ants' nests. 367 



which is constantly secreting albuminous, egg-shaped particles 

 (Miiller's corpuscles). These secretions are eagerly collected and 

 devoured by the Azteca; they are one of their chief articles of 

 food (ascertained through Fritz Miiller). The Cecropia which is 

 free from ants has none of Miiller's corpuscles. The species of 

 Cecropia are much frequented in Brazil by the leaf-cutting ants 

 (species of Attd) and are terribly injured by them, as has been 

 repeatedly ascertained by Belt and others. All those which 

 contain Azteca colonies are spared, because the vicious Azteca 

 pursue the Atta furiously and drive them away. 



All this is well ascertained. The plant, by an undoubted 

 adaptation, gives the ant food and lodging. The ant, in return, 

 defends the plant from its worst enemy. This symbiotic relation 

 did not, of course, arise all at once. Schimper found a Cecropia 

 which is not inhabited by the Azteca until later, and probably also 

 less regularly. This Cecropia has also, it is true, thinned boring 

 spots, but they are not formed until later, and it has not as yet 

 any Miiller's corpuscles. Last year in Bulgaria I watched in oak 

 woods and in old trees in general Liometopum microcephahwi^ Pz., 

 of Europe, which lives in trees. The trunks of the trees are 

 there, too, covered with ants, which attack fiercely all that approach 

 them. We have not in Europe any species of Atta that cut leaves, 

 but, on the other hand, we have so many more beetles and other 

 insects which delight to destroy the old oaks. I was charmed at 

 seeing near Aetos the finest oak forest that I have ever beheld, 

 with real, superb giants. Almost all of them were inhabited by 

 Lio7netopum colonies, whose running workers covered all the trunks 

 of the oaks. I have no doubt that these fierce ants, whose car- 

 nivorous habits Emery has described, drive away the enemies of 

 the oak. The symbiotic relations of the Azteca and the Cecropia 

 were probably formed from these simpler relations. Liometopwn 

 lives only in trees ; the trees, however, do not display the least 

 adaptation to that ant. 



Belt and Schimper have further proved, as to Acacia sphcero- 

 cephala, Willd.. and A. spadicigcra, Cham, and Schlecht, of 

 Central America, that ants of the genus Fseudomyrfua, Lund, not 

 only always live in the hollow thorns, but, owing to a peculiar 

 adaptation of that plant greatly resembling that of the Cecropia, 



