NEUROPTERA. 



165 



portable cases, agreeing with those of most of the " ITydroptilidse " in not showing any 

 difference between tlie two extremities; in one of them the water gains access by 

 a small, ujjright, cylindrical chimney in the middle of the case, in another by small 

 openings along the doi-sal side. The group of Hydroptilidse are divided roughly into 

 six genera, with cases of the following form : — 



1. Cases resembling those of the Euro])ean (and probably North American) species, 

 either naked or covered with fine sand, diatoms, etc, more or less compressed, and 

 with a slit at each end. 2. Very minute, nearly cylindrical, coriaceous, brown tubes 

 covered with line sand, fixed by either end to the underside of stones, and generally 

 showing two adhesive disks at the anterior, and one on the posterior end. 3. Diaulus. 

 Strongly compressed oval cases, elegantly covered by diatoms, with a narrow slit at 

 each end, and with two (or sometimes three) cylindrical chimneys for access of water. 



Fig. 240. — Different forms of caddis worm cases. 



4. In Larjenopsyche the cases resemlile a bottle with the bottom cut away and the 

 lower part compressed until only a slit is visible, the neck representing the mouth-end 

 by which it is held in an upright position. Before transformation to a pupa the case 

 is fastened by disks of silk, and the larva then turns itself in tlie case so as to keep the 

 head of the pupa uppei-most. 5. Cases of younger larvas of Rhyacopsyche are brown, 

 coriaceous, nearly cylindrical, widening in the middle. From one end proceeds a 

 silken thread, sometimes more than twice the length of the case, by which it is fastened 

 to stones so as to prevent its being swept away by the rivulets in which it lives. 

 Before transformation to the pupa the thread is much shortened, so as to sustain the 

 case in an ujjright position. 



The case of IlelicopsycJie merits attention from its resemblance to certain fresh 

 water shells, such as Valoata, for which it has been mistaken even by conchologists of 

 experience. The larvte of IMicopsyche ylabra may be found iu lakes in the north- 



