146 Annals of the Carnegie Museum. 



the latter name for the purpose of indicating a certain amount of 

 variation from the typical C. gracilis in the direction of greater depth 

 of body and coarser striation of fiank-scales. 



Concerning the extremely difficult subject of cranial osteology, very 

 little can be added to the few facts already known. The bones forming 

 the cranial roof are as a rule firmly coalesced and their sutures con- 

 cealed by the tubercular ornamentation. Apparently the superior 

 border of the orbits is formed by the large-sized frontals, which are 

 bounded behind by the parietals (the latter separated in the median 

 line by a small-sized supra-occipital) and squamosal. The inferior 

 border of the orbit is formed by the expanded posterior portion of the 

 maxilla, which is of relatively large size and decidedly paleoniscid-like 

 in form. This plate bears numerous fine, acutely conical teeth, and 

 there is also present a small dentigerous premaxilla, which is often 

 found detached from the other mouth-parts. 



Just how the facial plates are arranged in the space lying between 

 the orbit and shoulder-region (clavicle) is difficult to determine. 

 Newberry's interpretation of the elements covering this area in a single 

 specimen studied by him is open to serious question. At least one 

 postorbital is present in its normal position behind the eye. and there 

 may possibly be another (or suborbital) below it. Behind these 

 plates is the area commonly occupied by the operculum and suboper- 

 culum, but the pre-operculum was probably much reduced and nearly 

 concealed by adjacent elements. The general configuration of this 

 region is shown in one of the original co-types of this species now pre- 

 served in the Yale Museum, and also in a specimen belonging to the 

 United States National Museum, which has been examined by the 

 writer. 



On the Cranial Structure of Semionotus. 



The cranial osteology of this genus has been studied chiefly by 

 L. Agassiz, E. Schellwien, and Dr. C. F. Eaton. In general, the 

 arrangement of plates is not unlike that in Lcpidotus, except that the 

 circumorbitals are relatively very small, and the suborbitals are not 

 divided up into a number of polygonal plates. One of the best pre- 

 served specimens of American Triassic fishes showing the head-region 

 is that shown in PI. XXXII, Fig. i, which probably belongs to the 

 species S. micropterus Newburg. It is from Durham, Connecticut, 

 and bears the Carnegie Museum Catalog No. 5285. It is of interest 



