1891.] ASSOCIATION OF GAMASIDS WITH ANTS. 643 



them belonging to one of the same two genera, viz. Lcelaps or Uropoda, 

 and, as far as I have been able to ascertain, none of them are yet 

 known to science. The most freqnent and the most conspicuous of 

 the former genus was a pyriform species which had its dorsal sur- 

 face dotted at almost regular intervals with wedge-shaped hairs, 

 which give it rather an exceptional appearance ; I propose calling it 

 L. cuneifer. This Acarus was found chiefly deep in the interior of 

 the nest, on the sides and roofs of the passages and chambers, where 

 they were damp without being wet, although sometimes in dry parts, 

 but never in any instance did I find this Lalaps on the Ants them- 

 selves. The Lcelaps were of both sexes and in all stages ; but even 

 the immature stages of the mite were always on the wood, never 

 on the Ant ; yet I never found a specimen except in the Ants' nests. 

 The same remarks will apply to all the Gamasids of the genus 

 Lcelaps which I found in the nests of the Gamponotus. 



I thought that this species would be a favourable one to experi- 

 ment upon, in order to see if I could obtain any idea of the object 

 for which the Gamasid was present in the Ants' nest. In these 

 Gamasince, when the dorsal chitin is thin, as in this species, the 

 principal portions of the alimentary canal, consisting chiefly of the 

 small ventriculus and its four great caeca, can be plainly seen through 

 the dorsal surface as dark objects when they are full of food ; if, 

 however, they are empty they usually become invisible. I collected 

 several suitable specimens of the Gamasid and placed them in the 

 cells which I had formerly used with success in rearing Gamasidce 

 under observation to trace their life-histories — viz. glass rings 

 cemented on to an ordinary 3x1 inch microscopic glass-slip, and with 

 the bottom of the cell thus formed lined with blotting-paper, which 

 is kept moist, and a few pieces of suitable sterilized debris placed in 

 the cell ; the whole is then covered with a second glass-slip, and two 

 elastic bands or a clip added to hold all together. The Gam.asids 

 were healthy when I put them in and their alimentary canals were 

 fall of food ; I placed some living Ants with them and kept them 

 supplied with living Ants only. The alimentary canals of the 

 Gamasids soon ceased to contain food, and were not replenished, 

 while the creatures themselves became weak and unhealthy. I changed 

 my Gamasids, but with similar results. I then tried eggs, larvae, and 

 pupae of the Ants ; but in no case, as far as I could see, did the 

 Gamasids touch them, and their alimentary canals became or 

 remained empty as in the former case. This probably was only 

 what might have been expected, as the Ants would hardly have 

 tolerated in their nests creatures which destroyed themselves or 

 their young ; for such a Gamasid as L. cuneifer would not apparently 

 have any means of defence against so powerful and well-armed an 

 insect as Gamponotus herculeanus. I now tried the experiment of 

 killing adult Ants and putting their fresh dead bodies into the cells ; 

 very shortly the alimentary canals of all the Gamasids became well- 

 filled, and the creatures strong and healthy. I thought, however, 

 it would be better to avoid any possibility of mistake about this ; 

 so I removed the dead Ants and allowed the canals of the Gamasids 



