lO 



Memoir Sears Foundation for Marine Research 



1 54) to form one of the Subclasses of the Class Osteichthyes, and this same course is 

 followed here. 



In 1 86 1 Huxley proposed for this Subclass the name Crossopterygidae (subse- 

 quently changed to Crossopterygii) as a Subordo of his Or do Ganoidei (j^ : 23, 25). Other 

 names proposed later are: Amphiboidei by Hubbs in 1919 (J5: 589); Choanichthyes 

 by Romer in 1937 (72: 56), with Crossopterygii retained as an Order to include fossil 

 as well as living Choanichthyes (7J: 589); and Sarcopterygii by Romer in 1955 to 

 replace his earlier name Choanichthyes {j^: 126). 



Figure 2. Skeleton and basal part of pectoral fin rays of: top, Latimeria, after Millot and Anthony, somewhat 

 emended; lower left, Neoceratodus (lungfish), after Giinther, somewhat emended; lower right, Salmo trutta 

 (European trout), after Parker and Haswell, somewhat emended, c pectoral girdle; b basal segment, and d distal 

 segment of radialia; r fin rays. 



None of these names seems wholly appropriate. While Crossopterygii, in a restric- 

 ted sense, has been widely employed in scientific literature,'" its continued use seems 

 likely to lead to confusion because it has long been associated with the polypteroids, 

 which were the basis for Huxley's name but which have been excluded from the Cross- 

 opterygii, Goodrich {2g: 91) having shown that the affinities of the polypteroids lie 

 with the actinopterygian fishes. 



Amphiboidei does not seem a happy choice, for while the fishes concerned may 

 be "related to amphibian ancestry," the living coelacanths are about "as unamphibian 

 as a fish can be," as Romer has aptly expressed it (j^: 125, 126). 



Choanichthyes is a misnomer if the coelacanths are included, for internal nares 

 neither were nor are present among fossil {j^: 126) or living coelacanths, regardless 

 of the situation in their hypothetical ancestors; and it is doubtful whether the internal 



30. For a recent defense of this course, see Trewavas, et al. {8y: 126-127). 



