JUAN FINAS— TENANTS OF B ROM ELI ADS 241 



of the latter as: i. Those which attack the plant itself, 2. 

 Those which feed on the vegetable debris and the fungi 

 which develop there, and 3. Predaceous animals. He has 

 brought together a long list of about 250 species included 

 under these three categories. 



The inter-relations of members of these different elements 

 are interestingly shown by our "fourth bromeliad." The 

 ants brought from it were described by Prof. W. M. Wheeler 

 as a new species, Apterostigma caherti, of which we had also 

 other specimens (although not from a bromeliad), together 

 with fragments of their nest, from the Banana River dis- 

 trict, in early November. We had suspected these latter to 

 be fungus-growing. Professor Wheeler wrote of them, "As 

 you surmise It is a fungus-growing ant of the most primitive 

 and, at the present time, most interesting genus of Attii. 

 No species of this genus has ever been taken in Bromeliads. 

 All of the known species have been described from cavities 

 in rotten wood where they build a peculiar fungus-garden 

 using caterpillar excrement as a substratum, and enveloping 

 the whole garden in a mycelial web, which is not known to 

 exist in any of the other genera of Attiine ants. Your spe- 

 cies is very dark in color, all the other forms being yellow- 

 ish or brownish. The sculpturing of the thorax is more 

 pronounced and the head of a different shape." In his 

 technical description of this ant, whose workers are 3,5- 

 4 mm. and females 4.5 mm. long, he states: "Fragments of 

 one of the fungus-gardens, preserved in alcohol with the 

 ants, seemed to present the same primitive development 

 of the hyphal swellings ('ambrosia') growing on a sub-stra- 

 tum of insect (beetle.?) excrement as described by Mceller 

 for some of the South American species of Apterostigma.^^ 

 We have made no direct observations on the nature of the 

 fungus-garden of Apterostigma calverti in the bromeliad; but 

 in view of Professor Wheeler's statements it is apparent 



