1897.] On the Characters of Macropetalichthys. 495 
Two facts furnish additional corroboration of the view that 
the posterior cranial angles are formed by plates homologous 
with the epiotic and marginal of other Coccosteids: first, the 
course of the sensory canal system, which traverses the center 
of the epiotic in a straight line backward until it emerges upon 
and penetrates into an element called by Cope the “nuchal 
plate; ” and second, the reception of the postero-lateral angles 
of the cranium into a concavity on either side of the nuchal 
plate so that the latter becomes firmly articulated with the 
head shield. This mode of union between cranium and 
nuchal plate is, we believe, equivalent to the hinge-joint 
formed by the epiotic and antero-dorso-lateral in other Arthro- 
dires. The logical conclusion of this view is that the “nuchal 
element” represents collectively the dorsal plates of the body, 
and is homologous with the dorso-median and dorso-laterals 
of other Coccosteid genera; hence it is more properly styled the 
dorsal plate. 
The dorsal plate, as it will hereinafter be called, occurs as a 
distinct element immediately behind the cranium, from which 
it is partitioned off by a thin osseous membrane that forms a 
vertical wall in front, and slopes away under the posterior 
angles of the cranium on the sides. A thin layer of bony tis- 
sue covers the entire upper surface and also the posterior face, 
the latter’ corresponding to the cranial process in the Dinich- 
thyids. The existence of this element was known to New- 
_ berry, who regarded it as a “cerebral chamber.” Cope demon- 
strated the fallacy of this notion, and considered it as a pro- 
longation of the median occipital region backward in order to 
protect the anterior part of the vertebral axis; and it was held 
that “such a structure would indicate the presence of a num- 
ber of fixed vertebral elements, such as exists in the chimaeras, 
the rays and the sturgeons.” This author correctly observes 
that there are “ two angular elements on each side of the pos- 
terior region, which are also shown to be distinct. . . -o One 
of these is wanting in the specimen, showing its junction with 
the median element is by a smooth squamosal suture.” These 
angular eleménts he calls the “ lateral nuchal plates;” but it 
8 Loe. cit., p. 453. 
