ants' nests. 499 



it, and as, moreover, the large Formica works with much coarser parti- 

 cles than the puny l^olenopsis, the different character of the walls is at 

 once explained. 



I have already explained the frequent occurrence of imperfect, more 

 accidental compound nests of other species of ant, by ascribing- tbem 

 to the acquisition of favorable localities, especially the underside of 

 stones. From this competition frequently arise very murderous under- 

 ground wars, which I have often watched. 1 have noticed closely, in 

 glass apparatus, how they are carried on. The ants mine toward each 

 other. A battle begins where their work happens to meet. The con- 

 queror forces his way into the gallery of the coufpiered. The latter, 

 however, hastens, after he has retired a few millimeters or centimeters, 

 as tlie case may be, to stop up his gallery thoroughly with earth. The 

 victor does not then, by any means, always succeed in again finding the 

 entrance to it, but, in many cases, mines by the side of it, and thus 

 partial interlappings of the nests arise. The galleries of ^Solcnopsia 

 fuf/a.r are often broken through by the large ants. The little robbers 

 are, however, in the first place, very courageous and combative; and, 

 in the second place, they know how to mine rapidly and how to barri- 

 cade rapidly, and by this means to make a skillful use of all the parti- 

 tions, as I have been enabled to observe directly in the glass nest. 

 The digging and fighting spirit is at its highest pitch among the ants 

 in the first half of the summer, when the nests have to be enlarged for 

 the brood. It then ceases, and truces follow ; in the autumn there is 

 abundant space for all, and peace prevails. It is not without reason 

 that the females and males of Solenopsis fugax do not swarm until Sep- 

 tember, when the swarming time of their host ants (J uly= August) has 

 long been past. They can then, in spite of their size, go to the upper 

 surface of the nest and swarm undisturbed, as I have seen myself, 

 whereas they could not have done so earlier without great danger. 



A peculiar variety of the compound nest is formed by the dwelling 

 of the guest ant Formicoxenus nitidulus Nyl., with Formica rufa and 

 Formica prafcn.sis, which I first discovered in a fragmentary condition, 

 and which Adlerz subsequently found and described more fully. For- 

 micoxenuH hunts the large Formica., and even follows it up closely 

 throughout its changes of abode, as Wasmann first noticed, and as I 

 have verified. By Formica., on the other hand, it is merely tolerated 

 and superciliously ignored. The peaceable guest constructs in the 

 walls of the nest of its large host ant little chambers and i)assages, 

 which are, however, only imperfectly closed, and open freely into the 

 chambers of the Formica. In these little chambers lie the brood of 

 the Formicoxenus. The Formicoxenus' s mode of subsistence is still 

 unknown. 



8.— NESTS OF MIXED COLONIES. 



The mixed colonies of the slaveholding ants and parasite ants {Poly- 

 ergus rufescens Latr., Strongylognathus testaceus Schenk and S. huberi 

 FoREL, Anergates atratidus Sciienk, Xenomyrmex stoUii Fokel) have 



