• (g.) Kawau: No specimens very near the type were seen.

• (h.) Oamaru: The many fossiliferous localities in this district and the various horizons represented give Hutton's use of the name a very wide meaning. The Target Gully “ shell-bed ” was unknown to him, and, as he used “ Awamoa ” as a separate locality, it is hard to say from where his specimens came.

Thus it will be seen that there is room for considerable doubt as to the exact limits assignable to this species. The shells from the upper beds at Broken River, though more oblique than the Kanieri one, have the same strong anal sinus. Specimens similar to the Broken River ones are common at White Rock River, Pareora, and these, principally, were used in this paper for comparative purposes (see Plate 7, fig. 2).

Localities:—

• Kanieri. (Lectotype.)

• 165—White Rock River, Pareora. (Also a flat variety in Otago Museum collection.)

• 458—Lower Gorge, Pareora River.

• 237—“ Upper part of Mount Brown limestone, Trelissick.”

• 451a—Pareora beds, Thomas and Porter Rivers.

• Lower Gorge, Waipara (Marshall). (Shells more oblique and with sloping shoulders; perhaps a new species.)

• Muddy Terrace, Waikaia (Marshall).

• Opihi River, South Canterbury (north bank, two miles west of Pleasant Point) (Marwick).

• Grey Marls, Mead Gorge, 25 ft. below conglomerate (Thomson).

• Tutamore Ridge, Poverty Bay district (Ongley and Macpherson).

• Shingly Creek, Lake Heron (Canterbury Museum collection; two specimens, one with a high, sharp beak).

• Records to be eliminated from N.Z. Geol. Surv. Pal. Bull. No. 8:—

• Page 23. Onairo = G. manaiaensts.

• Page 45. Motunau beds, Weka Pass = G. manaiaensis.

• Page 48. Trelissick Basin, loc. 239 = G. trelissickensis.

• Page 49. Trelissick Basin, loc. 449 = G. trelissickensis.

Glycymeris subglobosa Suter. (Plate 3, fig. 3; Plate 4, fig. 3.)

1917. Glycymeris subglobosa Suter, N.Z. Geol. Surv. Pal. Bull. No. 5, p. 66, pl. vii, fig. 8.

For the description of this species Suter used, in addition to the Kakahti material from which he chose his holotype, a paratype from Target Gully. This latter shell belongs to a different species (G. robusta n. sp.), and has been responsible for such features as “ more inflated than globosa” and “ secondary ornamentation on the ribs ” in Suter's description.

The number of ribs on the holotype is 56, but other specimens have as few as 48. This species is distinguishable from G. hutioni by its somewhat rounded ribs and interstices. In this feature (seen only in well-preserved samples) it resembles the Cretaceous species—an interesting fact when its low stratigraphical position in the Tertiary is remembered. The angle of the beak is wider than that of G. huttoni, the anal depression is not so marked, and the outlineas more nearly equilateral.

Localities:—

164—Kakahu. (Holotype.) (Also in Dr. Marshall's collection.)

176—Black Point. (Poor specimens.)

Although this shell has been recorded from many horizons—as high, indeed, as Wanganuian—these determinations must be considered extremely doubtful. No specimens examined from other than the two localities given above can be .placed in this species.

Glycymeris robusta n. sp. (Plate 3, fig. 5; Plate 4, fig. 5.)

Shell very large, heavy, much inflated, subquadrate; beaks very prominent; anterior dorsal margin long, curved, descending gradually; anterior end bluntly angled above; posterior dorsal margin shorter, more rapidly descending; posterior margin with a flattened curve; basal margin meeting the posterior in a blunt angle, but curving regularly to meet the anterior; anal depression moderate; sculpture consisting of about 43 ribs, broad on the anterior and middle part of the disc, but narrow on the posterior; interstices narrow, but widening slightly towards the margin; crenulations of the inner margin obsolete posteriorly; ligamental area long and fairly wide, with four well-spaced ligamental grooves; teeth long and strong, horizontal at the extremities of the hinge-line; anterior muscular impression large, triangular; posterior smaller, subquadrate.

Dimensions of holotype (right valve): Height, 74 mm.; length, 75 mm.; thickness, 30 mm.

Localities:—

• Target Gully (Park, 1916). (Holotype, in collection of N.Z. Geological Survey.) Specimens collected also by Marshall, Benson, and Finlay.)

• 649—Paparoa Rapids. (Collected also by Marshall and Murdoch.)

• 1017—Raekohua Valley, Tangarakau. (Specimen of 108 mm. height.)

• Mead Gorge gravels (1 from conglomerate). Collected by Dr. Thomson. This specimen was (apparently accidentally) in the Geological Survey locality collection from Awatere Valley, loc. 126, and was accompanied by a label of Suter's, “Cucullaea alta.” This record must therefore be deleted from Pal. Bull. No. 8, p. 30.

• ? Orari River. (Canterbury Museum collection, fragmentary specimen.)

• Point Hill, Glenavy. (Otago University collection.)

