440 BIOLOGIC RELATIONS BETWEEN PLANTS AND ANTS. 



degree to continue to admit with Beccari that it is the presence of the 

 ants, dwellers upon the Myrmecodias (Iridomyrmex cordata v&r.mynne 

 codice), that leads to the formation of the first gallery in the tubercle. 

 These ants must have perforated the epidermis and penetrated into the 

 central cavity of the hypocotyl, the irritation they produced there 

 determining the hypertrophy of the latter and transforming it into a 

 tubercle. As the tubercle acts as a reservoir of water, the plantlet 

 that is deprived of it is necessarily condemned to perish as soon as the 

 drought affects it. The development of the leaves, which are organs 

 of transpiration, hence occasioning a loss of water, does not occur until 

 after the development of the tubercle, the water storer. Now the ants 

 invade the hypocotolous axis when it is surmounted merely by the two 

 cotyledons. 



Let us now examine another point. Are the ants capable of making 

 perforations and of hollowing out the galleries in the tubercles"? 



The examination of species of Myrmecodia is especially instructive 

 in regard to this matter. M. bullosa shows at the base of its tubercle 

 a few openings (one to four) of quite narrow galleries that end, at the 

 periphery of the organ, in cavities somewhat similar to the cells of a 

 honeycomb. In these recesses there are very numerous colonies of 

 ants, and ventilation is difficult. To avoid the danger of asphyxiation 

 the ants perforate the outer walls of the galleries with minute holes 

 that dot the surface. The surface of the tubercle of M. alata has 

 small gibbosities corresponding to the blind ends of certain galleries 

 (fig. 6). Around these eminences are also seen small, dot-like venti- 

 lating holes that may, by becoming confluent, cause the detachment of 

 the cover closing in the gallery, and thus make a new entrance to the 

 passage. 



Itwould be ascribing a very high degree of intelligence to the ants to 

 admit that they could corrode, from the outside, a circular series of 

 points exactly correspond! ug to the bottom of the gallery. They 

 undoubtedly form their ventilators by working from witbin outward 

 upon the inner wall of the gallery. The corrosive liquor (probably 

 saliva) secreted by the ants not only destroys the cells with which it 

 comes in contact, but also causes the formation of a cicatricial tissue 

 that borders the perforating opening. There also seems to be no doubt 

 but that the ants continually increase the diameter of the galleries 

 when the dimensions of these passages become too narrow for their 

 needs. In fact, once stripped of the dead flocculent tissue that at first 

 fills them, the galleries have a constant tendency to fill up by reason 

 of the internal proliferation of the layer of growing tissue that lines 

 them. The ants must remove this, as it constantly tends to invade the 

 gallery and decrease its caliber. 



It is evident that the presence of galleries within the tubercle does 

 not in any way assist the function of that body as a reservoir of water. 

 Still, it may be admitted that the galleries, as asylums for ants, serve 



