BIOLOGIC RELATIONS BETWEEN PLANTS AND ANTS. 449 



tible droplets in the interior of the pitcher may be reabsorbed by the 

 slender radicles applied to its internal wall. The rain water gathered 

 in the pendent pitchers evaporates slowly through the narrow month, 

 and thus constitutes a reserve that may be absorbed by the rootlet. 



The ants may, it is true, be indirectly useful to the plant by protecting 

 it against the attacks of phytophagous creatures. The pitcher presents, 

 indeed, various features favorable to the life of the ant. In particular, 

 the rain water, which does not enter in sufficient quantities to become 

 dangerous to the inhabitants, may probably be to their advantage, for 

 the ants that inhabit these pitchers are fond of water. But the only 

 protection the ant can offer in exchange for the shelter afforded by the 

 plant is that which we have before mentioned. Even in this connec 

 tion "nothing authorizes us to suppose that the colonies of ants exer- 

 cise a salutary influence upon the plant" (Treub). Indeed, the ants 

 when they multiply too greatly in a pitcher gnaw the rootlets, or, 

 indeed, cause their malformation. Still, Beccari has seen the DiscMdias 

 form iuextricable masses of pendent branches on the surface of trees, 

 which masses were so well defended by the ants and termites that 

 inhabit them that it was impossible to put the hand upon them. 



If at the present time the ants play no part in the normal evolution 

 of the leaves that become pitchers, they yet may have had something 

 to do with tlieir original formation. In a related genus, Dkchidia- 

 Conchophyllum, and in many species of Dischidia, all the leaves are 

 indistinctly suborbicular or reniform, convex on one face, concave on 

 the other, like a watch crystal, and applied against the bark of the 

 tree that serves them for support. Their inferior concave face is purple 

 (flg. 9, PI. XXII). At the level of the leaves the branches give off adven- 

 titious roots, very much ramified and sheltered under the concavity of the 

 leaves. These roots, arising near the insertion of the petioles, divide 

 dichotomously and serve, some to cause the plant to adhere to the bark 

 of the tree it inhabits and to absorb the nutritive matter they may find 

 there, others, covered by the leaves, to absorb water, for which they 

 are more especially designed. 



The ascidiferous Dischidias are certainly derived from types with 

 reniform leaves like those we have just described. Xow, the inferior 

 face of these concave leaves is often inhabited by acarids, and we may 

 suppose, with Beccari, that the irritation produced by these parasites 

 has caused a more marked concavity in the organ that shelters them. 

 There would be primitively formed there true galls, and the deforma- 

 tion of the leaf might become hereditary in the course of time. Ants, 

 finding shelter under these concave leaves applied to the surface of the 

 trees, have chosen a domicile there, profiting thus by an organ of lodg- 

 ment whose abnormal evolution may have had for its primordial cause 

 the adaption of the plant to the struggle against drought or parasitism, 

 or perhaps both causes combined. As soon as they were installed 

 under these lodging organs, the ants became useful to the plants by 

 SM 96 29 



