August. 1944 



Ross: Caddis Flies of Illinois 



19 



scales, with simple venation and only the 

 hypothetical number of crossveins. 



The Trichoptera belong to the holome- 

 tabolous group of orders. The reduced 

 crossveins will distinguish them from the 

 Neuroptera, Megaloptera, Mecoptera and 

 their allies; the two pairs of wings will 

 distinguish them from the Diptera; and the 

 narrow ventral portion of the meso-epi- 

 sternum will distinguish them from the 

 Hymenoptera. 



Members of this order are most closely 

 related to the Lepidoptera, many forms of 

 the two groups being quite similar in gen- 

 eral morphology and wing venation. Almost 

 all Lepidoptera differ from Trichoptera by 

 having a coiled sucking tube which may be 

 very long, fig. 19, or quite short. All those 

 Lepidoptera occurring in this region which 

 have no sucking tube or only a short one 

 have the wings entirely covered with closely 

 packed scales, as in fig. 19. The Trichoptera 

 of this continent do not have a coiled suck- 

 ing tube; most of them have no scales on 

 the wings, but a few species either have 

 patches of scales which do not cover the 

 entire wing or have scales which are scat- 

 tered and interspersed with hair. 



The Trichoptera larvae have a distinct 

 head capsule, full complement of mouth- 

 parts, single-segmented antennae which are 

 often difficult to see, a pair of distinct, single- 

 facet eyes, sclerotized pronotum (mesono- 

 tum and metanotum sometimes sclerotized 

 also), three pairs of distinctly segmented 

 legs, all provided with claws, and a pair of 

 anal hooks. Tracheal gills of various sizes 

 and shapes are sometimes present. The end 

 of the abdomen never has a long mesal 

 process as in some Megaloptera and Coleop- 

 tera. 



Family Groupings 



The Trichoptera are represented in North 

 America by 17 families. This division de- 

 parts in certain respects from the traditional 

 plan of dividing the order but has been 

 necessary because of the following circum- 

 stances. 



1. In attempting to formulate a key to 

 the larvae it was noticed that some sub- 

 families of the Sericostomatidae appeared 

 more closely related to other families than 

 to each other. A tentative key was made 

 up in which the various groups of this 

 nature were treated as separate families. 



2. One difficulty with past keys was the 

 uncertainty of diagnosing a female specimen 

 to family, especially when the specimen was 

 slightly teneral and the venation difficult to 

 determine. A search for other characters 

 which would circumvent the use of wing 

 venation in the key brought to light differ- 

 ences in thoracic sclerites, tarsal claws and 

 arrangement of spines and spurs on the 

 middle legs. Using these in the family key, 

 it was possible to key out both males and 

 females together, and the resultant group- 

 ing agreed almost perfectly with the inde- 

 pendent grouping suggested by the larvae. 



3. The pupae offered only little evidence 

 on the question, but what there was decided- 

 ly favored the new family segregations. 



It appears, therefore, that the old family 

 Sericostomatidae represented a heterogene- 

 ous assemblage of diverse groups such as 

 the Helicopsychidae, Goeridae and Brachy- 

 centridae; these had been considered as one 

 family solely on the basis of a secondary 

 sexual similarity, namely, the three-segment- 

 ed and curved or modified maxillary palpi 

 of the males. Certain other groups, such as 

 the Beraeidae, have been treated as separate 

 families; and the opinion of Betten and 

 others that the Odontoceridae and Calamo- 

 ceratidae are distinct is substantiated by 

 characters of all stages. There are many 

 points to be cleared up in the placement to 

 family of several genera from western 

 states. In addition, many immature stages 

 need to be discovered. These points, how- 

 ever, do not preclude an analysis of the 

 present material. 



The Rhyacophilidae, in particular the 

 genus Rhyacophila, appear to be the most 

 generalized family in the order. The simple 

 wing venation, fig. 21, well developed ocelli, 

 unmodified mouthparts and other characters 

 of the adults, together with the simple type 

 of larvae with either no case or a simple 

 one, seem to represent basic characters from 

 which developed other specializations of the 

 order. 



Three families, Philopotamidae, Psy- 

 chomyiidae and Hydropsychidae, are a nat- 

 ural group differentiated by the annuliform 

 maxillary palpi of the adults and the net- 

 building habits of the larvae. Of these fam- 

 ilies the Philopotamidae appear the most 

 primitive, having diverged relatively little 

 in adult structure from the Rhyacophilidae. 

 The sclerotized larval mesonotum and meta- 

 notum, and larval gills mark the Hydro- 



