20 



SCIENCE 



[N. S. Vol. XXVI. No. 653 



knowledge. Suppose a man should undertake 

 a review of the present status of the problems 

 of evolution ; would we not expect that he 

 would at least describe the various theories ? 

 Would it not be ridiculous for the reviewer to 

 give simply a list of books and papers, and 

 indicate in what libraries these are to be 

 found? Science can make no advance when 

 nothing but names are given, or only un- 

 recognizable descriptions, with the indication 

 where specimens are to be found. The reten- 

 tion of a type specimen is always desirable for 

 future reference, but the publication of a good 

 description is the sine qua non of scientific 

 advance. 



Those who take the pains to furnish ade- 

 quate descriptions, and who draw comparisons 

 with the previously known species, will always 

 be regarded as the original describers whether 

 the rules of nomenclature give them credit or 

 not. 



The suggestion of demanding a recognizable 

 description is, of course, open to the objection 

 that it is diificult to decide what constitutes 

 a recognizable diagnosis. The decision must 

 be made separately for each particular case. 

 All would concur in the fairness of the prin- 

 ciple to consider the adequateness of the de- 

 scription at the time when it was made. A 

 description would be recognizable if at the 

 time it was published it served to demarcate 

 the species from all the other then known 

 species of the genus, or the genus from all 

 the other then known genera of the family. 

 Questions of this kind can generally be de- 

 cided by any monographic reviser of a group. 



As the case stands at present there is all 

 incentive for hasty and insufficient diagnoses, 

 at least dur codes do not prevent them. We 

 have either to hunt up the type specimen, 

 to spend friiitless hours to try to read 

 some meaning into a description, or to try 

 by the unsafe method of elimination to deter- 

 mine what species an author intended. As 

 the A. O. U. Code of 1886 put it, " Zoological 

 nomenclature is a means, not an end, of 

 zoological science " ; it ought to make the 

 path clear and not maintain obstacles. Too 

 much of a premium is placed upon age, there 

 is too strong a tendency to resurrect the 



oldest name and try to fit to some particular 

 species, whether it was accompanied by a good 

 diagnosis or not. So long as this continues 

 so long will the names in use be unstable. 



Personally, I never have and never will re- 

 gard a name as tenable unless it is based on 

 a recognizable description or drawing, and 

 that whether type specimens are preserved or 

 not. There are many who share this view, 

 and in this connection attention may be 

 drawn to the very cogent recent arguments of 

 Looss.' We are all in sympathy with the 

 endeavors of the International Committee, 

 most of us realize the difficulty of the ques- 

 tions it has to decide, and we are ready to 

 relinquish personal views in order to reach 

 uniformity. But there will never be uni- 

 formity of opinion in regard to the matter 

 of allowing a name to be based simply upon 

 an " indication." New names are multiply- 

 ing in a geometrical ratio, some of them 

 newly coined and others raked out of the ash 

 heap of describers who deserve oblivion; few 

 genera have been thoroughly revised ; if m.ere 

 " indications " and inadequate descriptions 

 continue to be permitted the task of revision 

 will before long be hopeless. Then there will 

 be need for far more radical reform than the 

 one here suggested. Now while our rules are 

 still plastic let us insist on the absolute neces- 

 sity of adequate diagnoses of genera and 

 species. Thus insecurity may be abolished, 

 each describer be given his just due and no 

 more, and science as well as nomenclature be 

 benefited. 



One recommendation the committee might 

 embody in the code to clarify future system- 

 atic work, though, of course, it could not be 

 applied to the work of the past. That is, that 

 when structural characteristics enter into a 

 diagnosis they should be represented so far 

 as possible by drawings rather than by words. 

 A drawing is immediately clear, few descrip- 

 tions are. Above all it is often very difficult 

 to build a conception of a structure from a 

 brief Latin description since the Latin is too 

 poor in adjectives for our present needs. De- 

 scribers are too intent upon their own conveni- 

 ence, give too little attention to the con- 



^ Zoologischer Anzeiger, 1907, Nr. 19. 



