186 CH. WARDELL STILES, 



Three prominent helrainthological publications, in particular, would 

 be much more convenient for consultation if either of tbese plans bad 

 been adopted; reference is made to Dujardin (1845), Looss (1899), 

 and Braun's Vermes. 



IL In proposing a new genus. 



1) Use bold type, as mentioned above, for beading or subheading. 



2) Having selected a name, consult Scudder's (1882) Nomenciator 

 Zoologicus, and tbe index of the Zoological Record, from 1880 to date), 

 in Order to see whetber or not the name selected is available or pre- 

 occupied. Hundreds of changes of names could be avoided by follow- 

 ing this very simple plan. It is of course to be regretted tbat neither 

 of these publications is absolutely coraplete, but they are of great 

 value, nevertheless. 



Beyond a doubt, many authors will raise the point that neither of 

 these works is accessible to them. Both are, however, to be found in 

 nearly all important scientific libraries, and are surely accessible to 

 some one friend of every author. 



3) Select a species as type of the genus and clearly 

 State so in the article. Compare discussion under 5. 



4) Give a clear , Condensed diagnosis , showing the essential 

 characters. 



5) Give a differential diagnosis, showing the characters by which 

 the new genus dififers from its dosest relatives. An analytical key is 

 best for this purpose. 



Regarding 3 and 4, it may be remarked that opinion dififers among 

 authors as to whether the type or the diagnosis is the more import- 

 ant. Looss evidently looks upon the diagnosis as the more important, 

 and in taking this view he sides with many systematists of undoubted 

 ability. 



While much is to be said in favor of this view, the latter cannot 

 by any means be admitted as seif evident. In this connection it may 

 be noted: a) When a type species has been published for a generic 

 name, this is practically the only definite, unchangeable, and absolutely 

 objective point connected with the whole matter, b) Since the limits 

 of a genus are to no little extent subjective, the diagnosis must neces- 

 sarily be subjective, in the same degree. c) In tryiug to evolve a 

 natural Classification, the characters selected as of generic value are 

 subject to the existing State of general kuowledge regarding the group, 

 to the State of knowledge of the author who writes upon it, to the 



