ants' nests. 489 



also the secretion of the posterior glands of certain ants (tlie Doluiliod-- 

 erides, with whom it serves as a weapon for smearing the faces of their 

 enemies), is immediately decomposed at the first contact with the air, 

 with a violent production of gas bubbles and the development of an 

 aromatic odor which is very peculiar. As soon as this chemical decom- 

 position is completed tbe residue of the secretion is transformed into a 

 resinous, viscous mass, which is very sticky. There is no doubt in my 

 mind that the viscous substance formed in this way is not a smelling 

 mucus, as Wolff by a very far-fetched explanation, which is untenable 

 for many other reasons, would have it, but forms the cement with 

 which the nests and many other things are welded together. 



What is still too little known, however, is the manner in which a 

 genuine phylogeuetic evolution converts this gland cement gradually 

 into spun threads. The pasteboard of Lasius fidiginosus Latr. is very 

 rich in wood dust or earthy matter and very poor in cement, so that it 

 is very brittle. There is a drawing of it in my Fourmis de la Suisse, PI. 

 II, figs. 32 and 3.3. The pasteboard which JAometoiyum microcephalum 

 Pz. manufactures in the innermost hollow of venerable but, neverthe- 

 less, strong, handsome, hard, large trees, and which is also composed of 

 wood dust, is somewhat less brittle. They make it in oaks, poplars, 

 apricot trees, etc., in southeastern Europe. Mayr gives a drawing of 

 it, taken from a photograph, in the Proceedings of the Imperial Royal 

 Zoological and Botanical Society of Vienna, June 1, 1802, Vol. XLII, 

 PI. V, fig. 7. A great many species of the genera Cremastogaster Lund 

 and DoUclioderus Lund build only pasteboard nests on the boughs of 

 trees, aiul these nests vary very much in their nature. In some cases 

 the pasteboard is harder and more brittle, resembling wood, as among 

 the si)ecies just described; in other cases it is thinner and more elastic 

 or flexible, but at the same time has much greater power of resistance, 

 and is much more like paper or pasteboard, like thatof wasps. Cremas- 

 togaster stollii Forel, of Guatemala, builds very peculiar galleries of 

 pasteboard along the trunks of trees betw^een the projecting portions of 

 the bark. They were discovered in these galleries by my friend. Pro- 

 fessor Stoll, who communicated this circumstance to me. In Cremasto- 

 gaster ranavalonw Forel, of Madagascar, the pasteboard of the inside 

 of the large, round, tree nest, is thicker and more brittle; that of the 

 outer portion is always thinner, more elastic, and finally, in the outer- 

 most layers, even perforated, having a reticulate appearance, somewhat 

 like loosely woven packing cloth. The nest of Cremastogaster ranava- 

 lona^ is represented in my " Formicides de Madagascar" (from Grandi- 

 dier's Natural History of Madagascar, Vol. XX, part 28, PI. VI, fig 4, 

 4a, and 4&, and PI. VII). The nest of Bolklioderus bisjnnosus Oliv., 

 which is composed of the seed hairs of a tree of tropical America (the 

 wool treej Bomhax ceiba L.) woven together with gland cement, is very 

 similar in appearance to the outer portions of this nest, but still more 

 coarsely perforated and more net like. Fig. 18 represents a small piece 



