On the Fossil Fishes of the Carboniferous Limestone Series of Great Britain. 527 
remarked that “the extremely puzzling and protean character of the surface pro- 
portions, and frequent distortions of this ray are certainly due to its most 
extraordinary character, which has entirely escaped M. Agassiz, viz. :—the 
extreme thinness of the substance and the great size of the internal cavity. 
It is owing to this thin hollow construction that some specimens are flattened 
laterally, showing oblique ridges, and some flattened from before backwards, 
showing the ridges meeting at a salient angle. This explanation also dispels all 
the difficulties under which M. Agassiz laboured, as to the direction of the ridges, 
and the position of the faces of the spine.” This statement has been repeated by 
M. de Koninck in the “Fauna du Cale. Carbonifére de la Belgique,” (1878.) A 
reference to Prof. Agassiz’s descriptions of the species, however, shows quite clearly 
that he was well aware of the facts stated by Prof. M‘Coy, whilst the deductions 
of the latter, as will be seen by further reference to P]. LXIV., figs. 1 and la, are 
evidently fallacious. 
Two or three fragments of osseous dermal plates are described 7 Prof. M‘Coy 
(‘¢ Annals and Magazine of Natural History,” 2nd Ser., Vol. IL, p. 9), as repre- 
sentatives of the Devonian genera Coccosteus and Aetrolenis the latter consists 
of an irregular fragment an inch and a quarter long, and less than half an inch wide, 
but “it is impossible to suggest what part of the body it belonged to.” Two small 
specimens are very doubtfully referred to Coecosteus. The several fragments are re- 
presented on Plate LXIL, figs. 14, 15; they are from the Mountain Limestone of 
Armagh, and form part of the Jones Collection at the Geological Society of London. 
Platycanthus isosceles, M‘Coy, (Pl. LXIL., fig. 9), is described by M‘Coy (op. cit., 
p. 120), as a triangular ray, very wide, the length of the base nearly equalling 
the height of the spine, arched backwards, much compressed, sides flat, surface 
pustulated, and having two rows of sharp conical teeth on the posterior face. 
It very much resembles Oracanthus, but differs in its small size, arched form, 
and posterior rows of teeth, which also distinguishes it from Byssacanthus of the 
Old Red Sandstone. This species‘is also from the Limestone of Armagh. 
Since the preceding descriptions were written an extensive series of specimens 
have been added to the comparatively small number known to Prof. Agassiz ; and 
the imperfect specimens which served for illustration in the “ Poissons F onsite 
can now be supplemented by others moré completely illustrating the characters 
and peculiarities of the fishes to which they formerly pertained. Notwithstanding 
this increased mass of material there is so much diversity of form, and the structure 
of the objects hitherto regarded as dorsal spines is so widely aberrant from the 
normal plagiostomous types, such as Ctenacanthus, that, as will be explained here- 
after, it is very doubtful whether they can be regarded as spines, and not rather as 
the dermal covering of some portion of the body of the fish. It may be advisable 
before attempting a detailed description of the specimens to briefly glance at their 
general characteristics and more salient points of difference and if possible divide 
them into two or more groups. 
fRANS ROY. DUB, SOC., N.S.. VOL, I. ; ex Up 
