FOOD OF COAL MEASURE INSECTS. 11 



" Small amphibians, and numerous arachnidians, came to limit the swarm of 

 this articulate world in a country without birds or mammals." 



Pruvost does not accept the view of a primitive aquatic origin of insects, but 

 affirms his belief in a terrestrial origin, and thinks that even if an aquatic habit 

 be proved for certain of the larvae of the Coal Measure insects, the habit must 

 have been secondary, and derived from an earlier land ancestry (op. cit., p. 268). 



Scudder^ has observed in the case of the Blattidae that the venation of the 

 tegmina very closely resembles the surface-features of the Neuropteris pinnule 

 so strongly in fact as to suggest mimetism. 



Pruvost rightly urges that a mimetism is of little value unless the mimetic 

 insect frequents the plant mimicked. At the same time, it can be urged that the 

 stout compact bodies of the fossil Blattoids and their powerful walking legs were 

 equally admirably fitted for progression among rank and decaying vegetation, and 

 that in these conditions the Blattoids were quite as likely to have been omnivorous, 

 while finding some degree of protection among the Neuropteris pinnules lying on 

 the ground. 



The writer has previously commented on the association of the wings of Blattoids 

 with the leaves of Gordaites (1911, ' Quart. Journ. Geol. Soc.,' vol. Ixvii, pp. 164 

 165), and has made the following comment : " While Carboniferous Blattoids 

 may have been wholly phytophagous, it is interesting to note that the leaves of 

 Cordaites (in the present case) are impressed with shallow pits, which show faint 

 traces of a spiral. I have in many previous instances found that such pits owed 

 their origin to attached shells of Spirorbis pusillus. Whether these leaves were 

 partially submerged in water during life is an open question ; but in all cases the 

 plant-tissues of the pittings are depressed, and are accurate impressions of 

 Spirorbis. If the Carboniferous Blattoids were not wholly vegetable feeders, the 

 occurrence of Spirorbis pusillus upon the Gordaites may supply a reason for their 

 frequent association." 



CLASSIFICATION. 



The classification of fossil insects has presented great difficulties, both to the 

 palaeontologist, and to the systematist of living forms. Palaeozoic insects show to 

 the systematist a series of forms not strictly referable to any modern grouping, 

 but presenting certain generalised details of structure which link two or more 

 now widely separated groups, besides other features not met with in living 

 forms. 



The palaeontologist finds that he has not to deal with early and primitive 

 types, followed by a regular series showing a developmental progression, but with 

 an apparent sudden incursion of large series of highly modified and well-developed 

 insects, co-existent with others of more primitive type. 



