internal characters of the shell were unknown. From the external appearance Zittel referred the shells to Mytilus. Trechmann (1918, p. 202) suspected that the species might belong to the Myalinidae. Later Wilckens (1927) placed the species in Myalina, and proposed a new subgenus, Maoria, for the problem species, though the internal features were still unknown. However, the name Maoria was later found to be preoccupied, and the species was again referred to Mytilus (Marwick, 1946).

Two recent statements have been made on the generic position of problematicus. Marwick (1953, p. 66) has stated that there is no evidence to indicate that the species is Myalinid. On the other hand, Avias (1953, Pl. 24, figs. 1–3) considers that the species is more closely allied to Myalina than to Mytilus.

None of the writers who suggested that M. problematicus might be Myalinid have stated why they disputed the affinity with Mytilus. The reason is probably because problematicus is inequivalve, a condition rare in the Mytilidae, and common in the Myalinidae.

Mytilus mirabilis Trechmann

The relative inflation of the valves is connected with another problem of M. problematicus, that of its relationship to M. mirabilis Trechmann. Zittel (1864) had based the species Mytilus problematicus on a relatively little inflated right valve. In 1918, Trechmann erected a new species Mytilus mirabilis for high inflated shells (Pl. 19, figs, 1, 2.)^* that occur with M. problematicus. These had previously been confused with Gryphaea (Park, 1903, p. 397; Bell, Clarke and Marshall, 1911, p. 20; Marshall, 1911, p. 17; Ongley, 1940). Trechmann (1918, p. 202) stated that the two species were connected by a series of intermediate forms, and considered that mirabilis “suggests … an enormously overgrown specimen of Mytilus problematicus in which the valves have become strongly arched”.

New Name for M. mirabilis Trechmann. The name Mytilus mirabilis is a still-born homonym of Mytilus mirabilis (Lepsius, 1878, Das weslich Süd-Tirole, p. 365., Pl. 6, figs. 3 a-c). Originally placed in Gervillea? by Lepsius, the species was transferred to Mytilus by G. Boehm (1884). Even though Trechmann's species is considered to be a subjective synonym of problematicus it should be renamed, following the recommendation of the International Commission for Zoological Nomenclature (Hemming, 1950, p. 49). Therefore the new specific name trechmanni is here proposed to replace mirabilis Trechmann non Lepsius. The newly named type was figured by Trechmann (1918, Pl. 20, fig. 20 a, b), and is kept at the British Museum (Natural History).

Growth Stage of M. problematicus. Both Wilckens (1927) and Marwick (1953) treated “Mytilus” trechmanni (= mirabilis), tentatively at least, as a distinct species. Wilckens (1927, p. 14) referred to mirabilis as a “puzzling pelecypod”, and noted that in its early growth stages it closely resembled M. problematicus. Avias (1953) considered that the two were conspecific, stating that mirabilis was the gerontic growth-stage of problematicus.

To test the relationship between M. problematicus and M. trechmanni, all the relatively undistorted specimens of the two forms at the N.Z. Geological Survey have been measured and graphs of dimensions have been plotted. One of these graphs, simplified, is presented as Text-fig. 1. Judging from the graphs, it seems impossible to separate early growth stages of trechmanni, indicated by growth lines, from specimens of problematicus. It appears that the two “species” are merely different growth-stages of one species. Therefore, following Avias (1953), trechmanni (=mirabilis) is put into subjective synonymy of problematicus.

Text-fig. 1.—Graph of right valves of little inflated Manticula problematica Zittel and highly inflated M. trechmanni n. sp, plotted from specimens and their growth-lines, showing individual ontogenies.

The best illustration of the change from a form typical of problematicus to one typical of trechmanni is given by Avias (1953, Pl. 24, figs. 1–3). Growth lines show that the young form of trechmanni is little inflated, like problematicus. As the right valves, or growth lines of other specimens, increase in size the inflation increases rapidly, particularly at the posterior-ventral margin. In the growing animal this resulted in a considerable tilting of the first formed part of the shell. Little further increase occurred in length or height; indeed the margins of some shells became slightly constricted. As a result the valve became highly arched and semi-circular in outline.

