188 A NATURALIST IN BORNEO 



or feathers of birds hunting in the vicinity of the 

 plants for other food. The plants are all epiphytic, 

 growing upon trees, but not as true parasites like mistle- 

 toe, which sends its rootlets into the bark of the tree on 

 which it grows. The viscous seeds when dropped on the 

 branch of a tree adhere to it and send out little rootlets 

 and leaves. A full-grown specimen of Myrmecodia tubcrosa 

 has a large tuber measuring 30 centimetres in length 

 by 20 in diameter ; its surface is ridged, and on the 

 ridges are sharp spines, homologous with the numerous 

 rootlets whereby the plant is attached to the bough on 

 which it lives. The tuber has the consistency of an 

 apple, and it is tunnelled in every direction by an 

 elaborate system of galleries tenanted by ants, either 

 Cremastogaster difformis or Iridomyrmex cordata, var. 

 myrmecodice. These galleries communicate with the 

 exterior by means of numerous pores, the larger of 

 which are situated at the base of the rootlets. The 

 walls of the galleries are marked with pimples, which 

 are not, as originally supposed, glands secreting a 

 sweet fluid pleasant to ants, but lenticels or rudiment- 

 ary breathing organs. The species of Hydnophytum 

 have large spineless tubers, also galleried, and the leaves 

 are much more fleshy than those of Myrmecodia. 

 There are two views about these remarkable galleried 

 tubers. Beccari holds that the galleries are the work 

 of ants ; that it is impossible for young plants to come 

 to maturity without the intervention of ants ; that their 

 tunnelling causes the tuber to grow enormously, whilst 

 its weight is not proportionally increased ; and the 

 galleries increase the absorbent surface. Treub, on 

 the other hand, regards the ants as of quite secondary 

 importance. He has seen the young plants develop, 



