40 PROC. ENT. SOC. WASH., VOL. 23, NO. 2, FEB., 1921 



Isoptera to the Plecoptera (with the Embiida) on the one side, 

 and to the Protoblattida (with the Manticia) on the other, is 

 shown in the appended diagram, and simply indicates that the 

 common ancestors of the Dermaptera and Isoptera (together 

 with those of the other insects previously discussed) in turn lead 

 back to precursors which exhibited many of the primitive 

 characters of the Protoblattida and Plecoptera (with their 

 allies) since they were doubtless anatomically intermediate 

 between these two groups. These ancestral forms quickly 

 merge with those of the Plecoptera and Protoblattida as we 

 trace them still further back, and all of them were doubtless 

 eventually derived from forms closely allied to the Palaeodicty- 

 optera, which were among the first winged insects to be evolved 

 from the Apterygota. 



If the relationships as shown in the diagram were correct, we 

 would expect that the Hymenoptera would exhibit a number of 

 features in common with the Neuroptera and Psocida and nearly 

 as many with the Zoraptera and Coleoptera, or with the Isop- 

 tera and Dermaptera; while the resemblances to the Proto- 

 blattida (with the Mantida) and to the Plecoptera (with the 

 Embiida) would be somewhat more remote, and such is indeed 

 the case. Furthermore, the fact that all of these forms exhibit 

 some mutual resemblances, becomes readily comprehensible if 

 we consider that the Psocida, Hymenoptera, Neuroptera, etc., 

 were derived from ancestors like the Isoptera (with the Zorap- 

 tera) and the Dermaptera (with the Coleoptera), and that these 

 in turn were derived from ancestors like the Protoblattida 

 (with the Mantida) and the Plecoptera (with the Embiida), 

 since some of the inherent tendencies of the primitive forms 

 might be carried over into the successive ancestral types derived 

 from them, thereby appearing in certain primitive representa- 

 tives of even such higher types as the Neuroptera, Hymenoptera 

 and Psocida. This view, which takes into consideration the 

 undoubted mutual resemblances occurring in all of the forms 

 discussed above, and which also takes into consideration the 

 very primitive features exhibited by certain of the lower Hymen- 

 optera, is clearly more in accord with the facts than is the case 

 with that proposed by Packard, who would derive the Hymen- 

 optera from Lepidoptera, or that proposed by Smith, who would 

 derive the Hymenoptera from Trichoptera. Furthermore, the 

 views that the Hymenoptera are derived from Neuroptera alone, 

 as proposed by Haeckel, or that the Hymenoptera are derived 

 from Mantida alone, as proposed by Handlirsch, do not take 

 into consideration the fact that Hymenoptera exhibit affinities 

 with several other groups, not with one alone, and in this respect 

 they fail to set forth the true relationships of the Hymenoptera 

 and the other forms under discussion. 



