present even on the second whorl, separated from the lower part of the whorl by a distinct, broad angle in whorls three and four; this angle narrowed to about 90° by the growth of a strong spiral cord on the fifth and sixth whorls. This cord corresponds in position to the uppermost cingulus of P. acuminata, but is not nearly so strong, the holotype having it stronger than is usual in paratypes. The body whorl has three rather narrow, raised, rounded cinguli, the median slightly weaker than the others. Beneath these three is a fourth obsolescent cingulus, much lower and weaker; on the base four widely spaced, thin, distant, spiral cords, regularly decreasing in strength from the highest to the lowest, which is very weak. Paratypes usually have only three cords on the base. The entire shell covered by weak but distinct spiral lirae. Blunt, low tubercles are present on the uppermost cingulus (about 18–20 per whorl) and sometimes on the median cingulus. In some paratypes the tubercles are very weak.

Dimensions. Height, 35 mm. Diameter, 23 mm.

Localities. V.300 type and one other; V.301 two specimens; V.34 one specimen. G.S.2626, 4 specimens.

Remarks. The type has been selected as the best preserved specimen available. In some respects it represents a slightly atypical modification of the species. The whorl is more quadrate than in some paratypes which have rather convex whorls. The spiral lirae are less strongly developed, the cinguli are narrower, and the lowest stronger, than in some paratypes.

In aspect the shell is between acuminata and rotunda. The holotype and some paratypes are more nearly allied to the former species, other paratypes (with convex whorls, broader cinguli and stronger spiral lirae) are closer to P. rotunda. The specimens nearer acuminata frequently have slightly higher spires—e.g., the type, 35 mm. × 23 mm., while specimens nearer rotunda tend to have a more depressed spire—e.g., a paratype 33 mm. × 23 mm.

Because of this intermediate character and also because of its stratigraphic position P. mangaoparia is thought to be possibly the stock from which P. acuminata and P. rotunda diverged. From its broad cinguli P. rotunda may, however, be more nearly related to P. canaliculata (Zitt).

Pelicaria rotunda n.sp. Plate —, figs. 5 and 7.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Shell lower than acuminata but much broader, proportions more like those of media; spire same height as aperture. Whorls 6, first three quite convex, later three with a flat (not channelled) shoulder, but with sides still fairly convex; on body whorl, which is very broad, the convexity continuing over the base without any angulation at the periphery. Sculpture: First 1 1/2 whorls smooth, then developing about 10 fine spiral threads; on latter part of third whorl two broad ribs appear, increasing in strength on fourth whorl to two low, slightly rounded, broad cinguli, slightly narrower than the interspace. On body whorl a third cingulus is present at level of suture, somewhat broader and less raised than the other two, though this is not always so in paratypes. On the base is a very obsolescent cingulus which is better developed in paratypes but never very strong. The spiral lirae continue to be well developed over the whole shell except on the flattened shoulder where they are obsolescent.

In each of the grooves between the three main cinguli is a strong spiral thread. The upper of these is seen to arise on about the fourth whorl, by the strengthening of one of the spiral lirae. Numerous small nodules are present on the cinguli

(specially on the upper) of the fourth whorl, and these increase in strength to blunt low tubercles and become rather widely spaced on the fifth and body whorls (16 on body whorl of holotype, 16–20 on paratypes). Base ornamented with 8 fine, distant spiral cords. Some paratypes have fewer than eight, and stronger cords on the base.

Dimensions. Height, 43 mm. Diameter, 30 mm.

Localities. V.205 (holotype and 27 paratypes); V.7, V.25, V.59, V.204, V.216, V.272, V.274, V.278.

Remarks. P. rotunda is distinguished from P. acuminata by its relatively depressed spire, weaker, broader and tuberculate cinguli, stronger spiral lirae, more numerous spiral cords at the base, and the unchannelled or only slightly channelled shoulder.

Immature specimens of P. rotunda are very similar to P. media, but can be distinguished by the fact that corresponding whorls on rotunda are much larger, there are more initial spiral threads, usually rather narrower cinguli, and the tubercles are more strongly developed than on media, and appear earlier (third to fourth whorl on rotunda, fifth whorl on media). Mature specimens are much larger than media and rarely have the suture even slightly channelled (channelling is rather inconsistent in media however.)

Pelicaria rotunda is a very unstable species, giving many varieties most of which probably occur locally only. At present none of the varieties seem to be worth naming, but a few of the most noteworthy may be mentioned.

(a) Localities V.59 one specimen and V.216 (see Fig. 7, Plate —) one specimen: Pelicaria rotunda typical in all respects except that a heavy cord occupies each of the spaces between the three main cinguli. Each of these cords arises by enlargement of one of the spiral lirae, at about the fourth whorl.

(b) Localities V.274, 1 specimen and V.7, 1 specimen: As above, but with only one cord, between the two lower cinguli.

(c) Locality V.204: There are 4 specimens of typical P. rotunda. Besides there is one specimen in which the sculpture is obsolete, except the spiral lirae, and the whorls are convex, giving a form very like P. convexa. It lacks the flattened border to the suture on the earlier whorls and the channelling on the body whorl, and it has fewer and weaker spirals than convexa. One other specimen is intermediate between this extreme variation and the true P. rotunda.

(d) The number and strength of the spiral cords on the base is a variable feature, but there are rarely less than five cords, and rarely more than eight. They may be little more than threads or they may be quite heavy.

The convexa-like adaption of P. rotunda is from beds at about the middle of the Lower Nukumaruan. It is not suggested that P. convexa diverged at this early time, but that P. rotunda had a tendency (or capacity) to lose its heavy ornament, and very probably was the stock from which P. convexa eventually arose.

