NORTH AMERICAN SPECIES OF THE GENUS ATAX 245 



first two stages are passed in the substance of the gill, and the 

 next apparently in the spaces between the plates of which the gill 

 is composed. This stage is very short and the nymph is soon 

 formed either in the spaces in the gills or outside them in the 

 mucous over their surfaces. In the lacter case the nymphs ac- 

 cumulate in the mass of mucous at the exhalent aperature, and 

 the same species of mite has been reared both from nymphs 

 taken from the gills and from those collected in the latter situ- 

 ation. During the second larval stage the mite is very active 

 and moves freely about between mantle and gills. Generation 

 follows generation in the mussel, the only check to the increase 

 in number of mites being afforded by accidental escape or 

 voluntary migration. The latter, as Kramer (91) observes, 

 usually occurs during the second larval period when the body 

 is smaller, the legs longer, and the mites more active than 

 later. When free from the ancestral mussel they no doubt 

 swim here and there or clamber over the bottom with a chance 

 of finding another mussel and effecting a lodgment therein. 

 However, collections made with nets over beds of mussels have 

 never, so far as the author's experience goes, furnished speci- 

 mens of these migrating mites. 



There seems to be no particular time of year when eggs are 

 deposited as they are found throughout the season, but are ap- 

 parently most numerous in the summer, while during early au- 

 tumn the adult mites are most abundant. There also seems to 

 be no one time at which the eggs from any one female mite are 

 laid, but in the same shell and with but a single female, and in 

 which other conditions lead to the inference that she laid all 

 the eggs, they are found in various stages of development. 



Lodgment is effected probably with different degrees of fa- 

 cility in different species of mussels. The swiftness of the cur- 

 rent seems to have little effect in preventing it, if indeed it 

 does not assist. In lakes with practically no current, collec- 

 tions show the following percentage of those infested, no other 

 conditions being regarded, and all collections being made 

 during August of different years: 



