690 Bulletin Av^erican Museum of Natural History. [Vol. XXIH 



They are always excavated in clayey soil, and the raised entrances, which 

 are more or less cylindrical, are constructed with the particles of earth result- 

 ing from their mining operations and are about an inch in height. In young 

 colonies this entrance leads into a small chamber, about six inches below 

 the surface of the ground, situated not at the end of the gallery but either 

 to the left or right of it. As the colony increases the ants do not enlarge this 

 original chamber, but, piercing its side, form another chamber near it with 

 a small entrance hole. In large colonies, which never consist of more than 

 about 200 individuals, a nest consists of two or three chambers which open 

 on the original excavation. This is no longer used for growing the fungus 

 in, but forms a sort of ante-chamber which generally contains material 

 brought in by the ants to grow their mushrooms on, which is deposited here 

 and gradually made use of. The chambers adjoining are more or less round, 

 with a diameter of about 2--3 inches, and any small roots of plants growing 

 through them are not cut away but used by the ants to hang their mushroom 

 gardens on. These fill the interior of the chamber and consist of a gray 

 spongy mass consisting of a great number of little irregular cells and resem- 

 bling a coarse sponge, amongst which are scattered larvae, pupte and ants. 

 The walls of the cells consist of small round pellets resembling dust shot and 

 are penetrated by and enveloped in white fungus hyphfe, which hold the 

 mass together. Strewn thickly upon the surface of the garden are to be 

 seen round white bodies about a quarter of a millimeter in diameter. These 

 are what Moeller terms "Kohlrabi" clumps, and consist of an aggregation 

 of hyphpe with special swellings at their ends. It is on this that the ants 

 feed. The fungus found by Moeller in the nests of the Brazilian fungus 

 growers (Acromtjrme.r) is the Rozifes gotjgijlopJtora, Moeller, and if it is not 

 the same species cultivated by S. opacus it is, at any rate, very nearly related 

 to it. As material to grow their mushrooms on the ants make use of particles 

 of fruit, flowers, and leaves, but prefer fruit. They do well in artificial 

 nests, constructed on Sir John Lubbock's plan, and are easy to watch. I 

 have tried them with all kinds of vegetable products; they have taken orange, 

 banana, rose petals and leaves and once they even made use of the dried 

 glue from the back of an old book Ivino- near their nest, but that dav thev 

 had nothing else; if the choice be left to them they invariably take fruit and 

 seem to prefer the orange amongst these. Very small particles of the white 

 skin of the oranges are torn off, and after undergoing a slight kneading 

 process in the ants' mandibles, are planted in the nest. The neuters are 

 all of the same size, varying but slightly and never exceed 4 mm. in length. 

 They are more diurnal in their habits than other species of fungus growers, — 

 but also work a little at night. I have found winged forms in the nests in 

 the month of Julv." 



