of Edinburgh, Session 1876 - 77 . 265 
the lowest wrought coal in the Edinburgh district, between the 
first and second Marine Limestone beds in ascending order. 
Gonatodus, gen. nov., Traquair. 
Amblypterus, Agassiz (pars.). 
The body is fusiform, the scales rather large, rhomboidal, orna- 
mented with striae and punctures, sometimes nearly perfectly 
smooth. Rays of pectoral fin articulated ; base of ventrals short ; 
dorsal and anal large, triangular, dorsal situated behind the middle 
of the back so that the middle of its base is opposite the commence- 
ment of the anal ; caudal following closely on the anal, powerful 
and deeply cleft with well developed body-prolongation along the 
upper lobe ; rays in all the fins numerous, and closely jointed. 
Suspensorium not quite so oblique as in Palceoniscus, Elon- 
ichthys , and most other genera of the family, but more so than in 
Amblypterus ; operculum large, oblong, suboperculum wanting, 
interoperculum * quadrate ; branch iostegal rays numerous, with a 
median lozenge-shaped plate behind the symphysis of the mandible. 
Jaws stout ; teeth closely set and of very peculiar form, being 
cylindro-conical, first inclined slightly inwards, then bent outwards 
at an obtuse angle, the apex coming then by another curvature to 
point upwards (or downwards in the case of the upper jaw). 
Gonatodus punctatus , Agass. sp. 
Amblypterus punctatus (pars.) Agassiz, Poissons Fossiles, vol. ii. pt. 2, p. 
109 ; Atlas, vol. ii. tab. 4 c, fig. 4, but not figs. 3, 5, 6, 7, 8. 
(?) anconocechmodus, Walker, Trans. Geol. Soc. Edin., vol. ii. 
pt. 1 (1872), pp. 119-124. 
The length of the entire specimens varies from 5| to inches, 
but in none is the extreme point of the upper lobe of the tail pre- 
served. The length of the head is contained about four times, the 
greatest depth of the body about three times in the total length up 
to the bifurcation of the caudal fin. The head is short, with bluntly 
* The bony plate which I here denominate “ interoperculum ” is the same 
as that which, in the Palgeoniscidse, has hitherto been considered as “sub- 
operculum.” In Rhabdolepis there occurs between it and the operculum 
another and smaller plate, which I interpret as “suboperculum,” and which 
is wanting in most of the genera of this family. (See the author’s account of 
the structure of the Palaeoniscidae in the Mem oirs of the Palaeontographical 
Society for 1877). 
