372 Dr. R. H. Traquair on Pakeozoic Fishes. 



more curved, and having the ridg-es in greater number (fig. 3). 

 Though the slender form is more abundant than the stout one, 

 I am inclined to regard both as belonging to the same 

 species and as representing respectively the anterior and pos- 

 terior dorsal spines of the fish to which they belonged. 



From the Lower Carboniferous of the Edinburgh district. 

 Rare in nodules above the " Broxburn " oil-sliale at Straiton 

 and in shale in connexion with the "Blue^' coal at Niddrie. 

 More abundant in the roof-shale of the " South Parrot '^ coal- 

 seam at the last-mentioned locality. 



Acantliodidae. 

 AcanfJiodopsis mid'odoti, sp. n. (PL IX. fig. 7.) 



This is the articular moiety of an Acantliodian mandible 

 about I inch in length, and sliowing on its upper margin 

 eleven conical, blunt, incurved teeth, apparently to some 

 extent laterally compressed, and with a slight apical enlarge- 

 ment. Bases of teeth not so broad in proportion as in 

 A. Wardi. 



As the small size and somewhat different sha])e of the teeth 

 strikingly distinguish this mandible from that of the well- 

 known Accmthodopsis Wardi of Hancock and Atthey, its 

 claims to specific distinction seem to be clear and undoubted. 



From the Upper Carboniferous ; Woodhouse Coal, Cheadle 

 Coal-field. The specimen, from the Ward collection, is now 

 in the British Museum, but was previously lent to me by 

 Mr. Ward for description. 



Position uncertain, probably GhimEeroid : — 



Harpacanthus major ^ sp. n. (PI. IX. fig. 8.) 



One specimen only. Spine slender, pretty strongly curved, 

 and at the base again slightly recurved ; 2f inches in length, 

 but, being broken off distally, it must originally have been 

 longer ; rounded in transverse section at the base, but be- 

 coming laterally flattened further on, and with the convex 

 margin slightly sharper than the concave one. On the distal 

 third of the concave margin, presumably the anterior, are seen 

 three stout recurved denticles, their apices | inch apart from 

 each other; at the broken distal extremity is seen the cross 

 section of a laterally compressed internal tubular cavity. 



This spine obviously belongs to the genus Harpacanthus^ 

 which I proposed some years ago * for the TristycMxis Jiinhri- 



* Ann. & Mag. Nat. Hist. (5) vol. xviii. 1886, p. 493. 



