FOSSIL FISHES. 623 



Second, that the plates called jugulars are not such, but are the pos- 

 terior laterals of the plastron. They are the missing pair which are 

 needed to replace the sub-o bitals. 



Third, consequently the four plates supposed to form the jugular 

 armature are the four principal plates of the plastron in reversed position. 

 The fifth, a median plate, should doubtless remain in the plastron. 



Fourth, this arrangement shows the plastron of Dinichthys to have 

 a much closer homologj^ with that of Coccosteus than has hitherto been 

 supposed to exist. 



The outline figures on Plate VII give various views of the four 

 bones in question. The arched bones (3, 4, 6, 7, 9), are the "jugulars" 

 figured in Dr. Newberry's Monograph; the others are the anterior laterals 

 of the plastrons. 



Taking up the points above enumerated in their order I have' first 

 to show that the so called " post jugulars " are really the anterior ventrals. 

 The most perfect specimen of an anterior ventral of D. Terrclii, which 

 I have seen is that shown herewith in figure 8 ; the inner or dorsal side 

 being there represented. The four accompanying cross-sections show 

 the nature of the margins. This bone was figured by Dr. Newberry, in 

 the large chart accompanying the second volume of the Paleontology of 

 Ohio. His specimen, however, was incomplete, lacking five or six inches 

 of the triangular apex. The dotted line by which he indicated the sup- 

 posed outline lacks the triangular apex, which is fully preserved in my 

 specimen. Placing the latter upon his chart, the two coincide in outline 

 perfectly, with the exception mentioned. The corresponding bone for two 

 smaller species, D. intermedins and D. curtus, is figured on plates XLYHI 

 and L of the Monograph. In the first of these also the apex is wanting, 

 while in the second it is complete and its identity unmistakable. If the 

 apex had been preserved in Dr. Newberry's original specimen of D. 

 Terrelli, he would have been spared the discussion of the question as to 

 why the anterior ventrals of Coccosteus overlapped the posterior pair, 

 while those of Dinichthys did not* and his orieinal restoration of the 

 plastron of Dinichthys might have been quite different from that given. 



II. The plates with arched outlines which are described by Dr. 

 Newberry as "jugulars" in his Monograph are demonstrated by my 

 material to be the companions of his " anterior ventrals. " The three 

 cases of overlapping which I have figured on Plate VII furnished the 

 proof upon this point. The bones numbered 1, 2, 3 and 4, belonged to 

 a single individual of Dinichthys Terrelli. When the posterior apex of 

 the anterior ventrals is applied to the sunken triangular area on the so 

 called "jugulars," it precisely fits. The four bones are arranged in the 

 drawing, in very nearly their natural position. The anterior and pos- 

 terior plate on each side are in their exact natural relation to each other, 

 but the space between the two anteriors may have been a trifle wider, to 



* Pa'eontology of Ohio, Vol. 2. p. 11. 



