Mermis Parasitic on Ants of the Genus Lasius. 359 



in damp earth ; they rarely have more than one fertile queen in 

 each colony, the females are very much larger than the workers 

 and are reared in large numbers every year, and are therefore 

 specially suitable as hosts ; and, finally, the larvte hatched from egcrs 

 laid in spring and summer live through the winter and do not 

 complete their metamorphosis till the following summer. Thus 

 there are always larv;e in the nests of all three species all the 

 year round. Every colony that I have found to contain Mermis 

 (some 15 to 20 in all) was either in damp ground or near marshy 

 ground or standing watei-, and it is interesting to note tliat 

 Donisthorpe found his colony of nigcr near a stream running from 

 a marsh into the sea. These facts support the conclusion arrived at 

 from my experiments, viz. that water or marshy ground is 

 necessary to the development of the worm after leaving its host. 

 The behaviour of the worm when attempting to escape from the 

 ant's body in water and on moist earth shows that the emergence 

 is more easily effected in water. When the infected ant was 

 placed on moist earth, the Mermis first protruded its head and 

 attempted to burrow under the soil, but the movements of the ant, 

 always very agitated, usually prevented this. After a time the 

 worm succeeded in hooking itself round a projection of earth and 

 held tight until the ant by pulling in all directions drew it clear 

 from her body. This operation often took a considerable time, but 

 when the ant was placed on the surface of the water the evacua- 

 tion usually occupied but a few seconds, or at most a minute or 

 two. The ant after getting free from the worm invariably performed 

 a very complete toilet, and then seemed to take on a renewed 

 lease of life, which, however, in the majority of cases only lasted a 

 few days, and sometimes only twelve hours or so. A normal 

 female of these species after fertilization lives from five to ten 

 years and sometimes much longer. 



It should be mentioned in passing that I brought back from 

 Porlock in May 1920 (i.e. before the winged females were present) 

 a colony of L.jiavus containing two queens, and among the females 

 when they eventually hatched was a single mermithogyne, proving 

 that at least one larva had been infected, though I was unable to 

 find any traces of the worms in numerous full-grown female larvae 

 that I dissected. This colony was used in a further experiment 

 later on. 



The infected colony of Z. niger (the third species of this genus 

 whicli is a host of Mermis) referred to above was very kindly 

 handed over to me by Donisthorpe to enable me to continue the 

 study of the further history of the worm. These mermithogynes 

 are exactly similar, except for specific differences, to those of the 

 other two species, but there is one structural anomaly not present 

 in the former cases. In about nine-tenths of the L. nigcr mer- 

 mithogynes there is a hole in the anterior border of the mesonotum 



