PROC. ENT. SOC. WASH., VOL. 23, NO. 2, FEB., 1921 37 



tera, or any other insects simply because one Crustacean has 

 surpassed these insects in specialization! In any phyogenetic 

 study we must consider the lowest representatives rather than 

 the highly specialized members of a group, giving particular 

 attention to the primitive characters retained in its lowest 

 representatives, and when we consider the Hymenoptera taken 

 as a whole from this point of view, they are seen to be a much 

 more primitive group than they are generally supposed to be, 

 and their line of development apparently arose at the point 

 where the line of development of the Neuroptera began to 

 diverge from that of the Psocida, as shown in the diagram. 



The Hymenoptera are somewhat intermediate, anatomically, 

 between the Psocida and the insects grouped about the Neurop- 

 tera (/. e., Neuroptera, Mecoptera, Diptera, etc.), but their 

 closest affinities are with the latter insects. This is shown by 

 the similarity between the Hymenoptera and the "Neuropter- 

 oids" (/. <?., the insects grouped about the Neuroptera) in the 

 type of metamorphosis exhibited by the two groups, and in the 

 similarity between their pupae and larvae in particular, the 

 larvae of Hymenoptera being very like those of the Mecoptera 

 and Lepidoptera, for example. The head capsule in certain 

 Hymenoptera is very like that of certain Mecoptera and Neurop- 

 tera, and the same is true of the mouthparts. The neck plates 

 and thoracic terga of the Hymenoptera exhibit many points of 

 resemblance to those of the Trichoptera, Mecoptera, and other 

 insects related to the Neuroptera, and the genitalia of certain 

 male Hymenoptera are very suggestive of those of the lower 

 Diptera, Trichoptera and Mecoptera. These and many other 

 features would indicate that the Neuroptera and Mecoptera 

 (with their allies) are the next of kin to the Hymenoptera. On 

 the other hand, the Neuroptera serve to connect the Hymenop- 

 tera with the Coleoptera, although there are many features 

 aside from their mutual resemblance to the Neuroptera, which 

 are clearly indicative of a rather close relationship between the 

 Hymenoptera and the Coleoptera, ami there are very good 

 reasons tor considering that the ancestors of the Hymenoptera 

 were quite like the Coleoptera, on one side of their lineage. On 

 this account, it is necessary to take into consideration the line 

 of development of the Coleoptera also, in attempting to trace 

 the genealogy of the Hymenoptera further. 



As was pointed out in an article recently published in Psyche 

 (Vol. 27, 1920, p. 112) if we take into consideration the character 

 of the larvae (/. e., the head capsule, mouthparts, etc.) the 

 Coleoptera are remarkably similar to the Neuroptera, since it is 

 possible to find but few characters which will serve to distinguish 

 between the larvae of these two groups of insects; and the t\pr 

 of metamphorsis exhibited by the two orders is almost identical, 

 so that if these features are of considerable importance in 