Few Mytilid species are inequivalve (Nicol, 1958, p. 58) but there are exceptions. A recent Australian genus, Stavelia Gray (1858), is inequivalve, with a short ligament carried on very stout, almost shelf-like projections (Iredale, 1939, p. 410). The hinge has no teeth. Unlike Manticula, Stavelia is twisted in outlie, and the anterior ventral margin of either the right or left valve bulges into that of the other valve, so that the margin is sinuous.

Inequivalve genera are more common amongst the Myalinidae then in Mytilidae. But none are known to be so grossly gibbose as Manticula problematica, and it is generally the left valve of Myalmids which is inflated, not the right. The ligament area of Myalinid genera is broad as in Manticula, though as a rule it is longer and more externally placed. Usually it lies on a thickened hinge area rather than on a plate. The ligament area of the upper Palaeozoic genera Septimyalina Newell (1942) and Atomodesma Beyrich (1864) shows some similarities to that of Manticula, in that it overhangs an umbonal chamber. But in these genera the left valve is more inflated than the right valve, as in other Myalinids, and an umbonal plate is developed. In addition, teeth occur in Septimyalina.

According to Newell (1942, p. 33), the outer ostracum of both valves in Myalinid genera is typically composed of polygonal prisms of calcite at right angles to the surface. By contrast, the outer ostracum in Mytilids may comprise oblique prisms (Bøggild, 1930, p. 42). The outer ostracum of Manticula shows neither of these characteristic structures: it is very thin, and apparently homogeneous, as may be the case in some Mytilids (Bøggild, 1930), and in some Myalinids. Most of the ostracum is made up of a complex lamellar layer, unlike the shell structure described for any Mytilid or Myalinid.

Manticula problematica (Zittel). Figs. 1–4, 26–30.

• 1864. Mytilus problematicus Zittel: 28, Pl. 8, figs. 1a, b.

• 1910. Mytilus problematicus Park: 72, Pl. 3, figs. 1a, b.

• 1912. Mytilus problematicus Marshall: 183, fig. 102.

• 1918. Mytilus? problematicus Trechmann: 201, Pl. 20, fig. 8.

• 1918. Mytilus? mirabilis Trechmann: 202, Pl. 20, figs. 9a, b.

• 1927. Myalina? (Maoria) problematica Wilckens: 13, Pl. 2, figs. 3–6.

• 1927. Myalina? mirabilis Wilckens: 14, Pl. 3, fig. 1.

• 1939. Maoria problematicus Marwick in Ongley: 37.

• 1946. Mytilus problematicus Marwick: 27.

• 1953. Mytilus problematicus Marwick: 66, Pl. 4, fig. 1.

• 1953. Mytilus mirabilis Marwick: 66, Pl. 5, figs. 1, 2.

• 1953. Mytilus problematicus (= M. mirabilis) Avias, Pl. 24, figs. 1–3.

• 1956. Oretia sp. Marwick in Wood: 60.

• 1959. Mytilus problematicus Waterhouse in Kingma: 23.

• 1960. “Mytilus” problematicus Grindley and Waterhouse: 262.

• not 1878. Gervillea mirabilis Lepsius: 365, Pl. 6, Figs. 3 a-c.

Type Specimen. Specimen figured by Zittel (1864, Pl. 8, figs. 1a, b) and kept at the Vienna Museum. Flügel (1959) reported that the specimen could not be found.

Type Locality. Wairoa Valley, south-east of Spring Grove. The exact position is unknown.

Material. Several hundred specimens are available in the collections at the N.Z. Geological Survey, Lower Hutt.

Description. The most striking feature of the species is the grotesque inflation of large right valves, in which the width of the shell may exceed the length and height. Left valves of equal height are only about a quarter to a third as wide.

The umbones are anterior and prosogyrous. The anterior margin is wide and recessed, with a shallow byssal gape, and a weak beak ridge defines a shallow concave false lunule. In right valves the false lunule is subdivided by a second ridge. Behind the beak the short dorsal hinge is inclined at about 80° from the anterior margin. For about 5 mm the angle between the posterior dorsal margin and the line of maximum inflation (angle α of Newell, 1942, p. 25) is low, measuring 20° to 30°. This angle rapidly increases to 50°, and is fairly constant over the rest of the shell.

