CADDIS-FLIES 479 



exserted ; when, however, the larva ceased to raove the tracheal 

 gills, then these pouches were protruded. He is inclined to con- 

 sider them blood-gills. Similar structures are found in Uristalis 

 and some other Dipterous larvae that have to breathe under 

 difficulties. 



The imagines of certain species possess filaments — or some- 

 thing of the sort — on the abdomen. Palmen, who has examined 

 these organs in Hydropsyclie, thinks that they are the remains 

 of gills that existed in the larva and pupa, and that they are 

 functionless in the imaginal instar. M'Lachlan thinks that in 

 Diplectrona, where the filaments are elongate, they may be 

 functionally active even in the imago. ^ 



The skin of the nymph is at first very soft, but it soon 

 hardens, and when about fifteen or twenty days have elapsed the 

 nymph opens its case by means of the mandibular processes, and 

 swims through the water with its back downwards till it reaches 

 some soUd object by which it can ascend to the air ; the nymph 

 skin then swells and splits, and the thorax of the imago pro- 

 trudes ; this is soon followed by the disengagement of the head 

 and other parts, and the imago having thus escaped, the nymph 

 skin remains a complete model of the external structure of the 

 nymph, and contains a considerable number of tracheae. This 

 sketch of the metamorphosis of a caddis-fly does not apply 

 in all its details to all the forms of caddis-flies, there being 

 exceptions, as we shall mention hereafter. 



Dewitz has described^ the first appearance and development 

 of the wings in larvae of Phryganeidae. Each one appears at 

 first in the form of a small thickening of the hypodermis, accom- 

 panied outwardly by a minute depression of the chitin (Fig. 

 324, A). He compares the structure -in the earhest stage to the 

 entothoracic projections into the interior of the body. The 

 rudiment grows as the larva increases in size, the chitinous por- 

 tion being duly shed at the ecdyses. When the rudiment is larger 

 and more complex, a mesoderm layer appears in it (Fig. 324, B) ; 

 this is derived from a nerve-sheath near the rudiment. During 

 the resting state of the larva — after its case has been closed, but 

 before the pupal form has appeared — the wing assumes the form 

 and position shown in C, Fig. 324. Dewitz's description of the 

 process leaves much to be desired, and it is doubtful whether in 



1 Tricho]]tera europ. 1878, p. 356, note. ^ Berl. ent. Zeilschr. xxv. 1881, p. 54. 



