380 BOTANICAL GAZETTE [DECEMBER 
ants or are unsuitable for their purposes, and that other plants are 
provided with special means of defense. He concluded that the 
so-called bull’s-horn Acacias and other myrmecophytes are pro- 
tected by their guest ants, which drive away their leaf cutting 
relatives. MULLER (6) and ScuimpER (8) endeavored to prove 
that the myrmecophytic species of Cecropia are protected by the 
ants which inhabit their fistulose stems and branches. They found 
that plants which were colonized by Aztecas were not molested by 
Attine ants, whereas uninhabited individuals were more or less 
completely defoliated by them. 
That BELT, MULLER, and ScHmmPeEr tend, on the one hand, to ex- 
aggerate the destructiveness of the leaf cutting ants, and on the other 
hand to overemphasize the protection afforded by the guest ants, has 
been suggested by von IHERING (4), RETTIG (7), ULE(9), FrEBRIG(3), 
WHEELER (10), and other critics of the theory of myrmecophily. 
Several of these investigators call attention to the fact, previously 
noted by MOLLER (5), that the leaf cutting ants feed upon a great 
variety of plants, and show no particular preference for the foliage 
of Cecropia. Thus, although the Attine ants frequently are trouble- 
some pests in gardens and orchards, their feeding habits under 
normal conditions are such that they are not likely to exterminate 
indigenous species. According to von IHERING’s computations, 
183 nests of leaf cutting ants consume no more foliage during a 
year than does a single cow. In many cases the myrmecophytes 
grow in regions, such as swamps and periodically inundated areas, 
where the fungus growing Attas do not occur. Furthermore, 
during the earlier stages of its development Cecropia is not 
inhabited or protected by its putative guardians. In addition, it 
has been shown that plants which are not inhabited by Aztecas 
may remain unmolested by Attas for long periods, and that trees 
which are inhabited may be seriously injured by phytophagous 
insects and sloths. Of course it must be admitted in this connec- 
tion that the discrepancies between the conclusions of MULLER 
and SCHIMPER and those of ULE, von ImERING, and others, may be 
due to the fact that they were concerned with different species of 
Attine ants. Furthermore, it may be argued that the juvenile 