Variation in shape is considerable. At one extreme specimens are particularly high, with a mytiliform appearance. These high specimens have a narrow umbonal angle of 70°, and angle α is as much as 60°. The shell is well inflated near the anterior margin, and the posterior wing is well developed.

Other specimens are elongate and oval in outline. In these the beak is blunt, with an umbonal angle measuring 90° or 95°. Angle α is low, measuring 50° to 55°. The anterior margin is little inflated and extends a little in front of the beak. The posterior wing is not so well defined in these elongated shells.

The ornament is variable. Some specimens are almost smooth. Others have fine concentric wrinkles, and others strong concentric wrinkles, which pass into overlapping lamellae near the beak. In addition, the shell surface is covered with very fine radial filae, 3 to 5 occurring in 1 mm. Coarser radial costae are visible on internal casts of a few specimens.

Internal. Hitherto the internal features of problematica have never been described: they were not visible in Geological Survey collections in spite of the abundance of material. Most of the specimens which show internal details have been collected and prepared by the writer. Not all details are yet known. However, it is found that the hinge is edentulous, unlike Mytilus (s.s.), and that the ligament rested chiefly, if not entirely, upon a triangular plate, marked by growth-lines and weak subvertical striae (Fig. 3). On the right valve the area is set a little below the commissure, and is weakly concave. On the left valve the area is more obscure, and seems to be at the level of the commissure in one specimen. Because the specimens are somewhat distorted or broken, it is not clear if the ligament also lay at each side on narrow raised shelves as though modified from an Ostreiid or Pteriid pattern. However, this does not seem likely. A short ridge is developed along the shell margin behind the plate. It somewhat resembles the nymph-ridge in Mytilids, but probably served only to strengthen the shell rather than support a ligament.

Part of the posterior muscle scar is visible on a left valve collected by the writer from GS 7615 at the junction of Manganui and Turipoto streams, near the Awakino River, in N.Z.M.S. 1, Sheet N91 (provisional edition). The scar is relatively large, extending for a quarter of the height of the valve, and is lightly impressed. It is apparently oval in shape, though the posterior margin is not well defined, so that its shape and extent are incompletely known. A small pedal scar lies above the adductor closer to the hinge, and a pallial line passes forward from the base of the scar. Traces of a small anterior scar lie close to the anterior margin, below the beak. Faint concentric growth lines lie over the surface of the scar. A scar has been observed in a similar position on another left valve from the same locality (pers. comm., Mr. J. A. Grant-Mackie, Auckland University).

Shell Structure. In large specimens the shell reaches a thickness of 6 to 7 mm, but is usually 3 to 4 mm thick. Few are sufficiently well preserved to be sectioned. Two right valves from GS 2534 with shell 3 mm thick have been sectioned parallel and at right angles to the hinge, and a left valve from beds north of Wairoa Gorge has also been sectioned along the maximum dimension. These sections show that the shell is now made up entirely of calcite. In structure it is complex and variable, and crossed by faint growth-lines parallel to the growing edge (Figs. 26–30). The outermost zone comprises clear, apparently homogeneous calcite, .1 mm thick, well preserved only in the right valve. Within is a zone 2 mm thick of bifurcating and flexuous lamellae. The lamellae are inclined inwards away from the beak, and lie roughly at right angles to the growth-lines. They seem to consist of fine horizontal or weakly oblique fibres. Extinction is parallel to the length. Between these lamellae the calcite extinguishes very weakly, and is closely wrinkled in relief. Dr. W. A. Watters, N.Z. Geological Survey, suggests (pers. comm.) that the calcite is an aggregate of minute fibres, perhaps inclined in two oblique directions. The inner margin of the complex zone is irregular and merges with a zone of either massively crystalline or apparently homogeneous calcite, up to 1 mm thick. In some sections the calcite extinguishes parallel with the lamellae; in other sections it extinguishes weakly, and is possibly made up of the fibrous aggregate. The innermost narrow zone of shell comprises two or three layers parallel to the inner margin. These are homogeneous in some sections and fibrous in others. The structure of the left valve is apparently much the same as that of the right valve.

