
Genus Ethmocardium White.
1880. Proc. U.S. Nat. Mus. 2: 292.
Genotype (o.d.): Cardium whitei Dall. Upper Cret., Colorado and Montana. Figured (as speciosum M. & H.) by Meek, U.S.G.S. Territ., vol. ix, Pl. 37, Figs. 4, a, b, c.
Ethmocardium woodsi n.sp. (Plate 36, Fig. 21.)
1917. Cardium sp. 2. Woods, N.Z. Geol. Surv. Pal. Bull. 4: 33, Pl. 18, Figs. 4, 5.
The tubercles that Woods drew attention to and figured are a diagnostic feature. They are cylindrical, flat-topped projections of the matrix on the surface of the internal cast and represent small pits arranged in radial rows on the interior of the shell within the pallial line. They are not the internal reflections of tubercles, on the main radials at all events, because the rows are opposite the rib interstices of the exterior. On the holotype the pits are absent from much of the anterior part of the disc, but on the paratype (Woods, Fig. 4) there are three additional rows anteriorly.
Dall (1900, pp. 1071, 1072, footnote) has discussed a similar condition that characterises the genus Ethmocardium, based on the West American, Upper Cretaceous Cardium whitei Dall and represented in the Turonian of north-west France by C. alternata d'Orb. A further species, E. welleri Stephenson (1941, p. 195), has been described from the Navarro (Maestrichtian) of Texas.
Woods did not consider the New Zealand material worth naming, but the restricted generic identification that is here made alters the position, so the specific name woodsi is proposed. There are only about 25 radial ribs, which are consequently much broader than those of either of the American species.
The hinge of Ethmocardium does not seem to have been observed.
Height, 12 mm.; length, 12 mm.; inflation (1 valve), 4 mm.
Locality: G.S. 589. Selwyn Rapids. Piripauan, Upper Cretaceous.
Cardium sp. 1, Woods.
1917. N.Z.G.S. Pal. Bull. 4: 33, Pl. 18, Figs. 3a, b.
Locality: G.S. 13. Calcareous conglomerate, Amuri Bluff, Piripauan.
Only the single cast figured by Woods is known. Although the umbo appears to be opisthogyrous, the hinge, as far as can be deter-

mined, is on a Cardiid pattern. It well may be that the internal cast does not show the true twist of the umbones. Retention under Cardium sensu lato seems to be the best course for the present.
