196 



species {Cccropia adcnopns), but free from these ants, were visited 

 regularly by the leaf- cutting ants and their leaves cut into frag- 

 ments and destroyed. From this fact he drew two inferences: 

 first, that the Cecropia was a favorite of the leaf-cutters; second, 

 that the tree protected itself from their assaults by special adapta- 

 tion to the wants of the protecting ants. A close examination of 

 the anatomy of the tree confirmed the latter inference, showing 

 a most curious and wonderful contrivance to favor the entrance 

 of the protecting ants. The tree itself has been compared to a 

 huge candelabrum, the limbs growing out horizontally at first, 



afterward bending sharply upward with few, but large leaves. It 

 was found that the queen mother entered the hollow limb always at 

 at a certain place. This was through a little depression at the top of 

 the internode. This depression originates first from pressure of the 

 axillary bud; when this is grown out and the pressure thus 

 removed, the outer walls of the depression, instead of incrcasmg 

 in thickness like the ordinary outer walls, remain thin and soft, 

 no hardening takes place, the membrane remaining in this con- 

 dition till after the entrance of the ant has been effected, when 

 the before described abnormal growth takes place. Another 

 species was found to lack these colonies of protecting ants, but m 

 their stead was provided with a wax coating on the outside cells 

 so smooth as to effectually hinder the leaf-cutting ants from 

 reaching the leaves. Now in these trees the same depression is 

 caused by pressure of the axillary bud in the first stages of its 

 growth, but when the pressure is removed the subsequent devel- 

 opment of the thin-lined cavity entirely fails. The ordinary 

 thickening processes take place, and no chance is left for the 

 entrance of the protecting ants. 



Still another difference was discovered between these two 

 species, which Schimper also regards as an adaptation to the 

 needs of the protecting ants. On those Cecropia trees on which 

 they were found, under the petiole of the leaf just at its basis, the 

 surface, for the space of a square centimeter, is covered with vel- 

 vety hairs. On similar trees, lacking these ants, the same clus- 

 ters of hairs are found, but on their surface little egg-shaped 

 bodies loosely connected with the hairs. Fritz Miiller was the 



first to suggest the probable use of these bodies, that is, to serve 