Over the outer and inner surface, and locally within the inner recrystallized layer, are thin layers of biotite up to .2 mm thick. This association of biotite with the shell of Mollusca is common in fossils from the New Zealand Triassic (see also Waterhouse, 1960). Apparently the biotite has formed in cracks in the shell, and between the shell and matrix.

Occurrence. Manticula problematica is typical of the Otamitan stage of New Zealand. In the North Island, it occurs at many localities in the Huntly-Kawhia district (Henderson and Grange, 1926), Te Kuiti (Marwick, 1946) and the Mokau district (Henderson and Ongley, 1923). In the South Island the species is found on both limbs of a tight syncline in the Triassic beds near Nelson. It also occurs

Along the hinge of each valve is a longitudinally striated groove for the reception of the ligament, but only the anterior part of the area is visible in the specimens available. The shell is thickened under the beaks, but it is unknown if teeth are present, and the muscle scars are not seen. The shell is thin, and its structure unknown.

Rsemblances. P. spedeni is distinguished by its comparatively well rounded outline. It is less oblique than many species. P. tofanae Bittner (1895, Pl. 8, figs. 9–11; Broili, 1903, Pl. 18, fig. 20) is comparable in shape. This species occurs in the St. Cassian and Pachycardien faunas, and is also reported from the Rhaetian of the Apennines. Specimens from Sicily and Bakony have also been compared with P. tofanae by Scalia (1910) and Bittner (1901), and a left valve from the Tropites Limestone of the Himalayas was compared with P. tofanae by Diener (1906, Pl. 17, fig. 7). The European species is smaller than P. spedeni, and has a higher beak, a longer anterior ear, and longer obtuse-angled posterior wing. The concentric ornament is lower and the ligament area is higher.

Krumbeck (1924, Pl. 7, figs. 13 a, b, 14) considered that a new Ladinian species from Timor was close to P. tofanae, but the Timor shells have a larger anterior ear than either P. tofanae or P. spedeni.

A Carnian form from Bear Island, P. torelli Böhm (1903, Pl. 3, figs. 13, 16, 17, 21, 26), is moderately close in outline and ornament. It is less oblique than P. spedeni and has a much larger anterior ear and a more inflated right valve. P. torelli somewhat resembles P. angusta Saurin (1941, Pl. 1, figs. 7–13) from the upper Triassic of Indo-China. This species is also more inflated and less oblique than the New Zealand shell. Its beak is more prominent, and its anterior ears larger. As noted by Saurin, the specimen figured as P. cf. torelli Boehm by Patte (1926, Pl. 9, fig. 2) from Indo-China is close to angusta.

A closer species is P. caudata Stoppani (1858, Pl. 18, figs. 18–19) from Esino, and also described from the Pachycardientuffe by Broili (1903, Pl. 18, fig. 21), the St. Cassian beds by Bittner (1895, Pl. 8, figs. 17–18), the Grigne Group by Ronchetti (1959, Pl. 15, figs. 6, 7), as well as other Ladinian and Carnian beds of Europe. Diener (1906, Pl. 17, fig. 8) recorded a specimen with affinities to P. caudata from the Tropites Limestone of Byans, and Mansuy (1913, Pl. 3, fig. 16) compared a specimen from the Carnian of Tonkin, Indo-China. The anterior ear of P. caudata is moderately prominent, and the posterior margin meets the hinge at an acute angle. However, the species is more oblique than most of the New Zealand shells.

A. cassiana Bittner (1895, Pl. 8, figs. 6–8) from the St. Cassian beds, and Sicily, is only slightly more oblique than P. spedeni, but is smaller and more inflated. A specimen with some affinities to cassiana was recorded from Indo-China by Mansuy (1908, Pl. 17, fig. 8).

Pteria kitakamiensis Hayami (1958, Pl. 24, figs. 10, 11) from the Liassic of Japan has similar concentric ornament, but is more oblique, with a larger anterior ear.

