190 A NATURALIST IN BORNEO 



formed. The meristem has the power of producing 

 on its outer side new parenchyma cells, and it does 

 so with such vigour that the tuber quickly grows in 

 size, and new galleries are formed in it just as the 

 first gallery was formed ; and this goes on until the 

 tuber becomes nothing more than a scaffolding of laminae 

 separating galleries. When we examine the structure 

 of the swollen rhizomes of the ant-harbouring fern, Poly- 

 podium carnosum [Note, p. 205], we find that the galleries 

 therein are formed in very much the same way, and, more- 

 over, their arrangement is symmetrical to correspond 

 with the symmetry of the fern, and quite unlike the 

 random burrowings of ants in a more or less homo- 

 geneous structure. There is nothing, therefore, very 

 startling in the view that the galleries of Myrmecodia 

 and Hydnophytum are spontaneously formed by the plants 

 in their normal development. But, at the same time, 

 the function of the galleries seems obscure. If the 

 tuber is a reservoir of moisture in time of drought, 

 the removal of a large portion of it by the galleries 

 must surely impair its efficiency. Why not an un- 

 galleried tuber ? Treub regards the cavities as air- 

 shafts, and supposes that the fleshy tuber is aerated 

 by means of these moist chambers without undue loss 

 of water. But this mechanism for exchange of gases 

 is extraordinarily elaborate, its necessity is not proved, 

 and finally the tubers are not green, and therefore 

 cannot assimilate. Until we know more about the 

 physiology of these remarkable plants we must be con- 

 tent to theorize. Between the view of Beccari that 

 these plants cannot develop and cannot flourish with- 

 out the intervention of ants, and the ;view of Treub 

 that ants are not in the least necessary for their de- 



