1888.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 579 



reconnaissaut des relations ftmdameutales avec les Triples. Cette derniere situation 

 accepted par Cuvier est uae des preuvea les plus iut^ressantes de l'importance qu'il a 

 dfi attacher aux parties fondamentales du sqnelette. N'ayant pas eu a ma disposi- 

 tion <les squelettes d'Oreosouie, je les conserverai ici apres les Epiuoches, sur le dire 

 de Cuvier. (Pp. 532, 533.) 



SAUVAGE. 



In 1873, Dr. Sauvage published a special memoir upon the mail cheeked 

 fishes* and distributed them among three families — the Scorpwnidce, 

 Platycephalidce, and Triglidw. He availed himself of some anatomical 

 characters, but not in all cases happily. The family of Triglidce, for 

 example, was characterized by suborbitals covering the whole cheek, 

 hut, as Cuvier long ago showed, the suborbitals do not cover the binder 

 portions of the cheeks in the Dactylopterids. The nasals were said to 

 be soldered in a large plate covering the greater part of the muzzle, 

 but this statement does not appear to be strictly applicable to any 

 of the several types which are combined under the family called 

 by Sauvage Triglidw. Another of the characters given to Sauvage's 

 family, the development of four to six ganglionic tubercles at the origin 

 of the spinal marrow, is applicable to the typical Triglids and proba- 

 bly to the Peristediids, but not to the Dactylopterids, and there is no 

 reason to suppose that it belongs to the A^onids. Further, a subdi- 

 vision of the cataphract Triglids is made into two groups, distinguished 

 by the development of an interparietal bone (as in the Dactylopterids) 

 or the destitution of it (as in the Agouids and Peristediids). In fact, 

 there is no such difference, and the antithesis is probably due to the mal- 

 identification of the bones in the Dactylopterids, where a superficial 

 dermal bone was considered interparietal. The true interparietal, or 

 'supra-occipitiue, is entirely concealed from the roof of the cranium 

 in the Dactylopterids by a special system of dermal bones, while 

 on the other hand, in the Agonids, it is more than usually well devel- 

 oped (for the mail-cheeked fishes), aud extends forwards between the 

 parietals, in part uncovered, and meets the frontals. In the Triglids 

 and the Peristediids it is well developed, but visible only from behind, 

 its anterior or horizontal portion being covered by the overlapping 

 parietals. 



Credit is to be given to Dr. Sauvage for the characters derived from 

 the development of the pelvic bones, for using the number of branchia>, 

 and for utilizing the presence or absence of pseudo-branchire as family 

 characters. He has neglected, by name at least, the HexngrammhUv, 

 and the genus Agriopus has been presented in the Scorptenidce as 



* Sauvage (H. Emile). De la classification des poissous qui coniposeut la famillfl des 

 Triglides (Joues cuirass6es de Cuvier et Valeucieimes). <Comptes Rendus Acad. 

 Sc. (Paris), t. 77, pp. 723-726, 1873; also, ' Description de poissons nouveaux ou im- 

 perfaitement connus de la collection du museum d'histoire naturelle. Famille defl 

 Morpemdees, des Platycephalide'ea et des Triglide"es. <Nouv. Archives Mus. Hist. 

 Na-., Paris (2), t. 1, pp. 109-158, pi. 1, 2, 1878. 