• ? 166—Lower Gorge, Pareora. (A poor specimen, flatter than usual; listed by Suter in Pal. Bull. No. 8, p. 57, as G. subglobosa.)

• 483—McKay's “ Hutchinson's Quarry greensands,” Wharekuri. (A fragment.)

Remarks.—The holotype was used by Suter as a paratype for his G. subglobos a from Elakahu. The excellence of its preservation makes one wonder why he did not make it the holotype of his species, instead of the matrix-filled specimen that he used. Perhaps he doubted the exact specific identity of the Target Gully and the ELakahu shells.

Although a large shell, the holotype is not quite full-grown, and so does not present all the characteristics of the old individuals, which are as much as 108 mm. in height. These are thicker and heavier, the shoulders are more rounded, while the area is broad and has encroached on the teeth, which are thick and irregular. The flattened posterior margin is extended ventrally, making the shell more inequilateral.

This species is larger and more tumid than G. subglobosa; the ribs, too, are wider. Young shells of G. robusta would be very like mature G. subglobosa, but the old shells present quite a different appearance, having more rounded shoulders and an inequilateral outline.

Very young shells of, say, 20 mm. height are difficult to separate from the young of G. huttoni. The teeth of G. robusta, however, are slightly thinner and more horizontal, the hinge-line is narrower, and the anal sinus somewhat shallower. The full-grown shell is more inflated than G. huttoni, the hinge-line is longer, the anal sinus much weaker, and the teeth are longer and more nearly horizontal. The flattening of the posterior margin and its extension downwards give the shell a suggestion of a shortened Cucullaea, an effect strengthened by the horizontality of the outer teeth. Add to this the fact that the type of Cucullaea ponderosa was lost, and we have a fine little pitfall for the palaeontologist. Fortunately, the missing shell has been found, so the error can be corrected. The specimen used by Suter (1914, p. 36) for his description of C. ponderosa (and now in the Dominion Museum) is a Glycymeris robusta. It measures 102 mm. high, 94 mm. long, and 80 mm. thick (both valves). The measurements and locality given by Suter are taken from Hutton's description, and it will be seen that, while the height of both is 102 mm., the length of the Cucullaea is 109 mm., which can hardly be said to “ answer very well” to that of the 94 mm. length of the Glycymeris. The locality of the shell is unknown, but it may be Hutton's G. globosa from Hicks Bay (1873, p. 28).

In an unpublished “ List of Tertiary Mollusca in the Canterbury Museum ” Hutton gives a photograph of a shell from Porter River which., he identified as Cucullaea ponderosa. It is almost certainly a Glycymeris robusta.

Glycymeris cordata (Hutton). (Plate 2, figs. 1, 6.)

1873. Pectunculus (?) cordatus Hutton, Cat. Tert. Moll., p. 28.

1886. Pectunculus cordatus Hector, Outline of N.Z. Geology, p. 51, fig. 6.

Shell comparatively small, heavy, much inflated, roundly trigonal; beaks very broad and high; anterior end broadly curved, sloping steeply from the umbo; posterior end attenuated below; anal depression moderate; sculpture about 35 obscure ribs with linear interstices; ligamental area short and wide, with faint, spaced striae; teeth strong, steeply inclined, not greatly invaded by the area, 8 on each side.

Height (right valve), 34 mm.; length, 31 mm.; thickness, 15 mm.

Lectotype in collection of N.Z. Geological Survey.

Locality.—Wairoa River, Nelson, in large numbers in a brown sand at the mouth of the Wairoa Gorge.

This species is easily identified by its somewhat small size, and the very large, broad beak.

Records to be eliminated from N.Z. Geol. Surv. Pal. Bull. No. 8:—

Page 59. White Rock River, loc. 165 = G. globosa (worn).

Page 69. Left bank, Waitaki River = G. waitakiensis n. sp.

None of the other specimens recorded by Suter could be found, so that no example of G. cordata was seen from, other than the type-locality.

Glycymeris hurupiensis n. sp. (Plate 2, figs. 4, 5.)

Shell large, heavy, inflated, subquadrate when young but becoming obliquely subrhomboidal at maturity; beaks very prominent, anterior dorsal margin long, sloping gradually to the somewhat narrowly-convex anterior margin; posterior dorsal margin shorter, posterior margin broadly

convex; anal depression moderate; sculpture about 30 broad, flat radials with linear interstices generally obscured towards the edge of the adult by strong, waved growth-lines; inner margin coarsely crenate; ligamental area long and wide, with about 6 lightly-incised ligamental lines; teeth of variable strength, 10–12 on each. side.

Height of holotype (right valve), 58 mm.; length, 63 mm.; thickness, 25 mm.

Localities.—

1037—Sandstone, Hurupi Creek, Palliser Bay.

1039—Basal grits, Whakapirihika, east shore, Palliser Bay.

Holotype in collection of N.Z. Geological Survey.

This species may be distinguished from G. huttoni by its fewer and broader ribs, by its less-pronounced anal depression, and the ovate shape of the adult shell.

