BIOLOGIC RELATIONS BETWEEN PLANTS AND ANTS. 417 



more, and makes the entrance near the summit — a wise precaution that 

 is also shown in the location of their nests, which are placed beyond 

 danger from inundations. 



The ants then destroy the herbage entirely around the mound, level- 

 ing the surface for 3 or 4 feet all about the nest. Within this sort of 

 paved area no growth is tolerated except that of a species of grass 

 (Aristida stricta). This plant is sown all about the nest, while other 

 plants tli at start up in the vicinity are pitilessly gnawed. The grass, 

 thus aided in its struggle for existence, gives an abundant harvest of 

 small, white, flinty seeds quite similar to rice (ant rice). The plant is 

 harvested a little before maturity, bundled up and carried into the nest. 

 There the grain is separated from the husk, which is thrown out beyond 

 the paved area. If the ants are surprised by an early setting in of the 

 rainy season, their stores may be dampened. They then dry them in 

 the sun, preserve the sound grains and store them anew. 



Lincecum and Darwin thought there was no doubt but that this 

 species of grass was planted designedly. McCook affirms, however, 

 that it is not sown by the ants themselves, these insects merely pre- 

 venting any other species of plant from growing around their nest. In 

 autumn, after the harvest, the paved area is abandoned until the ensu- 

 ing autumn, when the grass again springs up, appearing about the ant- 

 hill in the same ring-like form, and is cared for by the ants in the same 

 manner. 



Mrs. Treat and McCook have also studied with the greatest care other 

 species of Pogonomyrmeoc in Texas (P. crudelis and P. occidentalls) 

 having similar habits. One of these collects the fruits of Composite. 



We see already by these examples that the vegetarian ants may, 

 notwithstanding the ravages they commit upon certain plants, yet aid 

 to a certain extent in their dissemination. The loss of a considerable 

 quan 'ty of seeds is compensated for by the dissemination of those 

 which, among the number collected, are necessarily overlooked by the 

 ants. The insects, especially the agricultural ants, manifestly aid the 

 plants of their choice in their struggle with the neighboring species 

 whose physicochemical requirements are the same. Certain tropical 

 plants make use of agricultural ants for the dissemination of their seeds ; 

 but, far from furnishing an aliment in return, they deceive their assist- 

 ants by the resemblance of their seeds to those of plants they are in the 

 habit of gathering. 



Sometimes, also, the insect is led into error by the resemblance of 

 seeds to the nymphal cocoons (vulgo, egg) of the ants. It is true that 

 the ants find under the leaves of these plants a saccharine liquid of 

 which they are very fond. An instance of this is our common Melam- 

 pyrum pratense, that often grows in the middle of ant-hills. Its dehis- 

 cent capsule contains a single seed, smooth and white, bearing a most 

 deceptive resemblance to the cocoons inclosing the nymphse. The 

 ants are deceived by this appearance, and bury these seeds with the 

 sm 96 27 



