60 Psyche [June 



structures, unless it be confirmed by an examination of many- 

 others from different parts of the body, since insects which are 

 primitive in respect to one particular feature, may be relatively 

 highly specialized with regard to certain other features, and the 

 same set of structures is not always equally well developed in all 

 insects. On this account, it has been a source of great surprise to 

 me that those who attempt to trace the lines of development of 

 insects confine their attention almost entirely to the difficult wing 

 venation, neglecting other no less vital features (which due to their 

 simpler arrangement are much easier to study), and they are appar- 

 ently unmindful of the fact that, due to the different degree of 

 development of a set of structures in different insects, no one set 

 of structures can be relied upon for such a study, since the evidence 

 must be drawai from all available sources, the evidence drawn from 

 one source merely serving to check that drawn from other sources. 

 On this account, I would present the evidence furnished by a study 

 of the terga and wing bases as merely a portion of the evidence of 

 relationships based upon the studj^ of as many and as widely 

 differing structures as possible, in an effort to determine the lines 

 of descent and the interrelationships of the insects comprising the 

 superorder ' ' Panneuroptera . ' ' 



The Neuroptera appear to be as primitive as any of the insects 

 here considered, and a study of their structures may therefore be 

 taken as the basis for that of the other forms, although the Neurop- 

 teron showTi in Fig. 5 is not as primitive as the Sialidse, etc., and 

 was chosen largely to illustrate the tendency among certain Neu- 

 roptera toward the development of the type of tergum and wing 

 base occurring in some Mecoptera and Diptera. In the Neurop- 

 tera (Fig. 5) the prescutum "psc" and scutellum "si" tend to 

 assume a triangular outline, and the apices of the triangles ap- 

 proach each other near the middle of the tergal plate in which these 

 sclerites occur. The same tendency is apparent in the sawflies 

 (Fig. 1) and in both sawflies and Neuroptera the prescutum "psc" 

 (Figs. 1 and 5) is divided by the middorsal suture "ms" into two 

 symmetrical halves. 



As is indicated by the extent of the broad black line bounding 

 the stumps of the cut off wings in Figs. 1, 2 and 4, the base of the 

 wings is comparatively broad in the sawflies, Trichoptera, Lepi- 

 doptera, and lower Neuroptera, although in the Neuropteron 



