128 NEW YORK STATE MUSEUM 



(SO). It is represented by actual bone, shown in outer, ornamented, 

 aspect. The body of the plate is deeper, that is, more nearly circular 

 in outline, than its homologue inC. decipiens. 



The present specimen throws light on one detail in the structure of 

 the Arthrodire head. It is a moot question whether the so-called 

 pineal canal, which is always seen on the inner aspect of the head of 

 Arthrodira, perforated the bone and opened on the upper surface 

 of the pineal plate. The state of preservation of the fossils has 

 always left it open to doubt whether the pore on the upper surface 

 had not been excavated by the collector while freeing the specimen 

 from matrix. The present head of Coccosteus shows indubitably 

 that the pineal canal did open on the upper surface of the head. 

 P'or in the matrix, which is a cast of the upper surface, there is seen 

 a small papilla (p) which is clearly due to the infiltration of sediment 

 into the original opening on top of the head. The function of the 

 canal is not known. But it may be inferred from its conical shape — 

 seen especially in Dinichthys where, owing to its large size, it 

 can be carefully studied — that it lodged a nerve organ which tapered 

 to a point as it rose through the bone to the surface of the head. In 

 the region of the external opening the surface of the cranium is 

 either smooth, or tuberculated, like any other superficial bone, and 

 not excavated for the reception of any sense organ, such as a 

 pineal eye. 



Mandibles. Portions of both mandibles (Mnd) are preserved. 

 They show only the toothed portions of the elements, with small 

 parts of the flattened posterior regions. There are seven teeth pre- 

 served in each mandible ; they are like those in other Coccosteids — 

 sharp, conical and slightly recurved. 



Dorsal armor. The dorsomedian (DM) is 8^ centimeters 

 long (excluding the process) and of the typical Coccosteid form. It 

 exhibits the inner (under) surface and is largely covered over with 

 fragments of other plates. I am not certain whether a posterior 

 spine is present, since the hinder end of the plate does not show ; but 

 a keel is present, as is clearly indicated by the small part of it (k) 

 preserved near the posterior end of the plate. 



Only one antero-dorsolateral (ADL) is preserved, that of the 

 right side. It retains its natural position with regard to the head and 

 dorsomedian. Its lateral line is noteworthy since it consists of two 

 branches which diverge at an acute angle a short distance back of the 

 articulating process. A similar branched canal occurs on the anterior 



I 