Glycymeris kaawaensis n. sp. (Plate 2, figs. 2, 3.)

Shell large, heavy, inflated, rounded, trigonal almost equilateral; beaks very prominent,. anterior dorsal margin sloping fairly steeply to the convex anterior end; posterior dorsal margin short, posterior end broadly convex; anal depression very weak; sculpture 30–35 broad, flat ribs with linear interstices, crossed by strong waved growth-lines towards the margin; ligamental area moderately long and very wide, with 6 somewhat shallow ligamental grooves; teeth short and stout, posterior ones curved; the holotype has 17 in front and 13 behind, 5 at each end free of the area (this, however, seems an exceptionally large number, as paratypes have as few’ as 3 whole, teeth at each end and 2 or 3 rudimentary ones truncated by the area).

Height of holotype (right valve), 61 mm.; length, 65 mm.; thickness, 27 mm.

Locality.—996—Kaawa Creek.

Holotype in collection of N.Z. Geological Survey.

Remarks.—This species can be distinguished from G. huttoni and G. hiurupiensis by its much more sloping shoulders, and the short, curved posterior teeth; from the former also by its anal depression being practically absent, and from the latter by its slightly narrower ribs.

This is the shell identified by Suter as G. globosa (Hutton): J. A. Bartrum (1919, p. 104).

Although most of the large shells are thick and heavy, giving a strong resemblance to G. huttom, this does not show the true nature of the species, which is much closer to G. manaiaensis. This is shown not only by the shape of the growth-lines of youthful stages, but also by some large individuals which have not become thickened. One specimen, collected by Mr. Bartrum, has a low beak and is very like a large G. manaiaensis. The area, however, is short and broad, as in other Kaawa shells. Viewed externally, the beak is seen to be somewhat narrower, the shoulders are more sloping, and the ribs fewer in number.

Glycymeris manaiaensis n. sp. (Plate 3, fig. 4; Plate 4, fig. 4.)

Shell large, of moderate weight, obliquely ovate, inequilateral, moderately inflated; beaks small, not projecting much above the dorsal margin, anterior end practically a semicircle, posterior end somewhat attenuated ventrally; anal depression absent; sculpture 30-35 flat radials with linear

interstices, becoming obsolete towards the margin, where there are numerous waved growth-lines; ligamental area short and narrow, almost smooth, with three scarcely discernible ligamental grooves; teeth about 10 on each side, fairly short and stout, bent in the middle, where there is on each side a short striated facet with raised edges.

Height of holotype (complete individual). 49 mm.; length, 53 mm.; thickness of one valve, 16 mm.

Holotype in collection of N.Z. Geological Survey.

Remarks.—This species may be distinguished from the three preceding ones by its ovate shape, very much lower beaks, moderate inflation, and small, almost smooth, area.

Localities:—

• 875—Manaia beach. (Holotype.)

• “Whakino, Hawera (Marshall and Murdoch).

• 1018—Cliff half a mile north—west of mouth of Patea River.

• ? 720—“ Upper band of Petane limestone.” (Fragmentary specimen.)

• ? 736—“ Petane and Scinde Island.” (Fragmentary specimen.)

• Lower end of Starborough Greek, Awatere Valley (Thomson).

• 218—Motunau. (Poor specimen.)

• 219—Motunau. (Poor specimen.)

• 126—Awatere Valley.

• Motunau beds, Middle Waipara (Thomson).

• Motunau beds, Lower Waipara (Thomson).

• Motunau beds, Weka Pass (Thomson). (Fragment.)

• ? 41 (red)—Moonlight Creek, Waiwhero Survey District, Westland.

The shells from Whakino, Hawera, are somewhat stronger in the beak than the type, and in some cases have more sloping shoulders. In this respect they approach the variety of G. haawaensis mentioned above. The general affinities are, however, with the Manaia shell.

Although the shells are not at all alike, this species has generally been identified as G. laticostata. This seems to have been brought about by an identification made by Zittel. The figure he published is that of an equilateral shell with flat ribs and wide shoulders. It is neither G. laticostata nor manaiaensis, but apparently belongs to Glycymeris sp. [see below, p. 74 (d)].

Glycymeris rapanuiensis n. sp. (Plate 3, fig. 1; Plate 4, fig. 1.)

Shell fairly large, not heavily built, disc longitudinally oval, inequilateral; beaks fairly prominent; anterior dorsal margin almost horizontal, anterior end broadly convex; posterior end slightly sinuated above by the anal depression, which is bounded ventrally by a low ridge running from the umbo and making a blunt angle in the posterior margin; sculpture, about 26 very broad flat ribs with linear interstices obscured toward the margin by waved growth-lines; ligamental area small, with 4 lightly incised ligaraental grooves; teeth short, curved, with short striated facets at the angle.

Height of holotype (right valve), 45 mm.; length, 52–5 mm.; thickness, 16 mm.

Holotype in collection of N.Z. Geological Survey.

Locality.—895—Mouth of Rapanui River. (Collected also by Mr. L. Grange.)