Notes on the Cases of some South Brazilian Trichoptera. 760 



the surface smooth, and the stones of the wall firmly connected. At either end 

 a few small openings may be detected, leading through the wall. Within the 

 case there is a cylindrical silken cocoon, which is loosely connected with, but 

 may be easily separated from, the walls of the stone-case, and which has a trans- 

 verse sieve at either end. 



Genus II. Tinodes (?). 



Cases similar in general appearance to those of Tinodes niaculicornis are 

 very common on rocks exposed to the spray of waterfalls. They consist of a 

 soft silken ribbon interwoven and covered with microscopical algae, diatoms and 

 mud, and curved into a semicylinder. These canals without a basal wall can 

 hardly be called "tubes". The larva is remarkable for its very long spinneret, 

 which projects beyond the head. I have not yet seen the imago. 



Genus III. 



I do not know the imago; in the pupa I found 2, 4, 4 spurs. The cases, of 

 which I have seen but very few specimens in the River Itajahy, are interesting 

 on account of their close resemblance to those of the Hydroptilideus genus Pelto- 

 psyche; indeed, before I had an opportunity of examining the larvae and pupae I 

 supposed them to be some new species of Peltopsyclie, or even unusually large 

 specimens of Peltopsyclie Maclachlani. They are flat, elliptical, smooth, buff- 

 coloured shields, with a small opening at either end, fixed to the upper side of 

 stones. 



Genus IV. Rhyacophylax, nov. gen. 



(Apears to be nearly related to Smicridea, but the number of spurs is dif- 

 ferent, being i, 4, 4 in the ?, and i, 4, 2 in the <J.) 



This is, no doubt, as to the cases, the most curious of all our Hydropsychidce. 

 The cases themselves are rather rude canals, covered with irregularly-interwoven 

 vegetable fibres, but at its mouth-end each case has a large funnel-shaped verandah, 

 covered with a very beautiful silken net. The larvae live in the rapids of various 

 rivulets, and the entrance of the verandah is invariably directed towards the upper 

 part of the rivulet, so as to intercept any eatable things brought down by the 

 water. Generally, a more or less considerable number of larvae build their cases 

 close together, so as to form transverse rows, on the upper side, of stones. Lately, 

 I saw, on a large stone, about half-a-dozen parallel rows, at some distance from 

 one another; one of them, being about 0.2 m. long, must have been composed of 

 about thirty cases. Before the end of the larval period the vegetable fibres are 

 replaced by small stones, and the verandah is destroyed, either by the larva or 

 by the current of the water. One day, when I was taking to my house a stone 

 with beautiful Rliyacophylax cases, some of the larvaa left their houses, crept to 

 the edge of the stone and then descended, suspending themselves in the air, like 

 spiders, by a thread of silk. The larvae of Grumichella, also, may be seen 

 suspending themselves in the water in a similar way. Such a faculty must prove 

 highly serviceable to larvae living in rapids, where they might otherwise be easily 

 swept away by the current. 



Fritz M u 1 1 e r s gesammelte Schriften. 49 



