CORALS FROM BRITISH DEVONIAN BEDS. 501 
Newton Abbot. Pl. XXIII. fig. 1, Lummaton quarry, near Tor- 
quay, both in Middle Devonian limestone. 
ZAPHRENTIS (or AMPLEXUs ?). 
General form cylindro-conical, curved. Height 4+ inches, diame- 
ter of the calice about 1? inch. LEpitheca strong. Principal septa 
64, with a secondary series which are very minute. The former 
are thick: they extend horizontally 4 to 4 inch from the epitheca ; 
vertically they extend down the sides of the calice, which, in part 
of the periphery, is complete. The upper tabula is quite smooth for 
a space measuring an inch in diameter, the septa terminating at its 
margin. The fossula is well defined, but restricted to the septal 
area, and appears scarcely, if at all, to depress the tabula. It is 
many degrees out of the median plane of curvature, but is nearest 
the greatest curve. 
Obs. This coral, from its large smooth tabule, approximates, in 
some respects, to Amplewus; but the most characteristic forms of 
that group (¢. g. A. coralloides, Sby.) are totally devoid of secondary 
septa. This is a more reliable point for their generic diagnosis 
than the greater or less extent of the septa across the tabule *. 
Loe. Orchard quarry, Dartington, in Middle Devonian limestone. 
ZAPHRENTIS SUBGIGANTHA, sp. noy. (? var. of Z. gigantea, Lesueur). 
(Plate XXII. figs. 2, 3, 4, 5.) 
Little or nothing is known as to the length and degree of curvature 
of these large corals. In transverse section they are circular, measur- 
ing about 2 inches in diameter. LEpitheca strong. Septa from 110 
to 124, of two orders. The principal ones can be traced almost to 
the centre as crests on the tabule. The secondaries scarcely attain 
4 millimetres, including the epitheca. They often bend towards the 
primaries, and are united by a few dissepiments (or rather the sub- 
divided margins of the tabule). These latter are large, and virtually 
extend across the entire visceral chamber. ‘The fossula (well seen 
in fig. 4) is formed by a deep inflexion of the tabule, the septa 
_ bending round with the margin of the depression. in that example 
four principal septa can be seen within the circumscribed area, but 
these minor appearances would not be constant in all transverse 
sections. 
Obs. There can be little doubt that these three specimens belong 
to the same species. Fig 3 has an illusory aspect of great stout- 
ness of the septa and other parts of the sclerenchyma. This is 
due to crystallization perpendicular to their surfaces, as the line 
representing the actual plate can be clearly observed in the midst. 
In the local (Devonshire) museums, almost invariably, the Za- 
phrentoids with wide tabulz have been indiscriminately referred to 
Amplexus tortuosus, Phillips; and this fact has been somewhat pre- 
* For opinions touching the relations of these genera, consult Rominger, J. ¢. 
p- 153, F. Romer, /. c. p. 362, J. Thomson, J. ¢., Zittel, &e. As a whole, they 
are regarded as broadly separable. 
