1898-1902. No. 33.] UPPER DEVONIAN FISH REMAINS. 39 



C. R. EASTMAN, to whom I sent photographs of the specimen, informs 

 me that he considers it to be an antero-ventra-lateral plate. In that 

 case the two upper margins of the figure can scarcely he the original 

 ones. They appear, however, to be the original limits of the plate. 



In addition we can see no traces of the sensory canals, which are 

 said to be characteristic of this plate. Thus this interpretation is not 

 satisfactory. 



It seems to me to be more probable that the plate is a portion of 

 a bilateral symmetric plate. A median dorsal plate is also not improbable. 

 The point which is turned up in the figures would in the latter case be 

 the backside and the facet that by which the plate lies over the right 

 anterior dorso-lateral plate. The lack of any median ribs on the inner 

 side argues against this theory, as does also the circumstance that the 

 plate grows thinner on the right side (fig. 3). The latter may be ex- 

 plained by assuming that the plate has been worn down. To judge by 

 the coarse ornamentation, a dorso median plate, however, would pre- 

 sumably be much larger. 



The plate may perhaps be best interpreted as a medio occipital plate. 



In whatever way the specimen is localised, it must be assumed to 

 belong to a large, new Coccostcmorph fish. Several large coarsely 

 ornamented forms have been known previously in certain incomplete 

 fragments. Possibly Aspidicldhys ingens v. KOENEN, (I) shows the 

 greatest resemblance as to ornamentation, although this form cannot be 

 identified with the preceding with any certainty. Aspidiclithys clavatus 

 Newb. has an even coarser sculpture. 



In my material there are also one or two other small fragments 

 which might possibly be placed here. 



Locality. From the abundant fish horizon in Skrap Valley. 



Asterolepidae. 



Bothriolepis cf. hydrophila, AG. 



(PI. VII, figs 16). 



From the dark schists of Galge Valley we have 3 remains of a 

 small Botliriolepis, which are of great interest. 



Description. The most complete specimen (S. I. 101 and 104) 

 shows the proximal portion of the left arm and the somewhat incom- 

 pletely preserved anterior ventro lateral plate. (Fig. 1.) The hinder 

 part of this plate is well preserved and shows distinctly the pointed 



