Syn SSR Sas Foren ae oper 
443 ~Mr.-C."'O; Fag rsinte Five Years’ Observations 
one of the ants intends for the other. I am confident that 
I shall see this happen. It is only a question of waiting 
till I can see the flies side-on. At any rate I am certain 
that only pairs of ants at regurgitation interest them, and 
that they never tackle solitary ants like Harpagomyia. 
Nor indeed do they ever come to rest on the surface of 
the trees as the mosquitoes do. (The mosquitoes keep in 
motion when actually soliciting food, but when not begging 
they frequently settle down on the trunk of the ant-tree.) 
Their hind-legs are backwardly directed when in flight, not 
unlike mosquitoes’ legs, but their flight is rapid and dancing. 
I have spent a good part of to-day at one of the ant-trees 
(an Alstonia) which I do not visit as often as the others, 
for it has not yielded Lycaenids like the others. It is in 
a shadier part than the others and somewhat inaccessible, 
but I have had a path cut to it. It is very rich in Harpa- 
gomyia, but extraordinarily so in the Cecidomyids. 
Aug. 25, 1918.—I spent a good part of to-day at one of 
the ant-trees just making sure of the habits of the Ceci- 
domyids. The web-hangers [Chaetodiplosis gymnastica, see 
below] I feel sure are closely related, though not the same, 
the chief difference being in the proboscis. I am not sure 
that I haven’t found a second species, or a variety of the 
first one [Farquharsonia], that appears to compete with 
a Cremastogaster (on the Teratoneura tree) for the secretions 
of Stictococcus. It does seem a most precarious mode of 
existence. Without doubt, the first ones sent do take 
advantage of the ants in the act of inter-regurgitation. 
I feel confident that the anatomists will agree that the 
extraordinary mouth-parts are for sucking only. 
[Mr. J. E. Collin wrote Feb. 13, 1920: “I believe 
Farquharson was undoubtedly right in considering that 
the proboscis of Farquharsonia is for sucking and not for 
biting. The tip of the tongue-like organ is microscopically 
pubescent and consequently better adapted for sucking 
up liquids than for piercing. Also I found no trace of 
maxillae, which I believe are always present in biting or 
predaceous insects.” 
2. Chaetodiplosis gymmnastica Collin, sp. n. (p. 507). 
[The material includes 8 ¢ 34 2 hanging from threads 
in the hollow at the base of the ant-tree Alstonia, at Moor 
Plantation, Aug. 11, 1918. Also taken with them 1 9 of 
a distinct species. See pp. 508-509. ] 
Aug. 11, 1918.—Near the base [of the ant-tree Alstonza, 