Intermediate forms between P. rotunda and P. media occur at the base of the P. media zone (see remarks under P. aff. media) which are taken as definite evidence that P. media is descended from P. rotunda.

It might also be indicated here that P. rotunda seems to be the species most closely related to P. vermis (Martyn) and Pelicaria tricarinata (Lesson).

Pelicaria media might be thought equally close but its smaller size is against this. Besides, Pelicaria media, unlike rotunda, seems to have been a comparatively stable species, not disposed to change.

The stratigraphic and geographic relations of P. rotunda and P. acuminata are rather involved, as a glance at the table (Fig. 3) will show. Once specimen of P. rotunda (none of acuminata) was found in the higher part of the P. mangaoparia zone. Both rotunda and acuminata have been collected from various horizons in the P. rugosa zone, but above this only P. acuminata occurs in the Mangaopari and Makara Streams. In the area from the Whakarua Stream northwards on the other hand P. acuminata is not known, and P. rotunda ranges from the Upper Waitotaran to the top of the Lower Nukumaruan.

On the crest of the Maungaraki Range Pelicaria acuminata occurs together with forms intermediate between P. acuminata and P. rotunda in the silts between the middle and upper limestones (see column 4). The facies at this locality has more in common with that in the Mangaopari and Makara Streams than with that in the Te Ahitaitai and Whakarua Streams.

Between the Mangaopari and Whakarua areas, that is in the valleys of the Ngarara and Te Ahitaitai streams, faunas from the base to near the top of the Lower Nukumaruan include fairly typical specimens of Pelicaria rotunda together with forms intermediate between P. rotunda and P. acuminata. These intermediate forms have the high spire of acuminata, and often have narrow, raised cinguli, but the body whorl is grossly enlarged and convex, the nodules are strongly developed on the cinguli, and the base has from 5 to 8 spiral cords. Only one specimen that could be regarded as a typical P. acuminata was collected, this from V.14, the concretionary sandstone (basal Nukumaruan) in the South Branch of the Ngarara Stream. Most of the acuminate specimens come from near the base of the Nukumaruan, that is near the concretionary sandstone and Chlamys delicatulus shell-bed, which are thought to indicate shallowing of the sea in this area. It would appear that here the two distinct faunas represented to the north and south have mingled and interbred to form a hybrid swarm. This implies that they are closely related forms.

From the distribution of the two species as outlined above it is judged that P. acuminata lived in shallow waters near the shore, as is indicated by the facies in the Mangaopari and Makara Streams and on the Maungaraki Range; while P. rotunda preferred a deeper water environment, being now found in the silty mudstones in what must have been (from the structural evidence) the centre of the basin of sedimentation in lower Nukumaruan and late Waitotaran times.

Being thus restricted to different habitats P. rotunda and P. acuminata might be considered no more than ecological varieties of the same species. However, the essential differences between the two—i.e., the ratio of height to diameter, and the sculptural differences, are marked and constant where the two forms are living separated from each other. In both these features each species is more closely related to P. mangaoparia than to each other. The cinguli of P. rotunda, by their broadness, suggest a possible relationship with P. canaliculata (Zittel). The tubercles on the cinguli, the development of spiral lirae and the non-channelled shoulder, however, show that P. rotunda is far removed from P. canaliculata. It is most probably evolved from P. mangaoparia or some other form close to it. The probable phylogenetic relationships of P. acuminata and P. rotunda are summarised in Fig. 3.

Fig. 3.—Table showing suggested phylogenetic relationships of some species of Pelicaria.

If this phylogenetic relationship is valid P. rotunda and P. acuminata are not simply varieties but must be regarded as different species, evolving along different phylogenetic lines, though closely related through a common ancestor not far removed.

Pelicaria rugosa (Marw.)

1926. Struthiolaria rugosa Marwick. Trans. N.Z. Inst., vol. 55, p. 189, Pl. 13, fig. 2.

Pelicaria rugosa ranges through about 300 feet of beds in the Mangaopari Stream. Specimens agree well with Marwick's description, except that they usually have 7 or 8 spiral cords on the spire whorls.

Localities. V.302, V.304, G.S.2624 (all the Mangaopari Stream); V.294, new cutting on Awhea Road, three miles south of Martinborough.

Pelicaria acuminata (Marw.) Plate —, fig. 11.

1926. Struthiolaria acuminata Marwick. Trans. N.Z. Inst., vol. 55, p. 185, Pl. 15, figs. 11, 12.

The name should be kept for forms with the high, turreted spire notably longer than the aperture, and with five high, narrow cinguli, smooth or very weakly nodulous, on the body whorl, 3 to 5 moderately strong spiral cords on the base, and the body whorl not grossly enlarged.

Localities. V.260 Maungaraki Range, Gladstone East-Coast Road; V.14 Ngarara Stream (South Branch); G.S.2904, V.306, G.S.2610, G.S.2611, G.S.1612, all in Mangaopari and Makara Streams.

Pelicaria media (Marw.). Plate —, fig. 4.

1926. Struthiolaria media Marwick. Trans. N.Z. Inst., vol. 55, p. 187, Pl. 15, figs. 13, 14.

The type is from Castlepoint (Nukumaruan), a shell about half the size of P. rotunda, with depressed, pointed spire, low, rounded broad cinguli, and spiral threads of varying strength over most of the shell. Many specimens corresponding closely in size and general form to the type, have been collected from a zone occupying the lower part of the Upper Nukumaruan. They vary to some extent in the strength of the cinguli, which may also be either slightly nodular or smooth; the flattened border to the suture may be broad, when the sides of the whorls are