The Problem of Megalodon globularis Trechmann

The true affinity of the species named Megalodon globularis by Trechmann (1918) has been one of the minor mysteries of the New Zealand Triassic. The lectotype of Megalodon globularis Trechmann (1918) is kept at the British Museum (Natural History), London. It is a damaged internal and external cast of a specimen with valves conjoined. Trechmann (1918) referred the species to Megalodon Sowerby (1827) on account of the thick hinge plate. However, the internal and external details of the hinge have apparently been destroyed (Dr. L. R. Cox, letter, 1958), and the real systematic position of the species is difficult to determine. Marwick, in 1953, reported that no further speciemens had been found, and left the species in the genus Megalodon.

of Wairuna trig. N.Z.M.S. 1, Sheet S170 (Provisional Edition), Fossil Sample S170/628. Collected by B. L. Wood, N.Z. Geological Survey.

Diagnosis. Unusually high in outline, anterior dorsal margin high and steeply inclined. Beak narrow and prominent, hinge short. Byssal sinus in the left valve weak. Radial ornament very fine and discontinuous.

Dimensions (in millimetres).

Description. External: The shells are relatively high, with prominent slightly prosogyrous beaks, of which the umbonal angle measures approximately 60°. The anterior margin extends almost to the lower third of the height of the shell, and the ventral and posterior margins are well rounded, the posterior margin being most extended close to mid-height. Angle α measures 60°. The greatest width lies close to mid-length in the lectotype, and at the lower third in the right valve from GS 5156.

A deeply excavated false lunule lies in front of the beaks. The byssal sinus is well developed in the right valve, and is deeply excavated under the beak (Fig. 15). In the left valve the sinus is weak though long in the specimen from GS 5156, and is apparently weak in the lectotype.

The hinge is almost entirely posterior, and has a small posterior wing. A high triangular ligament area is visible in the specimens from GS 5156. The area is marked by weak growthlines and bears a triangular ligament pit. The anterior margin of the area projects forward as a low ridge.

The ornament comprises sharply defined concentric growth-lines and steps. Very fine radial filae are present between the concentric growth lines. About two or three filae occur in 1 mm at the ventral margin of the shell, measured at abòut 20 mm from the beak.

Internal Muscle scars are faintly defined. Both the posterior and anterior impressions lie at mid-height. The anterior scar is rounded in outline, and placed close to the anterior margin below the byssal sinus. The posterior scar is apparently elongated and crescentshaped. The impressions are linked by a low ridge with a groove on each side.

The posterior part of the ligament area is supported by a short buttress, at least in the left valve from GS 5156.

The shell is more than 2 mm thick at the central margin. Its structure is unknown.

Resemblances. One of the closest species is Mysidioptera a kittlii Bittner (1895, Pl. 21, Fig. 15), from the Ladinian Red Marmor of Han Bulog and Haliluci in Bosnia, and also the Schreyeralm-Marmor of the Northern Alps and the Anisian of the Dinarides M. kittlii closely resembles M. globularis in its well rounded outline, long anterior dorsal margin, and high ligament area. Moreover, the byssal sinus of the left valve is small in both species and the ornament predominantly concentric, with fine radial striae. M. kittlii can be readily distinguished by its more roundly inflated profile, its more incurved beak, longer hinge, and larger posterior wing.

Wilckens (1909) made M. kittlii the type species of subgenus Latemaria. The subgenus is supposedly distinguished from Mysidioptera by the incurved beak and globose shell, but these characters seem to be of minor importance, and it is doubtful if the subgenus is warranted.

The well-inflated species, M. globosa Broili (1930, Pl. 22, Fig. 16) of the Pachycardientuffe is another species referred to Latemaria by Wilckens (1909) Diener (1913, p. 110) compared a shell to the same species from the Carnian Myophoria Limestone of Kashmir M. globosa agrees with the New Zealand species in its prominent beak, long false lunule, and ornament, but the European species is more elongated in outline, and has a more incurved beak than that of M. globularis.

Age. The lectotype of M. globularis and another specimen mentioned by Trechmann come from beds below Manticula problematica in Nelson, and are presumably Oretian, unless they were not in place. The specimens from GS 5156 are from

Otamitan beds. They were mistakenly mapped as Oretian by Wood (1956), but the fauna includes the Otamitan key fossil, Manticula problematica.

Discussion. The specimens from GS 5156 were referred to Oretia by Marwick in Wood (1956).

