30 THE CANADIAN ENTOMOLOGIST 



by cephalization of the abdominal ganglia and by adhesion and fusion of the 

 longitudinal connectives. We should, therefore, expect to find the Jugatse 

 exhibiting a tendency towards retaining a larger number of abdominal ganglia 

 and showing a lesser degree of fusion of the connectives, with the opposite 

 tendencies becoming more strongly marked in the higher families of the Frenatae, 

 In the Rhopalocera at least we should expect to find a decided advance over 

 the condition found in the Jugata?. The actual condition is that the Jugatae 

 have one more abdominal ganglion, and are in that respect more primitive; 

 but, on the other hand, they are infinitely more highly modified in that both 

 the thoracic and abdominal connectives are completely fused for their entire 

 length. 



It is usually assumed that the Lepidoptera and Trichoptera are closely 

 related in origin, and that the latter approximate more nearly to the original 

 ancestral type from which both orders have apparently arisen. We should, 

 therefore, expect the larval nervous system of the caddice flies to be more nearly 

 like that of the lowest families of the Lepidoptera. It is interesting to find that 

 the exact opposite is the case; the Caddice flies, as represented by the species 

 dissected, approximate most closely in this regard to the Rhopalocera, and the 

 Hepialidae and Micropterygidas stand out together remarkably distinct from 

 both the Frenatae and the Trichoptera. 



In all the Frenatae the connectives in the last two thoracic segments are 

 widely separated, with the oblique muscles passing out between them. This, 

 supposedly a primitive character, is especially prominent in the Rhopalocera, 

 but occurs throughout the subfamily. Even in the abdomen the double origin 

 of the connectives is indicated by an impressed median line, apparently through- 

 out the Rhopalocera, in the Sphingidae, and variably in other families of the 

 Frenatae. In the Jugatae, however, as represented by the two species of Sthenopis 

 and the eriocranid I have studied, these connectives are absolutely fused through- 

 out the thorax as well as in the abdomen. When it is considered that this 

 modification also involves a great alteration in the relations between the con- 

 nectives and the oblique muscles of the thorax, it would appear that these 

 larvae of the Jugatae are in this important character very much more highly 

 modified than any of the Frenatae. 



On the other hand, the larvae of all the Frenatae appear to be more highly 

 modified than those of the Jugatae in that they have only seven abdominal 

 ganglia in the ventral chain, although the last, the seventh, is always evidently 

 composite, and in some groups partly divided into two ganglia. In Sthenopis, 

 and also in the eriocranid dissected, there are eight abdominal ganglia, the 

 last evidently composite. 



The Trichopterous larvae dissected have only seven abdominal ganglia 

 and have the connectives in the thorax widely separated throughout their 

 length, much as in the Rhopalocera. 



The highly modified condition of the thoracic connectives indicates that 

 the Jugatae were separated from the main stock, from which the Lepidoptera 

 and Trichoptera were derived, at a very early period, even before the separa- 

 tion of the Trichoptera and, following a different line of development, have 

 retained markedly primitive characters in the wing-venation of the adult and 

 the eight distinct abdominal ganglia in the ventral chain of the larva, although 



