EVOLUTION 343 



gota ; this is Sycopteron from Commentry described by H. 

 Bolton (19 1 7) as a mecopteroid fly. In wing-neuration these 

 insects show features of " the primitive mecopteroid plan " 

 combined with *' strong hymenopterous characters," such 

 as approximation of the radial system of nervures to the costa 

 of the wing and the consequent development of the distinctive 

 pterostigma (Plate XII, D). From these Protohymenoptera 

 of the Carboniferous and Permian there is a long gap in 

 the record to the earliest known fossils that are certainly 

 referable to the Hymenoptera. These come from the 

 European Jurassic, and are sawfly-like insects of the sub- 

 order Symphyta. The highly organised Hymenoptera of 

 the sub- order Apocrita are not knov/n to have lived before 

 the Tertiary era ; ants and other existing families are repre- 

 sented in the Oligocene amber of the Baltic as well as in 

 the Miocene rocks of Colorado and Germany. We realise, 

 therefore, that the geological history of the Hymenoptera, 

 so far as we know it, supports the conclusions drawn from 

 the comparative study of living members of the order. It 

 is of much interest to notice that the three groups of insects 

 — Lepidoptera, cyclorrhaphous Diptera, and Hymenoptera 

 Apocrita — which in their manner of life and ways of feeding 

 are so closely dependent on the higher flowering plants, 

 attained their full development at that period in the course 

 of the earth's history which is distinguished by the domi- 

 nance of the section of the Vegetable Kingdom wherewith 

 their lives are so closely bound up. 



The mecopteroid features of the wings of the Proto- 

 hymenoptera suggest that they, Uke the Paramecoptera, had 

 a common origin with the primitive scorpion-flies during 

 or before the Carboniferous Age. Tillyard, however, in a 

 recent discussion (1926) on these extinct orders of insects, 

 expresses the opinion '' that there were three distinct groups 

 of holometabolous insects which evolved a pupal stage 

 independently of one another early in the Permian Period," 

 and that the Hymenopteroid stock had an origin apart from 

 the great assemblage of orders (Mecoptera, Neuroptera, 

 Trichoptera, Lepidoptera, Diptera) that make up the 



