152 



THE FACTORS 



[Part 



readily be suspected. In Capura alata the large spoon-shaped stipule 

 of each leaf arc bent in such a manner as to enclose within then 

 a chamber-like space, the marginal gap of which is closed as far as ai 

 entrance-aperture by a kind of web spun by the ants. I found th( 

 chambers nearlj' always inhabited by ants. Still more peculiar anc 

 requiring morphological investigation was the state of matters in a tree 

 described as Actinodaphne sp. coming from Salak, in which each twis 

 terminates in a roomy chamber lying above the minute terminal buc 

 and formed by a whorl of small scale-leaves. The phyllomes describee 

 here as scale-leaves are distinguished from the foliage-leaves by mucl 

 smaller size, absence of petiole, and different shape. I have ahvay 

 found the chambers inhabited by ants, which appeared to belong t( 

 , ^ a species very abundant else 



where in the garden. It ii 

 no wonder that such suitablil 

 structures should be inhabitec 

 by ants ; it appears mucl 

 more remarkable that the hoi 

 low stems of Triplaris ant 

 Humboldtia, provided wit! 

 entrance-apertures, should b 

 free from ants in the gardei 

 of Buitenzorg, at least so fa 

 as my observations go. 



Still further deviations froi: 

 the normal leaf-structure occuj 

 among the tropical America' 

 Melastomaceae in the gener 

 Tococa, Maieta, Calophys; 

 Myrmidone, and Microphys; 

 as well as, according to Schu 

 mann, in the rubiaceous Remijia physophora and Duroia saccifera, and th 

 tropical African sterculiaceous Cola Marsupium '. Here at the base c 

 the blade of the leaf, on both sides of the midrib and sometimes also c 

 the petiole, are found two hollow outgrowths (Fig. (S9), which in th 

 Melastomaceae are situated on the under surface of the leaf and are tj 

 be regarded as modified domatia, whilst in Duroia they belong to th! 

 upper surface and morphologically represent new structures. 



All the above-mentioned plants, and others besides that are inhabite 

 by ants, are, as Schumann first pointed out, provided with an abundar 

 brownish-red coat of hairs, which appears in some way to be connectc 

 with the symbiosis. 



' See Schumann, I, regarding all these plants. 



Fig. Sg. Tococa lancifolia. Base of leaf with 



utricles. A seen from below, showing the entrance a. 



, £ seen from above. Natural size. After K. Schumann. 



