187 



allel to each other ; the lamina is turned outward and upward (see Plate XLV, 

 fig. 7, and cut, fig. 9). In Iclithijodectes anaides, the outer portion of the 

 lamina is extended backward as well as forward; in a species of Portkeus 

 (fig. d.), forward only. The face of attachment of the ventral fin is infero- 

 lateral, and in antero-posterior line with the anterior rod ; it is therefore 

 within the line of the external margin of the lateral alu of the femur. The 

 face is expanded in a vertical direction, and is suhround. Its manner of 

 articulation with the spines of the ventral fin presents a close analogy with 

 the corresponding articulation of the pectoral spines witii llie scapula, so far 

 as regards the first rays. Thus, there are two plane articular facets, — the one 

 superior, the other inferior. Between these is a prominent and narrow 

 transverse tuberosity, which bears an articular facet directed partially back- 

 ward. Behind this, instead of the fossse of the scapula, there is another 

 tuberosity, which is directed posteriorly. Thus, there are four facets, of 

 which the largest pair supports the diverging halves of the bhse of the first 

 ventral spine. The middle tuberosity is probably in contact with a portion of 

 the second spine; and the posterior tuberosity with the base of the tliird. 



I have alluded to the -ventt'al rai/s as spines, and such arc tlie first three. 

 Whether there are additional spines is more than doubtful, as in Porlheus, at 

 least, there are no other articular surfaces than tho-se described. A p;iir of 

 ventral fins, found in relation with their lemoral supports, by Profigssor 

 Merrill, and .sketched on the spot before removal from the rock, embraces 

 only the three spines; the halves of the first being separated, as is oflen the 

 case in isolated specimens. Whether additional soft rays existed is also 

 improbable; but, if existing, they must have been very few, as the basis fitr 

 their support, such as exists in SalmonidcE and otiier physostomous fishes, is 

 absolutely wanting. It is therefore improbable that the internal basilar 

 bone was similar to that in the above recent fishes ; though from the analogy 

 of the pectoral fins, one would expect to find ventral basilars. There are 

 two small bones in the specimen described, which, according to the sketch, 

 occupied the position of basilars ; and such they prol^ably are. Each is an 

 irregular discoidal body, with one, a concave facet on one side. 



The osseous condition of the ventral rays is a remarkable character it) 

 fishes of the general physostomous affinities of the S(iiirodo)ilU}ce. It is a 

 point of resemblance to the physoclystous or spinous fishes, for which the 

 structure of the superior walls of the skull in some measure prepares us. 



A third kind of spinous ray was originally described by me as jiertaining 

 to the Portheus thaumas, and had been previously referred by Professor 

 Agassi/, to tlie genus Ptychodux (vide Poissons Fos.siles). This is coniposcd 



