ANTS AND PLANTS 201 



not absorbent organs, but are adapted to the purpose 

 of protecting the rootlets with which they are filled. 

 No doubt the ants which colonize the pitchers bring 

 in a certain amount of soil, dead insects, leaf particles, 

 and so forth, and from these the rootlets can derive 

 a little nourishment. After a heavy shower of rain 

 the pitchers become more or less filled with water, and 

 some have supposed that they are true insect-traps 

 like the pitchers of Nepenthes; the insects which fall 

 into the water in the pitchers, or the insects already 

 in the pitchers when surprised by the rain-shower, 

 being drowned, and the products of their decomposing 

 bodies being absorbed by the rootlets. As a matter of 

 fact there is practically no evidence for this point of 

 view. Dr. Treub examined numbers of pitchers of D. 

 rafflesiana growing in the Buitenzorg Botanic Gardens, 

 and found that in the wet season most of the pitchers 

 were more or less full of water, but only a very few 

 contained one or two drowned insects, Beccari 

 regards the pitchers as comparable to galls, and as 

 produced by the action of ants, in fact he supposes 

 that if it were not for the ants there would be no 

 pitchers. That the pitchers can only develop after a 

 stimulus applied to leaves by ants is certainly untrue, 

 for the development of normal pitchers has taken 

 place in specimens acclimatized at Kew, to which ants 

 had no access. Treub roundly asserts that the ants, 

 far from being of use to the plant, are positively 

 harmful, for they nibble the rootlets inside the pitchers ; 

 he regards the pitchers as part of the normal develop- 

 ment of the plant and adapted purely and simply to 

 the purpose of protecting the rootlets, and not at all 

 as ant-shelters. It seems more likely that the associa- 



