OF THE DIUllNAL LEPIDOPTEBA. 495 



and others, that wheu two genera are combined, a new name 

 should be employed instead of either of the old names. Mr. 

 Dunning subsequently argued, with some show of reason, that 

 in a linear arrangement the middle of a genus ought to be 

 considered the most typical ; but it may be replied that it is a 

 common custom to place the most typical species (i. e. those 

 most divergent iii one particular direction) at the head of their 

 genus. 



So many un characterized and badly defined genera have been 

 adopted in Entomology, that we cannot in fairness overlook any. 

 K. mere catalogue name, if any species belonging to it are in- 

 dicated, and much more a species figured with a new generic 

 name, would be entitled to consideration. Thus Gray has indi- 

 cated a new genus under the name of Mesapia for Pieris pe- 

 loria of Hewitson, which he regards as the type of an aber- 

 rant genus of Papilioninro, To reject a manuscript genus when 

 the type is well known, for no other reason than because it has 

 not been described, appears quite unnecessary. A manuscript 

 genus is on a totally different footing to a manuscript species. 

 The only genera which can be justifiably passed over altogether 

 are those which contain utterly discordant species — and not 

 even these if the name has subsequently been retained for any 

 of them. 



As the second proposition could not be maintained, I adopted 

 the following new set of rules : — 



1. Any sufficient evidence of the type of a genus is binding on 

 subsequent authors. 



2. "Where no such evidence exists, an author is at liberty to 

 restrict the name of a genus or subdivision to any section he 

 pleases. 



3. But the name of a genus may not be applied to any group 

 which does not contain at least one species placed in it by the 

 original describer. 



4. The original name of every homogeneous genus (if not 

 a synonym or preoccupied) has a right to be retained for some 

 part of it— preferably, where the type does not admit of positive 

 proof, either for the section which answers best to the definition 

 of the genus, for the largest section in it, or, finally, for that 

 section to which the name can most conveniently be applied. 



5. "When the application of the old generic name is perfectly 

 indifferent, the first species should be retained as the type. 



