244 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.52. 



of the shoulder girdle in modern sturgeons. Impressed with the 

 similarity which he observes between this conjectural arrangement 

 and that presented by the body armoring of Arthrodires, he draws 

 the inference that the latter group belong to the same stock as stur- 

 geons. Apparently this view, together with the acceptance of Jaekel's 

 reconstruction of a "shoulder-girdle" in Rhynchodus, influenced 

 Dollo ^ to declare in favor of associating Ptyctodonts with Arthro- 

 dires. 



It need only be said in this connection that the presence of an ossi- 

 fied pectoral arch in any genus of Ptcytodonts is an unconfirmed 

 speculation very difficult to maiutam in the lack of positive evidence; 

 and the elements which have been interpreted as constituting it in a 

 single genus are clearly of very different nature, being externally 

 situated. And in case of any near affinity between Ptyctodonts and 

 Arthrodires, Ave should expect the dental plates of the former to be 

 accompanied by an ossified head-shield and other hard parts similar 

 to those invariably associated with Arthrodires, which is precisely 

 what we do not find in nature. The formerly prevalent view that 

 Ptyctodonts are of Chimaeroid stock is in harmony with all the evi- 

 dence thus far obtained. It may be of mterest to introduce at this 

 point a diagrammatic scheme taken from a paper by C. Tate Regan ^ 

 for the purpose of showing graphically that author's ideas of the re- 

 lationships of the sturgeons and "Placoderms" (i. e., Ai-tlirodires 

 plus Antiarcha and Osteostraci, according to Regan's definition) to 

 to the other orders of Teleostomes: 



Teleostei 



Dipneusti Placodcrmi 



\ / 



Crossoptcry^n 



Chondrostei 



Genus HETERACANTHUS Newberry. 



Syn. Gamphacanthus S. A. Miller. 



This^is a provisional genus, founded upon detached fui-sphies 

 occurring in the Middle Devonian of Wisconsin and adjoining states, 

 the relationships of which are uncertam. Their association in the 

 same beds with large dental plates of Palaeomylus and other Ptycto- 

 donts affords reason for believing them to be of Chunaeroid nature, 

 and for the present they may be tentatively referred to the family 

 now under consideration. 



1 Dollo, L., Les Ptyctodontes sont des Arthrodires. Bull. Soc. Beige G60I., etc., vol. 27, 1907, pp. 1-12. 



2 The Phylogeny of the Teleostomi. Ann. Mag. Nat. Hist., ser. 7, vol. 13, 1904, pp. 329-349. 



