NOMENCLATURE OF AMERICAN BIRDS. 1 39 



not stated to what degree, if at all, we may be 

 allowed to select the type of such a genus by 

 (metaphorically) questioning its author as to which 

 species he himself would have regarded as typical. 

 This method of selecting a type for the compre- 

 hensive genera of Linnaeus^ and others has been 

 practised by certain writers ; and where the indica- 

 tions are unquestionable, this seems a very just 

 method of procedure. Nor is it clear from the 

 " Code," whether the results of the application of 

 Canon XXL (by which the earliest restriction of 

 a comprehensive genus is held to be valid) could 

 be set aside either by the application of the pro- 

 cess of elimination (as set forth in Canon XXIII. ), 

 or on account of the supposed views of the author 

 of the genus. There are thus three different ways 

 for selecting the type of a comprehensive genus, 

 all three of them sanctioned by the " Code." 

 These three methods will often lead to different 

 results, and the ** Code " nowhere states which is 

 in case of conflict entitled to precedence. My 

 own idea is that if the author has indicated in any 

 unequivocal way which species was in his mind 

 typical, that species should be regarded as the 

 type. If no such selection is unquestionable, 

 the species (originally contained in the genus) 

 which is selected as type by the earliest author 



1 Professor A. E. Verrill has called my attention to the fact 

 that Linnseus himself has distinctly stated that he regards the type 

 of each of his genera as being " the best known Em-opean or offi- 

 cinal species." I have not been able to verify this statement ; but 

 if it is correct, it ought not to be overlooked in the process of 

 selecting types for the Linnasan genera. 



