the HydroptilidcB. 131 
as tall plants and slirubs, palings and walls, present great 
attractions to them, and so does liglit. They retire to rest 
mostly amongst the grass and herbs, though crevices in 
bark and the uneven surface of shaded walls are frequent 
resorts of theirs during repose. Their vivacity is remark- 
able. As soon as they alight upon the ground they rush 
along helter-skelter with unusual impetuosity, flying off* 
for a short flight from time to time whilst in full career, 
or coming to a dead stop to rest awhile with antennae 
ascending and slightly separated (as suddenly as Stenoj)- 
socus immaciilatus is in the habit of halting), and then 
starting off" hurry-skurry once more. In deep repose some 
genera place their antennas in a different posture to that 
Avhicli they assume in these short pauses. The species 
inhabit rivers and lakes. Their localities are sometimes 
confined to the neighbourhood of rapids in rivers. 
The larva inhabits a portable case. 
I am inclined to remove the Hydroptilidce from the 
position usually assigned to them between the Sericosto- 
mida and the Lejjtoceridm, and to place them beside the 
Rhyacophilida. The structure of the head and of the 
ovipositor, the horny ventral process of the antepenulti- 
mate segment of the male and its rudiment in the female 
of nearly all the genera, and the presence in some genera 
of an appendage to the anterior wing near the base of the 
post-costa, are the chief points in the imago which appear 
to indicate affinity between these two families. 
But it may be objected to this view that the Hydropti- 
lidcB are ranked Avith the Inceqiiipalpida, whereas the 
RhyacophilidcB are amongst the ^quipalpidce. So far, 
however, as I have been able to ascertain, the maxillary 
palpi of the male are five-jointed in the Hydroptilidce, as 
Pictet stated. The fifth joint can be demonstrated with 
a Codrington lens Avithout much difficulty in the male of 
Ayr ay lea, and even in Fhrixocoma under favourable cir- 
cumstances. But if a § object glass and a compound 
microscope be brought to bear upon specimens immersed 
in fluid, the palpus is seen to be obviously five-jointed in 
both sexes in all the genera, although in specimens freshly 
killed and viewed in air it may have seemed to consist of 
only four joints. And so the HydroptilidcB being a family 
of the JEqtdpalpidcE this objection is disposed of; and I am 
not aware of any remaining to be raised against the pro- 
posal to rank them close beside the RhyacophilidcB. 
Since this was written I showed my figures to Mr. 
