1896.] NEW YORK ACADEMY OF SCIENCES. 59 



later discovery it seemed likely, therefore, that a renewed study 

 of Dr. Clark's fossil would demonstrate a trace of separateness 

 in the elements of the ensiform plate. The result of such an ex- 

 amination, however, proved to be an unexpected one, for the 

 fossil indicates that the ventro-median plate, although weathered,^ 

 was in all probability a single element. 



Evidently, therefore, if we accept the thesis that the median 

 ventral plates of Dinichtliys must be sepai-ate or fused in all 

 members of this genus, the present difficulty is by no means les- 

 sened. And a more critical examination should be first made as 

 to whether both fossils are unquestionably of Dinichthys. From 

 the region in which both were obtained there have been taken at 

 least half a dozen other genera of little known Coccosteids ; 

 might not the present specimens have belonged among them ? 

 This question occurred recently to the present writer, and it has 

 led him to review the material of Dinichthys in the museum at 

 Columl)ia. There he recalled having seen somewhere in the col- 

 lections an ensiform plate similar to the one he had figured, and 

 he remembered that this had been shown in one of Prof. New- 

 berry's earl}' charts.* This has been found, and is figured as 

 Plate III. in the present paper. 



There can be no doubt, he believes, that this fossil belonged 

 to Dinichthys. It was received from Mr. J. Terrell among 

 his specimens of D. terrelli. It ma}^ be made to articulate 

 with the ventral plates of this species, corresponding to them 

 in size, thickness, texture, and it agrees with them in the mar- 

 gin of overlap of its edges. And to no other known Coccosteid 

 could it well belong. This being the case we must conclude 

 that in this species of Dinichthys the ventro-median plates 

 were fused. 



Novv the specimen of Prof. Clark agrees in this character with 

 the fossil of D. terrelli, and even if the ventroraedian element had 

 been discovered alone it would be referred to the genus Dinich- 

 thys. Furthermore the remaining plates are so similar in shape, 

 thickness and texture to those of D. gouldi that they might rea- 

 sonably be regarded as having belonged to this known species. f 



On the other hand it cannot be doubted that other Coccosteids 

 of the Cleveland sliale possessed separated ventro-median ele- 

 ments. In Glyptaspis and Holonema these have been described by 

 Prof. Newberry (Mon. xvi., U. S. Geol. Survey, Pis. xiii and xvii), 

 and this condition doubtless occurred in all the more generalized 

 Arthrodira, and in those especially which like Trachosteus, 



* Chart VI, accompanying Vol. II, Paleontology of Ohio. The figure shows the 

 plate in visceral aspect, but lacking the hinder end. 



+ According to Mr. Kastmau the evidence of the material recently described by Prof. 

 von Koenen is inconclusive in regard to the ventromediau plate. 



