BY R. J. TILLYARD. 



169 



branch derived by extreme reduction from the very base of the 

 Lepidopterous stem, and having no near relatives within the 

 Order as now existing), the author selects the Hepialid type of 

 wing as the most archaic. Text-fig. 2 shows the interpretation 

 of the forewing of Hepialus eximius Scott. 



Text-%.3.t 



In this wing, the capture of M4 by the upwardly arching Cuia 

 (which is not a mere cross-vein, as hitherto supposed, but a true 

 branch of the main vein Cuj) is beautifully shown. The posses- 

 sion of this character is the essential basis for the productio7t. of 

 the closed Lepidopterous cell. After the capture of 1^4 by Cuja, 



* \Ving-veiiation of Hepialus tximins Scott. Note the close coiiespond- 

 ence in venations in fore- and hindwings (homoneurism). 



t Four phylogenetic stages in the formation of the tri'jamma (T) — a, 

 primitive stage, Avitli no fusion between Cuia and M^ ; J>, fusion between 

 Cuia and M4 completed (N.B., the wing in Text-fig. 1 lies phylogenetically 

 between a and />); c, begintiing of alignment of the trigamma-stem; d, com- 

 pletion of the trigamma (T), with its strong stem and triple fork. (In d, 

 the media, M, and its first dichotomy are omitted, as in most Lepidoptera). 



