DETERMINATION OF GENERIC TYPES, ETC. 47 



cases in which we must have recourse to the still less satisfactory 

 method of "type by elimination." 



It may be advanced against the principle of "type by inclusion" 

 that the included type may be a little known or even an invalid species. 

 If, however, the species was invalid from the standpoint of the author 

 who included it in a later genus, or if he doubtfully referred it to his 

 new genus, it would of course be excluded as type; if on the contrary 

 it was simply a slightly known form, and he still unreservedly included it 

 among his valid species, without showing that he did not consider this 

 species as type, the case still represents a renaming of ah earlier genus. 



It seems quite clear that a "type by original designation" (see p. 

 30) should take precedence over a " type by inclusion," since the 

 former is intentional and results from a desire to conform to the rules 

 of nomenclature, while the latter is either an accident or due to ignor- 

 ing the rules of nomenclature. Accordingly we might have the fol- 

 lowing case: 



X-us, 1890, containing . 



albus, 1890, type by original designation, and 

 niger, 1885, type of Y-us, 1885. 



In this case an author who would go so far as to explicitly designate 

 albus as type would probably have adopted Y-us if available had he 

 known of its existence and that niger was its type. It seems but just, 

 therefore, to bind X-us to albus, to stand or fall according to the later 

 history -of attnis, not only from the point of view that the author of 

 X-us has complied with the requirements of the case, but because of 

 the fact that by such designation the author of X-us has explicitly 

 stated that he considered albus the standard of reference of X-us. 

 Thus a case of this kind would come under the principle enunciated 

 under type by original designation, page 30. 



9. GENERA CONTAINING TYPES OF SEVERAL EARLIER GENERA. 



RULE. If a genus without a designated type contains types of two or more 

 earlier genera, the type of the new genus is to be selected from the contained 

 types (the case being the same as a genus with two or more species, according 

 to the number of types in question), unless it can be shown that snch pro- 

 cedure is directly contraindicated by the origintil author's intentions. 



Under this heading may be cited one of the most unnecessary 

 renamings of genera that is known in helminthology. 



Prosthecosacter Diesing, 1851a, contained four species, three of which 

 were known to Diesing to contain the types of three monotypical 

 genera: 



inflexus contained as synonym filum, type of the monotypical genus Pzeudalius, 



1845; cited by Diesing. 

 minor contained as synonym wflexus, type of the monotypical genus Stenurus, 



1845; cited by Diesing. 

 convolutus. 

 alatw, type of the monotypical genus Pfiarurus, 1848; cited by Diesing. 