Mysidioptera riceae n.sp. Figs. 8, 12–14, 16–20.

Named after Miss Mabel Rice, former Editor of the New Zealand Journal of Geology and Geophysics.

1939. ? Gonodon mellingi non Hauer Marwick in Ongley: 40 (list).

1958. ? Atomodesma Grindley: 43.

1960. Mysidioptera n.sp. Grindley and Waterhouse: 262.

Holotype. External cast of the left valve, with the hinge and most of the internal cast, specimen TM 2101. Fig. 20.

Type Locality. GS 1431, Waiwera River, 20 to 30 chains below hairpin bend, N.Z.M.S. 1, Sheet S179 (Provisional Edition), Fossil Sample S179/475. Collected by M. Ongley.

Measured Specimens. TM 2096–7; 2101–2109, 2148, N.Z. Geological Survey.

Material. This species is represented in the collections at the Geological Survey by a few valves from a number of localities, and by nine specimens from GS 928.

Localities See Table I.

Diagnosis. Moderately small shells. Usually slightly higher than long, anterior dorsal margin high and steeply inclined. Beaks anterior, moderately prosogyrous, inconspicuous. Umbonal angle close to 90°. Posterior wing poorly developed. Byssal sinus wide in both valves. Ornament of strong concentric wrinkles and subordinate fine radial filae. Ligament area short and high.

Description. External: The shells are moderately small for the genus. They are probably equivalve or nearly equivalve. Only one specimen with valves conjoined is available, and in this the right valve appears to be a little more inflated than the left, but the specimen is slightly distorted. The zone of maximum inflation is inclined at 45° from the hinge.

The beaks are placed well forward near the anterior fourth of the shell length. They are moderately prosogyrous, and curve towards each other over the hinge, without projecting above the dorsal margin. The umbonal angle ranges from 75° to 110°, and is usually more than 90°. From the beak the anterior margin sweeps with marked concavity to below midheight, and the ventral anterior, ventral and posterior margins are well rounded. Virtually no posterior wing is developed along the dorsal margin.

In front of the beaks is a wide and deep false lunule, defined by a weak beak ridge and ornamented by strong concentric growth ridges. A byssal sinus is present in both valves. That of the right valve is deeply notched immediately in front of the beak and ligament area.

Low concentric growth ridges and steps lie over the shell, crossed by fine radial filae. About three filae occur in 1 mm at 20 mm from the beak.

The ligament area is less than half as long as the shell. It is triangular, deep-set, and between a fifth and a sixth of the shell in height. The anterior edge of the ligament area projects slightly in a sharp ridge in front of the beak. The surface of the ligament area is marked by horizontal and faint vertical striae, and one or two steps of growth. A shallow, triangular resilifer slopes obliquely over the area from the beak posteriorly to the hinge margin.

Internal: In most shells the inner shell below the anterior margin of the ligament area is slightly thickened and reflected to form a small triangular shelf.

Muscle scars are moderately impressed. The posterior scar lies close to the posteriormargin at about the upper third of the shell height, and is somewhat oval in shape, with a straight anterior margin. The anterior scar lies at the lower third, near the lower end of the false lunule, and is more rounded in outline. In some specimens it is raised. The two scars are linked by a simple pallial line.

The shell is as much as 2 mm thick at the anterior margin. In appearance it is platy, like that of an oyster. A thin section of a slightly decorticated specimen showed that the shell is composed chiefly of confused crystals of calcite, possibly recrystallized from aragonite. A thin inner layer .5 mm thick is comprised of crossed lamellae, and a thin outer layer .5 mm thick is comprised of clear homogeneous calcite. Another valve has a layer of crossed lamellae 1 mm thick.

Variation. Insufficient specimens are available from any one locality to enable the variation within the species to be adequately studied. Most specimens are similar to the holotype in outline, but the length:height ratio, and the length of the hinge vary slightly. The umbonal angle varies in specimens from one locality, and in some specimens the beak overhangs the ligament area more considerably than usual. Strong concentric undulations are developed in some shells (Fig. 19), but other specimens from the same locality are comparatively smooth.

