FOSSIL FISHES. 625 



. In the second place the nature of the overlapping joint between these 

 bones and the anterior ventrals is one which was not adapted to furnish 

 the freedom of motion which the jugulars under the jaws of the preda- 

 tory Dinichthys would need. The same, in a less degree, may be said of 

 the overlapping joint between the tw T o "jugulars," the nature of which is 

 shown in figure 9. 



To this must probably be added the overlapping joint between the 

 "ventro median" or "sternal" plate which bound all four of the bones 

 together into a piece of armor not especially flexible. 



These points might not be against the idea that this portion of the 

 armor was drawn back to a pectoral or intermediate position between the 

 jugular and the ventral, by which the arched outline of the jugulars 

 would still be anterior in position. 



And thirdly, if we consider all these plates as forming the ventral 

 aimor, retaining the anterior and median ventrals as Dr. Newberry placed 

 them, and making the "jugulars" the posterior ventrals, we shall have a 

 most complete arid convincing homology with the plastron of Coccosteus, 

 removing discrepancies which were previously supposed to exist, and 

 adding some harmonies which were not before suspected. 



All the points of harmony which Dr. Newberry so admirably makes 

 out* concerning the anterior laterals and median plate remain in their 

 full force. If now the "jugulars" be placed as the posterior plates there 

 will be added: First, the overlapping of the posterior plates by the ex- 

 tremities of the anterior plates; second, the overlapping of the right pos- 

 terior plate b}' the left posterior plate along the median line; third the 

 sinuous line of overlapping between the posterior plates in both genera; 

 fourth, the less breadth of the posterior plates behind than in front. 



This brings the whole plastron of the two genera into striking ac- 

 cord and it is difficult to resist the belief that whatever position upon 

 the body was occupied by the "plastron" of Coccosteus must likewise 

 have been occupied by the ventral armor of Dinichthys. If it was really 

 ventral in the one, it doubtless was in the other. 



A real point of difficulty in this view may seem to exist in the ab- 

 sence of an}- projecting angles upon the arched margin of the posterior 

 plates of Dinichthys, with which other plates could have articulated. I 

 do not detect any evidence that other plates, either by overlapping or 

 articulation, were connected with the outer margins of the posterior ven- 

 trals of Dinichthys. Such a bold, curved outline might seem to be more 

 easily explainable in an anterior position than in a posterior. But the 

 balance of evidence is so strong upon the other side that it must be ac- 

 cepted as best expressing our knowledge upon the subject. 



To recapitulate, then, this discussion shows that the four bones, 

 which have been supposed to constitute the jugular armor of Dinichthys, 



* Paleontology, Ohio, Vol II, p. II. 

 40 G. O. 



