472 ANNUAL EEPOET SMITHSONIAN INSTITUTION, 1912. 



transitions in the size of an ants' nest. Likewise there is hardly a 

 location where ants can not establish their nest, hardly a material of 

 which it can not consist. Through Forel and other investigators the 

 exceeding diversity in the manner of nest construction among the 

 ants has been long known. There are distinguished, earth nests, 

 nests under stones, earth heaps above ground, heaps of dry plant 

 material mixed more or less with earth; nests under bark, in hollow 

 galls, in hollow stalks, and hollow trees; nests in rotten stumps or 

 chiseled out of solid w^ood; carton nests, which are either placed 

 between roots or in hollow trees or hang free from the branches; 

 finally web nests, which may consist of leaves spun together or of 

 other hollows carpeted on the inside with a web. Furthermore, any 

 already existing hollow space may be transformed into an ants' nest, 

 should it be a piece of roofing paper, the cover of a tin of preserves, 

 dried cow manure, or an old skull of a horse, in which last P. Schupp 

 once found a nest of Camponotus rufipes in Rio Grande do Sul. Also 

 numerous are the stolen nests which formerly belonged to other 

 species of ants or to termites and were either taken possession of 

 after having been vacated by the builders or already before that. 



The photograph of a gigantic nest of Formica nifa, near Luxem- 

 burg, which is 17 meters in circumference, may serve as an example 

 of a typical ant hill (fig. 31). A carton nest of Cremastogaster 

 Stadelmanni dolichocephala Santschi, 1.10 meters in length, which is 

 in the Natural History Museum of Luxemburg, is shown in figure 32, 

 as it was photographed in its natural situation, hanging upon a high 

 tree in Kondue, by E. Luja. Finally, a web-nest of Polyrlmchis 

 laboriosa from Kondue (E. Luja) is shown in figure 33. This con- 

 sists of leaves spun together, the surfaces of which are carpeted with 

 web. The upper outer layer, into which wood-mold has been 

 abundantly introduced, appears to be prepared by the ants, as accord- 

 ing to F. Kohl's observations is also the case with Oecophylla longi- 

 noda on the upper Congo, by means of their mandibles and the secre- 

 tion of the mandibular glands, while the web itseK comes from 

 another source with which we shall now become acquainted. 



These web nests of the ants are of high psychological interest. 



For the spinning substance utilized in them does not come from 

 the ants themselves, but from their larvae which the workers grasp with 

 their mouths and employ as ''weaver's shuttle"! They conduct the 

 mouth of the larva, from which the spinning substance issues, from 

 one leaf margin to another and thus weave their nest. When 20 

 years ago Ridley's first information concerning this reached Europe 

 from the East Indies, it sounded hardly credible. Now, however, 

 they have been concord an tly verified by many investigators, for 

 Oecophylla smaragdina in Ceylon through Doflein and Bugnion, for 

 OecophyUa longinoda on the Congo through F. Kohl, etc. With the 