Resemblances. M. riceae is distinguished from other non-costate species of Mysidioptera by the wide umbonal angle, low umbo, poorly developed posterior wing, and wide byssal sinus in both valves.

The new species resembles M. globularis (Trechmann) in size, poorly developed posterior wing, short hinge line, and ornament. It is distinguished from globularis by the more elongated outline, and the more prosogyrous broader beak that possibly projects less about the hinge. In the left valve of riceae, the byssal sinus is better defined, and no buttress is developed under the hinge. These differences are slight, and riceae may prove to be a subspecies or variety of globularis.

An indeterminable specimen figured by Diener (1913, Pl. 12, fig. 13) from the Lower Muschelkalk (Anisian) beds of Kashmir is close to the new species in outline. Like M. riceae this specimen has a low beak, a long anterior dorsal margin that is steeply inclined from the hinge, and a rounded outline. The ornament comprises concentric lamellae. The Kashmir form is distinguished from the New Zealand species by the larger, less prosogyrous beak, larger posterior wing, and entire lack of radial ornament. The depth of the byssal sinus is unknown in the Kashmir shell. Possibly the Kashmir shell is conspecific with M. exima Diener (1913, Pl. 11, fig. 8) from the same horizon. M. exima is distinguished from M. riceae by its large size, small byssal notch and lack of radial ornament.

A large right valve associated with Manticula problematica (Zittel) in Carnian beds of New Caledonia was compared with M. exima by Avias (1953, Pl. 23, Fig. 4). The New Caledonian shell is well inflated, lacks radial ornament, and is very large, but shows some similarity in outline to the New Zealand species. Unfortunately the details of the hinge and byssal sinus are not described.

A less elongated shell with a steeply inclined anterior-dorsal margin, low beak, and small posterior wing is the chiefly Ladinian species M. vixcostata (Stoppani) of Europe. Bittner (1895, Pl. 20, Figs. 24–28) figured the species as M. cf. and aff. vixcostata Stoppani, and Broili (1903, Pl. 20, Figs. 17, 18) recorded the species as M. incurvostriata (non Woehrmann, fide Cox, 1924). Cox (1924) has described the species from the Carnian of Jordan. Costae are stronger than in the New Zealand form.

The Ladinian and Carnian species, M. woehrmanni Salomon (1895, Pl. 5, figs. 15–17; Bittner, 1895, Pl. 20, figs. 7–10; Broili, 1904, Pl. 20, figs. 12–14) from Italy, Sicily and the southern Alps of Europe, is somewhat similar in outline to elongated specimens of the new species. The European species is distinguished from M. riceae by its slightly protruding beak, and by its short anterior-dorsal margin that is less steeply inclined from the beak. Moreover the byssal notch of the European shell is less well defined. No radial ornament is present.

Two supposed varieties of M. ornata Salomon from the Ladinian Esino beds are moderately similar to the new species in outline. These are var. lombardica Bittner (1895, Pl. 21, fig. 13) with weak radial ornament, and var. laevigata Bittner(1895, Pl. 21, fig. 12) with a deep byssal sinus and no radial ornament. The varieties have similar low, but less prosogyrous beaks than the new species, and the anterior-dorsal margins are shorter.

Like the New Zealand form, M. obliqua Broili (1903, Pl. 22, figs. 13, 15) from the Pachycardientuffe, and also reported from the Ladinian by Wilckens (1909), is a well-rounded and moderately well inflated species, with weak radial ornament. However, the anterior dorsal margin of M. obliqua is short, and the beak is more prominant than in M. riceae.

Discussion. Specimens from the type locality GS 1431 were tentatively referred by Marwick in Ongley (1939) to a species from Nugget Point that was identified as Gonodon mellingi Hauer by Wilckens (1927). The specimens from Nugget Point lack the wide byssal sinus and well defined false lunule of Mysidioptera.

A specimen from GS 6043, Eglinton, was mistaken for Atomodesma by Grindley (1958). It is recorded as Mysidioptera n.sp. by Grindley and Waterhouse (1960).

Occurrence. M. riceae is a rare to frequent species in Otamitan beds, and is commonly associated with Manticula problematica (Zittel). It occurs in the North

of the right valve is convex and bears two low ridges. A strong convex buttress lies under the beak on the left valve; it is not certain if grooves are developed on the buttress. In front of the buttress is a deep notich to allow the passage of part of the byssus.

