123 



distal pairs were green and still intact. The queen in the lower- 

 most pair was guarding a few larvse and pupœ ; the two others 

 had as yet produced no 3-oung, and, curiously enough, the orifices 

 through which they had excavated and entered the thorns had 

 grown over and closed, though their position was still marked 

 by a scar on the outside. This closure, which I observed also in 

 some of the other seedlings, recalls the conditions in Cecropia, 

 each young internode of which is perforated by the Azteca queen 

 at a preformed pit, which then closes over by the growth of the 

 plant, so that the insect is imprisoned till the wall is again per- 

 forated at the same spot, but from the inside, b^' the worker 

 brood, and the voung colony establishes its cf)mmimication with 

 the outside world. 



It is evident, therefore, that it is the queen Pseudomyrma 

 that attaches the ant colon}' to the acacia, by a type of behaviour 

 which is mereh' repeated by her offspring when they hatch and 

 enlarge the colon}' by excavating and moving into additional 

 thorns as fast as these mature on the more terminal portions of 

 the trunk and branches. My observations also show that even 

 when the seedling acacias grow in localities where the leaf-cutting 

 Attas abound, they cannot be protected b}' the Pseudomyrmœ 

 till they are more than a foot high, for the queens do not leave 

 their thorns, since they are at this period as timorous as all young 

 isolated ant queens. Moreover, the closure of the openings of 

 the thorns would prevent man}' of them from defending the 

 plants, even if they were so inclined. 



A more difficult question is that relating to what mu-t occur 

 when the voung broods, produced b}' the various queens that 

 occupy successive thorns, come forth on to the surface of the 

 plant in search of food. Do these broods fight with one anoth(T 

 till only one and its queen survive, as vox Ihering believes to 

 be the case among the Cecropia Aztecas ? Or do the various 

 broods fraternise and fuse to form a single, large, polycladic 

 colony ? 1 am unfortunatel}' unable to decide between these 

 alternatives, but I am inclined to believe that the various broods 

 unite, and that the thousands of ants which occupy all the thorns 

 on a single tree represent a colon}' which arose b}' a coalition of 

 the broods of all the queen> that peopled the few available thorns 

 on the very young i:)lant plus tlie broods that have been produced 



