62 CLASSIFICATION OF INSECTS. 



introduction of the quiescent pupal stage in the Neu- 

 roptera, their obvious resemblances to the Odonata, 

 and the fact that they still retain the Thysanuroid 

 form of larva should not be overlooked. Diagram I. 

 recognizes these similarities, and presents the least 

 modified and most ancient branches of the genealogi- 

 cal tree of the Insecta as near together as practicable. 

 The placing of Thysamira near the centre, by means of 

 a short vertical line,^ indicates the essentially general- 

 ized and larval character of the order, and does not 

 necessarily imply any nearer relationship to Neurop- 

 tera, which stands on the right, than to Coleoptera on 

 the extreme left. The height to which the vertical 

 bars have been carried above the plate is a rough 

 approximation to the specialization attained by the 

 adults, and also to the removal of the mode of de- 

 velopment from the primitive Thysanuroid mode. 



The orders existing to-day are regarded as parallel 

 series differing from each other in structure, and not 

 as yet connected by well-known intermediate forms. 

 Where the probabihty exists that certain orders have 

 had a common origin, they are placed on the same 

 radiating line, as seen in Diagram III., orders II.-III. ; 

 also VI.-VIL, and VIII.-IX. ; and this rule has been 

 departed from only where the data seemed to justify 



genus, and since it has characters allying it both to Orthoptera 

 and Coleoptera, it may be the existing descendant of some com- 

 mon forms from which both of these orders originated. The 

 Thysanura stand, according to Comstock, in a similar position 

 with relation to the Hemiptera. 



1 See also the diagram given by Packard in Thii-d Rep, 

 U. S. Ent. Com., 1883, p. 295. 



