180 SCIENCE PROGRESS. 



vastly increased. 1 And it has been frequently assumed that 

 in all these cases there is a social or symbiotic relation 

 existing between the plants and the ants. Particularly is 

 this the case with those remarkable epiphytic Rubiacese, 

 Myrmecodia and Hydnophytum, described by Rumphius, 

 and which always harbour large numbers of ants in the 

 numerous cavities of their tubers. Beccari recognised in 

 this a marked case of symbiosis, stating that the tubers only 

 arose in consequence of the effect which the ants exercise 

 on the young plant ! But no one hitherto has been able 

 to trace the enemies against which the Myrmecodias are 

 protected by their guests, and there are no leaf-cutting 

 ants in the districts where Myrmecodias grow. The ex- 

 amination of a fern inhabited by ants, such as the Poly- 

 podium sinuosum, shows 2 that the hollows, which the 

 ants inhabited, originated from the disappearance of a 

 watery tissue, in which water is stored up for reserve. 

 Such a reservoir of water must be especially necessary 

 for epiphytic plants, as it is more difficult for them to con- 

 stantly obtain it than is the case with terrestrial forms. 

 I have put forward the same theory with regard to Myr- 

 mecodias, and in this the distinguished botanist Dr. 

 Treub 3 agrees with me. He also showed that the ants 

 did not exercise any influence on the forming of tubers. 

 The occurrence of ants in these plants seems to be in this 

 case only incidental ; 4 and if we suggest that they were 

 perhaps useful in former times, it may be said, on the 

 contrary, that we may only legitimately take refuge in 

 such a hypothesis when we can find no other explana- 

 tion. The suggestion of a botanist to settle a protecting 

 army of ants on our fruit-trees, by means of sweet-tasting 

 matter, in order to destroy caterpillars, shows clearly how 



1 Such cases are seen in several ferns, not only in Pteris aquilina but 

 also in some species of Polypodium, further in Platyceriuvi. 



2 Annates dujardin. botanique de Buitenzorg, vol. vii., p. 18. 



:f Nouvelles recherches sur la Myrmecodia de Java, ibid., p. 191. 



4 As also the occurrence of rotifers in the auricles of some Liver- 

 worts (cf. Ueber die Blattbildung der Lebermoose, Flora, 1893). 