Shell Structure. Insufficient specimens with shell preserved are available for a full study of the shell structure. A radial section of a left valve shows three layers. The inner and outer layer comprise clear homogeneous or weakly prismatic calcite. These layers are less than .1 mm thick and extinguish at right angles to the length. The medium layer, more than 2 mm thick, is complexly lamellar. The lamellae pass outwards at 45° from the inner layer away from the beak, and are undulous or form small zigzags. At the outer third they turn through 90° (Fig. 31).

A radial section of a right valve near the byssal sinus shows a thick layer of finely lamellar shell, with an inner homogeneous layer that possibly represents the inner ostracum. Away from the anterior margin, the outer layer becomes completely recrystallized, and the inner layer is twinned.

Synonymy. Some changes are necessary for the synoymy of Hokonuia limaeformis Two specimens may be added to the list given by Marwick (1953, p. 60). As O. Wilcknes (1927, p. 15) suggested, a specimen figured by Zittel (1864, Pl. 7, Fig. 2) as a steinkern of Astarte, and refigured by Park (1910) without a name, belongs to Hokonuia limaeformis. Also one of the specimens described by Trechmann (1918, p. 209) as Megalodon globularis is a right valve of Hokonuia. The hinge of this specimen, a right valve, was excavated by the writer at the N. Z. Geological Survey, and found to possess the long anterior ear typical of Hokonuia.

A specimen identified by O. Wilckens (1927, p. 28, Pl. 5, Fig. 10; Pl. 6, fig. 4) as Gonodon mellingi Hauer from Nugget Point was referred to Hokonui limaeformis by Marwick (1953, p. 61). This determination was repeated by Campbell (1955). But the Nugget Point specimen examined at the Geological Survey is not Hokonuia. It is equivalve, and the right valve lacks the prominent anterior ear and the byssal ctenolium of Hokonuia, and has only a weak byssal sinus.

Marwick (1953, p. 48) suggested that Cardiomorpha nuggetensis Trechmann (1918, p. 189, Pl. 21, Fig. 7) might be conspecific with Hokonuia limaeformis. A comparison of a plaster cast of the holotype of C. nuggetensis from the British Museum (Natural History) with Hokonuia shows that the two are not congeneric C. nuggetensis is equivalve, and lacks the right anterior ear and the strong byssal sinus of Hokonuia. Unfortunately details of the hinge in C. ? nuggetensis are unknown. In outline the species resembles a cast from Nugget Point that was erroneously referred to Hokonuia by O. Wilckens (1927, Pl. 6, fig. 3). Marwick's synonymy cl. parki and rotundata is accepted uncritically.

Affinities Trechmann (1918, p. 204) has already commented on the similarity of Hokonuia to Pergamidea Bittner (1891) from the Triassic of Balia Maaden, Asia Minor. Like Hokonuia, Pergamidea has a large anterior ear on the right valve of the type species, and a wide byssal sinus. A deep notch lies under the beak of the left valve, as in the left valve of Hokonuia. In Pergamidea attalea Bittner (1891, Pl. 3, Fig. 4), a synonym of P. eumena according to Krumbeck (1924, p. 209), the ligament area of the left valve is concave, and is divided under the beak by a low convex buttress, somewhat like the anterior part of the ligament area in Hokonuia. Pergamidea has an anterior ear in both valves, unlike Hokonuia, and the right ear is directed forward in the plane of the commissure.

The same genus is reported from the upper Triassic of Timor by Krumbeck (1924). In the Timor species the anterior ear is small and the ligament area of the left valve has an oblique resilifer, and is raised anteriorly. A buttress lies at the anterior end of the ligament area and the notch below the left beak of the Timor shells is scarcely developed.

Superficially Hokonuia is remarkably like the Permian genus Eurydesma Morris (1845). Both genera are similar in outline and large size. A “dental process” projects from the right valve of Eurydesma into the left valve, recalling the right anterior ear of Hokonuia. A deep byssal notch is present in the right valve and a notch and buttress lie under the beak of the left valve of both genera.