1030 An Interesting Connectiug Genus of Chordata. {December, 
centic, the flat or concave border inwards; the long diameter an- 
teroposterior. Nares close together, each with a subquadrate 
outline, and separated from the orbit by a space a little greater 
than its own diameter. Their borders are slightly elevated, espe- 
cailly above a shallow groove that connects the orbits behind 
them. Nuchal plate twice as wide as long, its posterior outline a 
little convex, its posterior external angle reached by the acute 
apex of the cephalic spine. The median anterior border of the 
cephalic buckler is damaged so as not to give its exact outline, 
but the muzzle was probably broadly rounded. The lateral bor- 
ders are nearly straight, and they diverge to near the base of the 
spine. The external border of the latter is gently convex, and 
turns inwards posteriorly. The surface is marked by longi- 
tudinal lines of flat tubercles, or raised areas, which are separated 
by narrow grooves, and have various lengths. Those on the 
head are usually as wide as long, while those on the nape are 
generally much longer than wide. Those near the borders are 
always broken up, and those at the side and in front of the orbits 
are irregularly distributed. Cephalic border and spine smooth. 
The size varies. The type specimen has the head as large as a 
fully-grown Amiurus catus, but parts of others indicate individ- 
uals approaching double that size. 
Returning to the presentation of the systematic relations of 
this form, it may be observed that in spite of its resemblances 
to the Pterichthyide and the Cephalaspidide, it must be dis- 
tinctly separated from both families. Supposing it to possess 
a ventral plastron like that of the former and Coccosteus, v; 
which is probable, we must not attach too much importance | 
associated with the latter. er 
ers of its orbits and nares, a” 
