BUFFALO SOCIETY OF NATURAL SCIENCES 



67 



them are poorly preserved, and the actual bone is missing in places; 

 still their outlines can in most cases be clearly made out. We have 

 restudied the specimen, and made a careful tracing of it which is re- 

 produced in figure 21. This represents all that can be made out; we 

 have only omitted, for the sake of clearness, some hazy indications of 

 plates, especially near the right margin of the dorsomedian. 



The plates present in the specimen are : the cranial shield, dorso- 

 median, both antero- and both postero-dorsolaterals, both mandibles, 



Fig. 22. Stenognathus doUchocephalus (Eastman). Restoration of Dorsal 

 Armor of Specimen Shown in Figure 21 



one antero-superognathal (not mentioned in Eastman's description), 

 a suborbital, a median ventral, and portions of several other plates, 

 perhaps ventrals. 



To determine the genus and species to which the specimen belongs, 

 we must obviously look to the mandibles as the most distinctive 

 elements of the animal. Fortunately thfese are fairly well-preserved. 

 They are much elongated, and anteriorly rise into a beak. Eastman, 

 in his original description of the specimen, decided that the mandibles 

 were of the Dinichthys type, and referred the specimen to that genus. 



