98 Mr. E. Meyrick on the classification of 



number of individual specimens have been examined in 

 those cases where variability seemed to exist, and where 

 the specimens were obtainable for the purpose ; and the 

 detailed results of this examination are added hereafter 

 in the form of an appendix. Typically 10 and 11 are 

 free and separate, but there exists in this group a strong 

 tendency to connection or anastomosis of these veins 

 with one another or with 9 and 12. Vein 6 is almost 

 always separate from 9. In the hind wings, in propor- 

 tion as 5 becomes obsolete, 4 and 6 tend to approach 

 more nearly together, so that no unduly broad inter- 

 sj^ace is left. Hence 6 and 7 are drawn apart, and are 

 almost always separate. The posterior tibiae are very 

 frequently dilated in the c? , and then usually contain a 

 large hair-pencil concealed in a longitudinal groove, but 

 the character is of little generic value. Occasionally 

 the median spurs are absent ; in these cases the allied 

 forms commonly have the spurs very small, and are in 

 general so closely related in all other characters that it 

 is evidently unadvisable to separate them generically. 

 The tongue is sometimes rudimentary, and in some in- 

 stances (as Gonodontis) it has been clear that the 

 character does not involve generic separation ; in others 

 it has proved possible to make use of it. 



Owing to the variability and uncertainty of these 

 characters, the family is one of the most difficult of all 

 the Lepidoptera to classify. It is only after repeated 

 examination and reclassification (most of my material 

 has been recast six times) that I have been able to deter- 

 mine the most essential and reliable points of structure, 

 and the mutual affinities of the genera ; and it is quite 

 probable that in some respects my views may yet be 

 materially altered by the discovery of new forms. There 

 is a great variety of superficial appearance amongst the 

 species, and often considerable difference even between 

 rather nearly allied forms. The Prosopolopha group 

 appears to be the most ancestral ; in its typical form 

 (stout-bodied species, with triangular anterior thoracic 

 crest) this is little represented now in Europe, and is 

 principally characteristic of Australia. From this group 

 originate three others : (1), the group of JFi/hcrnia and 

 Crocota, giving rise to PseiulopantJicra, Abraxas, &c. ; 

 (2), the Ennomos group, developing into Metrocampa, 

 J>('iliiiia, i'(,'C. ; and (3), tbe Sclidoscma group, of which 



