22 PALEONTOLOGY OF OHIO. 



As I have remarked on a preceding page, before the homologies of the 

 anterior three plates of the ventral shield of Dinichthys can be accurately 

 determined, we must know something more than we now do of the other 

 elements of the shoulder-girdle in this genus. First of all, the homology 

 and relations of the great bones, which I have called scapulo-coracoids, 

 must be ascertained. These evidently correspond wholly, or in part, to 

 the so-called coracoids of Polypterus (C of Prof. Huxley's diagram, 

 " Classification of Devonian Ganoids," p. 22, Fig. 17), and many other fishes ; 

 and probably to the " Clavicles " in Parker's diagram of the shoulder-gir- 

 dle of Lepidosiren* How these bones terminate below, whether in an 

 epicoracoid cartilage, or in interclavicular plates anterior to the ventro- 

 pectoral shield, we have not yet learned, but must ascertain before we can 

 fully reconstruct the shoulder-girdle. In the absence of proof to the con- 

 trary, we may accept, provisionally, the view of Parker that the antero- 

 lateral plates of the shield of Goccosteus (and hence of Dinichthys) are the 

 homologues of the broad, bony plates which form the basal supports- of 

 the pectoral fins in the sturgeon, and those which meet to complete the 

 pectoral arch in Gasterosteus, the interclavicles of Galamichthys, etc. In 

 all these cases, however, there are no median plates in the pectoral shield, 

 and, to find this element, Parker sends us to the Lophobranchs, where, in 

 the " abdominal line plates," he sees the exact counterparts of the median 

 bones of the shield of Cocoosteus. It seems to me, however, that this ver- 

 sion of the homologies of the 'plates of the plastron of the great extinct 

 Placoderms cannot be strongly insisted on, though it would be difiicult to 

 disprove it. The plastron of Dinichthys, for example, composed of but 

 five, large and ponderous, bony plates, is so simple and symmetrical that 

 the effort to find its honiologues among the multitudinous scutes of the 

 little Teleost pipe-fish, appears somewhat hopeless ; little less difiicult, 

 intieed, of satisfactory accomplishment than to identify the liomologues 

 of these great plates among the rhomboid or circular scales of a scaled 

 Granoid. 



Leaving the anterior series of the ventral plates of Cocoosteus and 

 Dinichthys with the provisional interpretation given above, and passing 

 to the posterior pair, we have still greater difiiculty in following the lead 

 of the great anatomists who have written on the affinities of the Placo- 

 derms. 



It will be remembered that Mr. Parker says they may either belong to 

 the post-clavicular cincture, or may represent the second pair of inter- 

 claviaular bones of Syngnathus ; while Huxley considers them the equiv- 

 alents of the posterior pair of bones of the ventral shield of Loricaria, 



* Monograpli on the Shoulder-Girdle, Plate 11, Fig. 1. 



