OSTEOLOGICAL CHARACTERISTICS OF THE FAMILY 

 AMPHIPNOIDiE. 



BY 



Theodore Gill, M. D., Ph. D. 



The genus Amphipnous has beeu generally' associated with the typical 

 Symbranchidcv in the same family. By Dr. Bleeker it was, indeed, long 

 ago isolated as the representative of a peculiar one (Pncumohranchoidei), 

 but even he subsequently reduced it to the rank of a subfamily of his 

 Synhranchoidci. Professor Cope again combined it with Symhranchus in 

 his family Symbranchidce, segregating the two in an order (Holostomi) 

 contrasting with the Monopteridcc constituting his order Ichthyocephali. 

 The order Holostomi was defined in the following terms: 



HOLOSTOMI. 



"Epiclavicle suspended to fourth vertebra, post-temporal wanting. 

 Parietals in contact. Mouth bordered by the j^remaxillaries, which 

 are in contact medially and bounded behind by maxillary. Symplectic 

 present; vertebrae unaltered ; no pectoral fin. Third superior pharyn- 

 geal not smaller than fourth. 



"One family, the Symhranchidcv, with the genera Amphipnous aud 

 Symhranchus.'''' 



The " epiclavicle [is notj suspended to fourth vertebra " in the skele- 

 ton of Amphipnous examined by myself, but is nevertheless free and 

 not " suspended to post temporal and to cranium " as in the Ichthy- 

 ocephali; Amphipnous is therefore doubtless the type of the Holostomi. 

 Symhranchus, however, has been asserted by all except Professor Cope 

 to have the "humeral arch attached to the skull." If this is the case, 

 the true Symbranchidw do not belong to the Copean order Holostomi, 

 but rather to the Ichthyocephali. Possibly Professor Cope based his 

 conclusions on an erroneously named skeleton of Amphipnous. At any 

 rate, there appears no reason to doubt that Symhranchus actually" has 

 the humeral arch connected with the cranium. In such case, Amphip- 

 nous should be accepted as the representative of a peculiar family, 

 Amphipnoida\ Inasmuch, however, as it agrees in so many characters 

 with the Symbranchidce and Blonopteridw, it would apjiear better to 

 retain all in the same order, in spite of the fact that generally the attach- 

 ment or freedom of the scapular arch is of ordinal importance. 



Proceedings United States National Museum, Vol. XIII— No. 825. 



