THE PROBLEMS OF TRANSFORMATION 267 



is satisfactory to notice how several recent students, 1 from 

 comparative studies of wing-nervuration and other imaginal 

 structural features, have agreed in regarding all the meta- 

 morphic insects as derived from a common parent stock. 



All the scanty available evidence from fossils confirms the 

 belief founded on the study of the structure and life-history 

 of recent insects, that the Endopterygota appeared later than 

 the Exopterygota in the history of life on our globe. While 

 examples of the latter are comparatively abundant in Car- 

 boniferous rocks, the known metamorphic insects of the 

 Palaeozoic era are represented for certain only by two genera 

 (Permochorista and Belmontia) both from the Upper Permian 

 coal-measures of Belmont, New South Wales. 2 Of these the 

 former is referable to the Mecoptera (scorpion-flies) and does 

 not differ materially in wing-nervuration from members of 

 that order still living in Australia. The latter appears to have 

 possessed a type of wing-nervuration probably modified from 

 that of the Carboniferous Palaeodictyoptera and from which 

 the nervuration of both the Trichoptera and the Lepidoptera 

 might readily have been derived, while the special order 

 (Paramecoptera) to which it is referred by its discoverer stands 

 close to the root of the Neuroptera. In these ancient insects, 

 therefore, we have an indication of the origin of the meta- 

 morphic insects generally, and from a consideration of the 

 larvae of existing scorpion-flies (Mecoptera) and alder-flies 

 (Megalopteroid Neuroptera) we may reasonably conclude that 

 the larvae of the unknown common ancestors of these two 

 groups were polypod like the caterpillars of the former and 

 well-armoured with long thoracic legs like the larvae of the 

 latter exactly the type of larva which our comparative 

 studies, summarized in previous pages, lead us to regard as 

 primitive. 



From these Permian insects a succession incomplete, 

 indeed, but suggestive leads on to the insectan orders and 

 families of the modern world. Triadosialis, from the Lower 



1 C. Borner : " Zur Systematik der Hexapoden ". Zoolog. Am., XXVII. 

 1904. G. H. Carpenter: " Hexapoda ", inEncycl. Brit, (nthed.), Vol. XIII. 

 1906. 



2 R. J. Tillyard : " Permian and Triassic Insects from New South 

 Wales". Proc. Linn. Soc. N.S. Wales, XLII. 1917. "A Fossil Insect 

 Wing belonging to the new order Paramecoptera ". Id. XLIV 1919. 



