:uAv. VI] 



ANIMALS 



151 



lave occupied themselves with myrmecophytes, and who have in some 

 :ascs obtained certain proof of adaptations in them, will prefer to consider 

 Vlyrmecodia and Hydnophytum as m^'rmccophytes. A proof of this 

 •iew has not yet, however, been obtained. 



PliyUoinc-stnicinrcs serving as dwelling-places for ants are even con- 

 iderably more varied than axial parts that are used for the same purpose. 

 They are in some points extremely peculiar ; yet, in all cases coming under 

 his head, the m\'rmecophily is highly conjectural. Even in the certainly 

 nyrmecophilous acacias, only the nutritive corpuscles, and not the hollow 

 tipular thorns, may be considered as indubitable adaptations. 



7. Capura .ilata. Myrmecophyle. 

 garden, Buitenzorg. Natural size. 



Fig. 88, Actinodaphne sp., from Salak. Botanic garden, 

 Buitenzorg. Natural size. 



In many plants the transformation of a leaf or a petiole into a 

 lambered structure suitable for and actually used as an ant-dwelling 

 lay be shown to be associated with other factors, for example in 

 iiphytic ferns, Asclepiadaceae and Bromeliaceae, in which the chambers 

 ;rve for storing water or earth. 



Figs. 87 and (SS are illustrations of plants that I studied in the 

 otanic garden at Buitenzorg, in which mj'rmecophily might more 



