III. — A REVIEW OF THE SCIMIM1 OF AMERICA AND EUROPE. 



By David Starr Jordan and Carl H. Eigenmann. 



Iu the present paper we .have attempted to give the synonymy 

 of the species of Scicenidce found in the waters of North and South 

 America and of Europe, together with analytical keys by which the 

 genera and species may be distinguished. The paper is based on the 

 collections in the museum of the University of Indiana, on a large series 

 belonging to the National Museum, the most valuable part of this 

 series being the collections made by Professor Gilbert at Mazatlan aud 

 Panama, and on the collections in the Museum of Comparative Zoology 

 at Cambridge, Mass. This collection is especially rich in South Amer- 

 ican forms, and nearly all of our information regarding the South 

 American species has been drawn from it. All the representatives of 

 this family in the museum at Cambridge have been examined by the 

 senior author of this paper, and for all statements regarding the South 

 American species he is responsible. 



We wish to express our special obligations to Prof. Alexander Agas- 

 siz, Director of the Museum of Comparative Zoology, and to Mr. Samuel 

 Garinan, curator of the fishes, for the free use of the material iu the 

 museum, and to Dr. Tarleton H. Bean for a loan of special desiderata 

 from the United States National Museum. Through the aid of these 

 two great museums we have been enabled to examine nearly all the 

 species included in the present paper. The only species not seen by 

 us are the following: Cestreus obliquatus, Larimus staJdi, Scicma gilli, 

 Scicena heterolepis, Pachyurus frcmcisci, Pachyurus schomburgki, Pachy- 

 pops trijilis, Umbrina reedi, Lonchurus lanceolatuSj and Eques pulcher, 

 ten of the 113 species recognized. 



There is room for much difference of opinion as to the proper sub- 

 division of the Scicenidce into genera. There are few families in which 

 the various types are more definitely joined together by intermediate 

 forms than in the present one. The subdivisions must be more or 

 less arbitrary, or else the great bulk of all the species must be thrown 

 into two genera, Scicena and Otolithus. Such an arrangement, however, 

 tends to obscure the inter-relations of the species, and so we have 

 adopted as distinct genera all the subordinate groups which we are 

 able to restrict and define by structural characters of some importance. 

 [11 343 



