APPENDIX 



external wall of the basal tumefied part of Balanophora, a part which 

 is always in contact with the humus, from which I suspect that 

 Balanophora and allied plants combine parasitism with saprophytism. 

 By analogy the formicarian Riibiacece would be both epiphytes and 

 saprophytes, for it appears to me quite possible that the surface of the 

 galleries of their tubers may serve for the direct absorption of nutritive 

 substances for the plant. The galleries might also act as subsidiary 

 organs for the absorption of water, as Dr. Treub opines, it being not 

 impossible that at certain moments the aqueous vapour diffused in the 

 atmosphere may condense in them. 



In any case I consider the tubers of hospitating Riibiacece. to be complex 

 organs which primarily originated from a thickening caused by the posi- 

 tion which those plants have sought, and by the necessity of having a 

 reserve of liquids against a dry season. All the other advantages which 

 I have enumerated have come afterwards, and must be considered as 

 secondary adaptations obtained in an indirect way. The tubers of these 

 Riibiacece may thus be organs which originated through the action of 

 simple physiological causes inherent in vegetation, were enlarged later 

 by the hypertrophy produced through irritation caused by the ants, and 

 finally became hereditary. 



The Myrmecodia and Hydnophytum just mentioned are not the only 

 hospitating plants found in Borneo. There are not a few others, and 

 very interesting ones, but these I need not dwell on here. For further 

 information I must refer my reader to my special memoir on the subject 

 already quoted. I must not omit, however, to mention Clerodendron 

 fistulosum, Becc., 1 a handsome plant which has a stem about five feet in 

 height bearing a few big opposite leaves of a purple colour beneath and a 

 terminal bouquet of long tubular white flowers. Each internode of the 

 stem is swollen and shelters internally a colony of ants. Of a similar 

 kind is a euphorbia, Macaranga caladifolia, Becc, 2 a small tree with few 

 branches, which are hollow and formed to shelter ants. This plant 

 has also, like the Clerodendron, large glands on the leaves which secrete 

 a sweet juice, and act as an attraction for the ants, so that they may 

 appreciate and not abandon the plant. Or perhaps it may have been the 

 presence of these glands which first attracted the ants and induced them 

 to form a safe refuge within reach of the prized delicacy. 



There are also palms in Borneo which are provided with permanent 

 shelters for ants. These are the Korthalsias, rotangs or creeping palms, 

 which along the stems, at the insertion of each frond, at the end of the 

 sheath, have an appendage called an " ocrea " or sock, of a paper-like 

 structure, swollen out so as to form an entirely closed receptacle in which 

 certain species of ants find a home, making an entry into the cavity 

 by using their mandibles. These ant-shelters in the inflated ocrese of 

 Korthalsia are permanent and constant organs, undoubtedly not 

 now produced by ants ; but I feel sure that originally (I mean during 

 the plasmative epoch) it was irritation caused by ants which occasioned 

 them. The swollen ocreae of the Korthalsia would thus be hereditary 

 pseudo-galls or ant-homes. In Borneo several species of hospitating 



1 Malesia, vol. ii., pi. iv. 



2 Ibid., vol. ii., p. 45, pi. iii. 



406 



