120 



supply of Beltian bodies, which they were now busily plucking and 

 carrying home, over the barbed-wire bridges, to their nests in the 

 thorns. Later I found that the Beltian bodies are, as a rule, so 

 eagerly sought and so quickly removed from the young leaves of 

 trees inhabited by vigorous colonies, that none of these structures 

 is to be found on the leaflets by the time they unfold. 



The liquid food-supply is derived by the ants from the extra- 

 ñoral nectaries on the upper surface of the leaf-petioles, and in 

 all probabiUty also from the pulp in the young thorns. As in 

 other plants, the nectar is produced most abundantly on the 

 3^oung leaves and in the early morning, so that the ants are most 

 assiduous in collecting the supply at this time, though some of 

 them may be seen exploring and licking the dry surfaces of the 

 nectaries and visiting other parts of the leaves, both old and 

 young, at all hours of the day. 



Shaking or roughly touching the branches at once excites 

 the ants. The shock itself, or possibh' some stridulator}^ signal 

 emitted by the insects that first feel the shock, is transmitted to 

 the ants in the thorns. Without a moment's hesitation the 

 angry creatures issue from the small elliptical apertures, rain 

 down upon the intruder, and thrust their burning stings into his 

 flesh. While stinging the Pseudomyrma curves its body in an 

 arc and bites with its mandibles at the same time, often persist- 

 ing in this position till it is torn away from the skin. The pain 

 thus produced is considerable and may endure for hours, though 

 it is confined to a ver}- limited area. The sting of Ps. spinicola 

 is somewhat more painful than that of Ps. belli or its subspecies 

 fulvescens. 



The foregoing observations agree very closely with Belt's, 

 and certainly at first sight suggest an intimate symbiotic rela- 

 tionship between the ants and the acacias. There is, of course, 

 nothing remarkable in the ants' utilising the nectar and food- 

 bodies, because almost any dendrophilous ants would do this, 

 but the uniform and purposeful method of excavating and 

 inhabiting the thorns certainl}' implies a singular degree of 

 familiarity with the suitability and consistency of these structures. 

 But the matter assumes a different aspect when we consider Ps. 

 gracilis. This ant, which is highly variable in colour and one 

 of the largest and most abundant species of the genus throughout 



