18 PROCEEDINGS OF THE NATIONAL MTTSEUM vol. los 



well be the brokcn-off anterior part of another branchio-cornual plate, 

 shows a wide canal in its interior that may, perhaps, be a lateral line 

 canal (lie?, pi. 2, fig. 8). If this interpretation is correct, there may be 

 some reason to believe that m the Eriptychiida the lateral line system 

 had no such superficial position as in Astraspis Orvig, in MS., b), 

 but that, like in several younger Heterostraci, it lay partly, or enth-ely, 

 in the middle, vascular layer of the exoskeleton. 



Apart from the plates dealt with here, the material of Eriptychiida 

 gen. and sp. indet. from the Pycnaspis-hear'mg beds also includes a 

 certain amount of scales of much the same kind as those of Eripty- 

 chius americanus from the Harding Sandstone of Colorado (Walcott, 

 1892, pi. 4, figs. 5, 6, 11; Bryant, 1936, pp. 423-424, pi. 8, fig. 2; 

 pi. 9, fig. 2) and Eriptychius spp. from the Winnipeg formation of 

 the Williston Basin in Montana and the Whitewood formation of 

 South Dakota. These scales, which vary somewhat in size, consist 

 of an anterior overlapped portion without ornamentation and a pos- 

 terior exposed portion exhibiting a system of roughly parallel dentine 

 ridges. As in the scales of Eriptychius from the Harding Sandstone 

 and other formations, the dentine ridges are highly reminiscent of 

 those of the exoskeleton in the Cyathaspida and Pteraspida (see 

 Lindstrom, 1895, fig. 2; Kiser and Heintz, 1935, pi. 37; Gross, 1935, 

 fig. 5c; and others) in that they contain, in their basal part, a longi- 

 tudinal vascular canal from which a series of short vascular canals 

 issue with fau-ly regular interspaces to both sides (can.l, can.d, fig. 

 5,b,c; concerning this canal system see also 0rvig in MS., a). In 

 the material from the Pycnaspis-hearing beds there are also a few 

 scales of large size ornamented with fairly coarse tubercles (pi. 3, 

 figs. 5, 6) that are to some extent suggestive of the median dorsal 

 and median ventral ridge-scales of the trunk in several post-Ordovi- 

 cian Heterostraci, e. g. the Cj'athaspida, Pteraspida, and Psammos- 

 teida (see Kiaer and Heintz, 1935, figs. 41-44, md and mv of fig. 50; 

 White, 1935, figs. 1, 2, 03, 65; Obrutchev, 1945, figs. 2-5; and others). 

 As far as one can tell at present, scales of this particular kind are not 

 met with in the Eriptychius material from the Harding Sandstone of 

 Colorado and the Winnipeg formation of the Williston Basin in 

 Montana. 



Discussion 



There is surely reason to believe that, as they become better known, 

 the Ordovician vertebrates of North America will prove highly use- 

 ful for con-elation pm-poscs, but up to now there has been very little 

 information in the literature as regards their stratigraphical distribu- 

 tion. It may be of some interest in this respect that, according to 



