( XXXV ) 



end is practically within the ant's jaws. I have seen the 

 ant's palpi (not the antennae) vibrating on it. The swollen 

 portion of the proboscis is undoubtedly capable of independent- 

 mo vement. I may say that the mosquito is not ' nervous,' 

 and I have had no difficulty in several cases in standing with 

 my eyes sufficiently close to the two insects to make out the 

 ant palpi and the proboscis movement in the mosquito. The 

 ant raises its head slightly when the exchange takes place." 



Dec. 27, 1917. — ■" On Xmas Day also I made another quaint 

 observation. I have been looking about for other Dipterous 

 myrmecophiles. In my search I came on a large crowd of 

 Cremastogaster sp. on the trunk of a Saman tree (Enter olobium 

 saman) — an introduced shade tree. I thought from the 

 appearance of them that they were about to start up a new 

 nest. All were workers, but I think they came from another 

 large nest not far away. On closer examination I was 

 astonished to see that there were a large number of dead 

 ones sticking to the rough bark of the tree in all sorts of 

 attitudes, but looking as if they had died there and had 

 not been carried up and afterwards ' dumped ' by tired 

 workers. Some were in fact still moribund. I soon noticed 

 some curious little Diptera [Rhynchopsilopa sp., Ephydridae] 

 which I thought might be the same as those I had seen on 

 the Harpagomyia hunt. They are, however, quite different. 

 I soon became interested in their doings, for they settled 

 among the ants, dodging about when a worker approached 

 them, but refusing to go far away. I thought I had only 

 to wait and see more able-bodied beggary. What I saw 

 was something quite new. I saw it repeatedly, as often as 

 I liked, and so tame were they that I could actually study 

 their doings through my pocket lens. When they saw a dead 

 ant, that had expired in such a position that its abdomen 

 was easily accessible, they alighted on it. For such small 

 flies they have a huge proboscis. This they thrust into the 

 appropriate orifice and fed, not this time on stomadeal food, 

 but on proctodeal. Pirouetting neatly on the abdomen of 

 the dead ant, they were themselves — but for the wings, which 

 when in rest stick up from the body at rather a high angle — 

 not unlike ants. When in the act of ingesting the abdomen 



