170 



PSYCHE. 



[Fcbriuiry — March 1SS9. 



some cases aphides may be held as serv- 

 ing their host-plants in the capacity of 

 wandering nectaries (i6, 84). It is 

 certain that they are sometimes a stronger 

 somxe of attraction to ants than either 

 nuptial or extra-nuptial glands. 



3. Myrmecophilous Plants, 



PROPER. 



Some of the earlier travellers, in de- 

 scribing their collections, make mention 

 of the fact that cavities in the stems and 

 stipules, or pockets on the leaves, of 

 some tropical plants are tenanted by 

 ants. This was recorded for Cecropia 

 as early as 1648 by Marcgravius (17), 

 and for Acacia in 1651, by Hernan- 

 dez (12). Though scattering observa- 

 tions of more or less biological interest 

 occur in the literature of the succeeding 

 225 years, it was not until the early part 

 of the last decade that these plants re- 

 ceived careful study. In 1S72 Professor 

 Caruel published a short paper (3) 

 on species of Hydnophytum and 

 jMvrmecodia^ two rubiaceous genera 

 that had been known to harbor ants at 

 least since the studies of Rumpf in 1750. 

 The field notes and material for this 

 paper were obtained from Beccari, who 

 was led to believe that shortly after ger- 

 mination the bases of these plants are 

 pierced by ants (subsequently identified 

 as species of Crefnatog'astc?' and Iri- 

 dojnyrmex) which tunnel the gall-like 

 enlai^gement in various directions, mak- 

 ing a permanent residence there. The 

 plants were even thought to die while 

 quite small if not pierced, though these 

 attacks might have appeared necessarily 



injurious, to another observer. Later 

 observations by Forbes (9) and es- 

 pecially by Treub (33) who has 

 made good use of the unusual facilities 

 afforded by the botanical garden at 

 Buitenzorg in Java for cultivating tropi- 

 cal plants, seem to show that this punct- 

 uring of the stem is not so essential to the 

 life of the plant as was supposed by 

 Beccari ; for in a series of cultures it was 

 found that not only do young plants and 

 seedlings develop ^vhen removed from 

 all possibility of ant visits, but the exca- 

 vations and perforations in their stems 

 appear spontaneously. According to 

 Treub, these elaborate structures, which 

 in a state of nature appear to constantly 

 serve as domiciles for ants, represent in 

 reality a highly developed water-tissue, 

 by which the plant is adapted to its 

 epiphytic habit. The view that they 

 are not primarily connected with the 

 maintainance of a body-guard of ants, is 

 accepted by Schumann (28,419) in the 

 last extensive study of myrmecophilous 

 plants that I have seen, so that what 

 have for years passed for ant-plants 

 par excellence^ seem likely to lose 

 even a subordinate position in the grow- 

 ing category of plants of this class. Yet, 

 as it seems to me, granting the full ac- 

 curacy of Treub's observations, it by 

 no means follows that the curious struc- 

 ture of Myrmecodia and related rubia- 

 ceous plants is not to be looked upon as 

 an adaptation for providing a body-guard 

 of ants with lodging ; and it has been 

 abundantly proved that they are ready 

 to fight whenever they are disturbed in 

 these residences. 



