, but regularly 

 have two hosts. During autumn and whiter the beetles live with 

 the little red ant, Myr^nica, and then in the spring, at the time 

 of propagation, pass over to Formica, where they have their 

 larvae reared; and furthermore every Atemeles species or race has a 

 definite Formica species or race as larval host. The double host 

 relation of Atemeles postulates a much higher degree of initiative of 

 these beetles toward the ants than we find with Lomecliusa. The 

 Atemeles, by "active mimirry," imitate the behavior of the ants to 

 a high degree, particularly in demanding to be fed (fig. 15). The 

 damage which their larva3 inflict on the Formica brood is similar to 

 that of Lomecliusa. In North America the fiomecliusini are repre- 

 sented by the genus Xenodusa (fig. 16), the species of which have a 

 double host relation, like Atemeles, but with Camponotus as second 

 host in place of jVIyrmica. Their larvae are reared with Formica at 

 the expense of the brood of the ant, as in the above genera. 



How seriously the Formica species are harmed by tlie larvas of 

 the Lomechusini is also shown by the fact that through their con- 

 tinued rearing the normal brood-nursing instinct of the ants is 

 patliologically altered — namely, in place of true females they rear 

 malformed individuals, intermediate between workers and females, 

 the so-called pscudogynes, which are perfectly useless for the ant 



