24 Annals of the Carnegie Museum. 



American forms three types of double-armored herrings, representing two, 

 or perhaps three, distinct genera. These groups may be thus compared: 



DiPLOMYSTUSi (type Diplomystus dentatus Cope). (Plates V and VI.) 

 Vertebrae forty- two in typical species; the caudal vertebrae about 

 twenty-three in number; anal fin long, with twenty-five to forty rays; 

 ventral fins small, inserted before dorsal; ventral region very prominent, 

 compressed; cleft of mouth very oblique, the chin prominent; teeth 

 present, moderate; dorsal scutes pectinate; scales small, about sixty in 

 number. 



Ellipes (type Diplomystus branneri Jordan). 



V'ertebrae about thirty-two, the caudal vertebrae twelve to seventeen; 

 anal fin short, of about eight to twelve rays; ventral fins very small, much 

 smaller than pectorals, inserted below or before dorsal; ventral region 

 prominent, compressed; cleft of mouth oblique, maxillary narrow; no 

 teeth so far as known; dorsal scutes entire; ventral scutes not serrate; 

 scales apparently large, very thin, and deciduous. 



Knightia- (type Knightia eocczna Jordan). (Plate VII.) 

 Vertebrae about thirty-six, the caudal vertebrae twenty-three; anal fin 

 short, of about fourteen rays; ventral fins well-developed, as large as 

 pectorals, inserted opposite front of dorsal; ventral region not at all 

 prominent, body more or less elongate; mouth little oblique; maxillary 

 narrow; no teeth so far as known; dorsal scutes entire; ventral scutes 

 long; scales large, smooth, about thirty-five. 



*In the fine specimen of Diplomystus dentatus, figured in Plate V by the courtesy 

 of Dr. John P. Merriam of the University of California, the opercle is covered with 

 large irregular scales. As no scales occur on the head in any other known herring, 

 these are probably loose .scales out of place in the specimen? 



IEditor's Note. In December, 1898, I received from Mr. Henry L. Ward of 

 Rochester, N. Y., as a Christmas present, a slab from the Green River Shales 

 at Fossil, Wyoming, containing, as he humorously wrote me, "a painting by the 

 oldest of masters." This slab represents a specimen of Diplomystus dentatus 

 Cope, in even finer preservation than the one figured by Dr. Jordan, and rep- 

 resented in Plate V. I have had the slab photographed and take the liberty of 

 annexing a reproduction of the photograph as Plate VI to this articles, and trust 

 that Dr. Jordan will forgive me for the act. — W. J. Holland.] 



^Histiurus Costa, Atti Accad. Pontan., V, 1850, 288 (elatus); not Histiurus 

 Agassiz, an emendation of Istiurus. 



Knightia Jordan, Univ. Cal. Publ., V, No. 7, 136 1907, (type Knightia 

 eocana ]ov<larv=Cliipea humilis Leidy, 1856, not of Meyer, =C/M^ea pusilla Cope, 

 1877, not of Mitchell). 



