10 



PROCEEDENGS OF THE NATIONAL MUSEUM 



either mean that the lateral line system was rather incompletely 

 developed as compared with that in Astraspis, or, perhaps still more 

 probable, that it was situated to a large extent in canals in the interior 

 of the exoskeletal elements so that it was invisible in external view 

 (cf. in the Eriptychiida, p. 18). 



can pea 



Figure 3. — Pycnaspis splendens, new genus, new species. Vertical section of a tubercle of the 

 ezoskeleton (section No. S 1477, from 0rvig, in MS., a; X150). Explanation of sym- 

 bols: b.as, aspidin; bzul, inner part of dentinous tissue; can. pea, vascular canals issuing 

 from the upper part of the "pulp" cavity; dt", fine ramifications of dentinal tubes; ezul, 

 outer part of dentinous tissue; fi, fibers of Sharpey; lam, concentric lamellation of aspidin; 

 pea, "pulp" cavity. 



The polygonal plates of the carapace, which are most commonly 

 met with in the material, are generally elongated rhomboid or more or 

 less distinctly pentagonal in shape. They are often somewhat convex 

 externally, but as far as I can find they are not raised upwards into a 

 central cone-like prominence as frequently is the case in Astraspis. In 

 some places they overlap each other to some extent, but as a rule 

 they show no imbricating arrangement, being situated beside each 

 other in much the same manner as the exoskeletal units, e. g., in the 

 carapace of Tesseraspis (Wills, 1935), in the cephalic shield of certain 

 late Osteostraci (Prvig, in MS., a), and in the armor of various Ostra- 

 ciontidae. 



The ridge-scales, which arc s^nnmetrical, var}^ to some extent where 

 theh* shape is concerned. Thus, some of them — which, judging by 

 the condition in Astraspis desiderata (0rvig, in MS., b), qidte con- 

 ceivably occupied a position in the anterior half, or thereabouts, of 

 the longitudinal crests of the carapace — are longer than broad and 



