The Waitotaran Faunule at Kaawa Creek—Part I

Transactions and Proceedings of the Royal Society of New Zealand, Volume 66, 1937, Page 38

The Waitotaran Faunule at Kaawa Creek—Part I By C. R. Laws, D.Sc., University College, Auckland (1934). [Read before the Auckland Institute, June 19, 1935; received by the Editor, June 28, 1935; issued separately, June, 1936.] Introduction. Owing to the lengthy nature of the full treatment of the faunule at Kaawa Creek, it has been found necessary to submit the results of the work in two separate parts. The present paper embodies general discussion on the faunule as a whole, along with detailed consideration of the Pelecypoda. Part 2, which will appear later, contains treatment of the Gastropoda. For a number of years Professor Bartrum and his students have frequently visited Waikato Heads and Kaawa Creek for the purpose of studying the general geology of those districts. It was found impossible to devote more than a very limited amount of time to the collection of molluscs from the fossiliferous beds at Kaawa Creek, but each party brought back some material, and in this way a number of new forms has been added to the faunal list. In April of 1934 Professor Bartrum and the writer visited the beds with a view to investigating the fauna more thoroughly, and as a result the number of species and genera has been very considerably increased. Most of the new records have been recovered from sievings of matrix. In all, 119 species have been added since the faunule was described by Bartrum and Powell in 1928. Of these 86 were obtained by Professor Bartrum and the writer during their collecting this year, the sieving of matrix yielding over 70. The total number of species so far known to constitute the faunule stands at 212. Relation to Miocene Faunas. An interesting feature is the presence of a substantial Miocene element in the faunule, shown by the occurrence of Finlayella, Notacirsa,* The gastropods referred to throughout this discussion are treated in Part 2 of the work on the faunule, to appear later, in Part 2 of this volume. Parvimitra, Inglisella, the “Liracraea” of this revision, Kaweka, Waikura, Pareora, Zclandiella, Zeacuminia, and the large Polinices. Kaweka, Waikura and Hipponix centrifugalis were previously known only in the Tertiary deposits of Gisborne District—Kaweka from the Upper Miocene (Taranakian), Waikura from the Upper Oligocene (Hutchinsonian), and Hipponix centrifugalis from the Lower Miocene (Awamoan). The remaining genera, along with Pareora striolata, are very commonly widespread in Miocene beds in the South Island. The first eight genera named above constitute new records for the Pliocene of New Zealand.

A similar though perhaps not so marked Miocene element is present in the Waitotaran beds at Hawera, as has been remarked by Powell (1931, p. 90), who has drawn attention to the presence there of the following “long-range stragglers from Miocene times”—Zeacuminia, Marshallena, Comitas, Sinum, Eucrassatella, Pholadomya, Dentalium solidum, Laevidentalium pareorense, Cardium spatiosum, Pallium waikohuensis,† Previously known only from the Taranakian (Upper Miocene) of Gisborne District. Miltha, Zelandiella, and the large Polinices. Of these genera the Kaawa beds contain Zeacuminia, Miltha, Zelandiella, and a large Pilinices. The Kaawa faunule, however, shows a closer approach to Miocene faunas than does that at Hawera, for there is a greater number of species that are indistinguishable from Miocene ones, while at Hawera the approach to the Miocene is indicated mainly generically. The following species found in the beds at Kaawa are either identical with, or else exceedingly closely allied to, Miocene forms—Navicula cf. subvelata, Anomia trigonopsis, Finlayella subtriquetra, Pareora striolata, Hipponix centrifugalis, Tanea notocenica, Tanea socia, Tanea cf. inexpectata, Poirieria primigena, Austrotoma minor, Thoristella n.sp. (named subsequently). In this connection the evidence supplied by the Kaawa foraminifera is interesting, as it is in close agreement with that furnished by the mollusca. Professor Bartrum sent samples of matrix to Mr W. J. Parr, of Caulfield, Victoria, who has made a preliminary examination of them. Mr Parr states that the commonest species is an undescribed form occurring in shore-sand at Timaru, and that there are other shallow-water forms of a Recent aspect. He finds two species that occur also in the Awamoan of the Gisborne area, and that do not, so far as he knows, occur higher. There is one species that ranges from Eocene to the present day; the remainder have not so far been recorded from the Pliocene of New Zealand, although known from deposits of this age elsewhere. Discussion Concerning Correlation. The percentage of Recent species is 26.6, a figure much lower than originally suspected, for examination of the list given by Bartrum and Powell (1928, p. 160) shows the percentage at 37.0. In his recent investigation of the faunule of the Waitotaran beds at Hawera Powell (1931) gives the percentage of Recent species there at 35.62. He points out (loc. cit., p. 86) that the high Recent percentage of 59, given by Marshall and Murdoch (1921, p. 87), is due to erroneous records based on shells derived from an uplifted Pleistocene beach, and also to the confusion of ancestral with Recent forms. No such explanations of the reduction apply in the case of the Kaawa faunule, for the decrease in the Recent percentage is due almost entirely to the great number of extinct species recorded since the faunule was originally examined. Investigation of the minutae has played no small part in bringing about this substantial fall in the percentage of Recent forms present in the beds.

