1891.] ASSOCIATION OF GAMASIDS WITH ANTS. 645 



I did not find this Uropoda in the nests of any other Ant, but 

 Mr. Bostock has since found it in England in the nests of Formica 

 fusca, where also it is found on the cocoons and in the nest. 



There was one ver}' good nest of Camjionotus herculeanus, nearly 

 a thousand feet above Igls, which I kept as a kind of store-house 

 for some time ; the tree had been cut down near the ground and the 

 greater part of the stump was beneath the surface or beneath the 

 fallen debris of the forest. I used to dig and cut down into the 

 centre of this nest, take home parts for examination, and carefully 

 put back and cover up the remainder. In this way I kept the Ants 

 in it for some considerable time, and as long as the Ants remained 

 the Gamasids were to be found there ; but at last the Ants seemed 

 to get tired of being so frequently disturbed, at any rate they 

 abandoned the nest, and from that time the Gamasids vanished also. 

 I could not find any more. I also frequently examined other 

 abandoned nests, but I never found the Gamasids in them except 

 possibly a single specimen once or twice. 



It will thus be seen that I have found seven species of Gamasince 

 and two of Uropoda in Ants' nests (one species previously found by 

 Berlese), and that two other species have been found by others ; that 

 none of these have been found elsewhere ; and that, so far as has 

 been ascertained, each Gamasid was associated with one or two 

 particular species of Ants only. 



I did not find any other Acariua in the nests except a few 

 Oribatidae, which were in much larger numbers on other stumps 

 where the Ants were not present. 



From the above observations I come to the following conclusions : — 



1. That there is an association between various species of 

 Gamasince and certain Ants. 



2. That one species of Gamasid usually associates with one or two 

 special species of Ant only, or at least preferentially, although this 

 may be a little affected by the presence or absence of the Ant in 

 different localities. 



3. That the Gamasids found in Ants' nests are not usually to be 

 found elsewhere, except probably rare and scattered specimens on 

 careful search. 



4. That the Gamasids usually abandon the nest if the Ant 

 does so. 



5. That the Gamasids live upon friendly terms with the Ants, 

 who do not attack them and even show signs of taking care of them. 



6. That the Gamasids are not true parasites and do not reside 

 upon the bodies of the Ants. 



7. That, in the cases investigated, the Gamasids do not kill or 

 injure the Ants or their young. 



8. That the Gamasids will eat the dead Ants. 



9. That the Gamasince are not improbably either scavengers or 

 else messmates sharing the feast off any insects which the Ants 

 may kill. 



10. That we do not know what the Uropodince teed on nor the 

 object of their presence in the nests. 



