represent points of insertion of scales; all veins and wino--maro-in 

 narrowly but distinctly edged with brownish; apparently a strong- 

 .iiK'ous margin running all round eosta and terinen, on latter deli- 

 •i.r.'ly transversely ribbed. The absence of the transverse vein between 

 o and 4, and the peculiar auxiliary cell formed between 5 and 6, are 

 the characters relied on by the author for the justification of a 

 new Family. 



In order to appreciate the position I looted up the record of 

 Palaeontina oolitica Butl., from the Jurassic, hitherto the oldest 

 known Lepidopteron, described as a butterfly by its author, apparently 

 iwith the approval of Westwood and other entomologists (the opinion 

 of Scudder that it was an insect allied to Cicada can only be regarded 

 as ludicrous) ; I must express my opinion that there is little doubt it 

 belongs to the Hepialidae ; the specimen is a complete fore-wing, in 

 which the irregular and ill-composed structure of the upper part of 

 the cell is very different from the compact arrangement of the same 

 part in a buttei-fly, 9 and 10 are stalked as characteristic of the family, 

 and the jugum appears to be perceptibly indicated in the proper posi- 

 tion. Other species described from Tertiary strata are genuine butter- 

 flies. Now it is clearly proved on structural grounds that the Micro- 

 pterymna (the small group to which the Hepialidae belong, otherwise 

 termed Jiujatae) were the primitive form of the Lepidoptera, and the 

 nature of Palaeontina is in accord with this conclusion and confirm- 

 atory of it. Parenthetically, I take occasion to notice that Goss, 

 Tillyard and others, are troubled as to what the early Lepidoptera can 

 have fed upon, when flowering plants were not yet in existence, and 

 honey was not available ; it is a problem easy of solution, since the 

 more primitive forms (including the HepiaUdae) have in general no 

 proboscis or feeding apparatus, and therefore fed on notliing in the 

 perfect state. 



With the evidence recounted above, Dunstania, if regarded as 

 Lepidopterous, is violently discordant. As the hindwing possesses neither 

 frenulum nor prominent basal angle of costa, it must be assumed that 

 the forewing was furnished wdth a jugum ; all existing Lepidoptera 

 exhibit one or other of these structures. But all jugate Lepidoptera 

 have the hindwing similar in neuration to the forewing, with at least 

 11 veins, whereas this wing seems to be of the modern 8-veined type, 

 though conspicuously different from any known form, and in fact 

 quite as highly speciahsed as any now existing. Finally, there is the 

 apparently corneous margin round the costa and termen, which is 