The species constituting the faunule at Kaawa may be classified as follows:— Pelecypods. Gastropods. Totals. Species peculiar to Kaawa beds 42 81 123 Species occurring Recent 33 15 48 Species with strong Recent affinity 3 5 8 Species previously known only in the Miocene 1 5 6 Species with strong Miocene affinity 1 6 7 Species occurring in Pliocene elsewhere in New Zealand 8 4 12 Species not definitely determinable 2 6 8 The high percentage of species that are peculiar to the beds, together with the small number found in Pliocene rocks elsewhere in New Zealand, shows that the faunule is definitely an isolated one, and suggests that correlation may prove a matter of some difficulty. When the faunule is compared with that at Hawera the difficulty of correlation becomes further apparent, for the Kaawa and Hawera beds belong to different facies, and on the whole the assemblage of species differs at each locality, there being only 19 species that are common to both. A striking feature of the faunule at Kaawa is the entire absence of such typical Waitotaran fossils as the large pelecypods Chlamys triphooki, Sectipecten crawfordi, Ostrea ingens, Cardium spatiosum. Not one of the species listed by Marwick (1931, p. 7) as mollusca characteristic of the Waitotaran has been found in these beds; and only two (viz., Miltha neozelanica and Olivella neozelanica) out of 17 characteristic mollusca in the list given by Allan (1933, p. 102). The revised faunal list for Hawera given by Powell (1931, p. 87) shows a total absence of minutae, and as minute forms are well represented in the faunule at Kaawa, correlation is made still more difficult. Not until a facies of the Taranaki Waitotaran is found in which minutae are represented can precise comparisons be made. The following analysis of the faunules at Hawera and at Kaawa is interesting:— Hawera. Kaawa. No. of species of pelecypods 36 91 No. of species of gastropods 35 121 No. of Recent species of pelecypods 19 34 No. of Recent species of gastropods 6 19 Percentage of pelecypods that are Recent 52.8 37.3 Percentage of gastropods that are Recent 17.1 15.7 Percentage of Recent species in whole faunule 35.6 26.6 As regards the Kaawa faunule, it is noteworthy that, though the total number of gastropods is considerably in excess of that of pelecypods, yet the number of the latter occurring Recent is almost double that of the former. In the case of the faunule at Hawera a similar condition holds: pelecypods and gastropods are approximately equal in numbers, but the number of Recent species of the former is more than three times that of the latter. These facts suggest that where pelecypods constitute an important element in a fossil fauna, the percentage of Recent forms should be found to

be higher than in the case where the fauna is mainly a gastropod one. Pelecypods as a whole may be less susceptible to change than gastropods; but it must not be overlooked that the former usually live embedded in the substratum and are therefore less likely to be affected by stimuli (e.g., changes in temperature; amount of light; quietness of the water) that would immediately tell on a gastropod fauna, living as it does largely on the surface of the substratum. In this connection one may recall the extended range in time of the brachiopod Lingula, an organism that inhabits the stable environment of the substratum. The analyses given above show that the discrepancy between the Recent percentages of the whole of each faunule is practically to be ascribed to the different percentage of species of pelecypods at each locality. There is a fairly close agreement between the percentages of Recent species of gastropods at the two localities, but this agreement alone could not permit exact correlation, for a serious discrepancy becomes apparent when the percentages of Recent pelecypods are considered. At first sight this is puzzling, and appears to be actually the reverse of what one would expect, for, remembering the suggestions made above concerning the stability of pelecypod species, if the Recent percentages of gastropods of each of two horizons differ but little, it is to be expected that the Recent percentages of the pelecypods will differ still less. It is suggested that difference in facies may in part at least explain this divergence. As has already been remarked, the Hawera faunule notably lacks minutae, whereas at Kaawa these constitute an important element in the faunule. Our knowledge of the larger Recent species in New Zealand waters is much more complete than that of our Recent minutae (for almost every haul of the dredge still continues to bring to light new species of minute molluscs), so that, where one faunule contains a higher percentage of minute forms than another, it is expectable that the percentage of Recent species will be lower for the former. At Kaawa the percentage of minute species of pelecypods is greater than the percentage of minute species of gastropods, and, in view of what has just been said, this fact must play some part in explaining the divergence indicated by the figures. The beds at Hawera were laid down at a distance from littoral communities and at a depth ranging from 25 to 30 fathoms (Powell, 1931, p. 90), while those at Kaawa, as will be shown later, accumulated near a shoreline in waters that are believed not to have exceeded a depth of 15 fathoms. Species inhabiting deep-water stations live in a more stable environment than those inhabiting shallow waters along a shoreline, so that one would expect, in comparing the percentages of Recent species of two fossil faunas, that the one that existed at greater depths would contain the higher percentage of forms that are still living. If, however, two fossil faunas of the same age and from like facies are compared with a Recent fauna from a similar facies, the fossil faunas should yield Recent percentages that approximate more or less closely. Summarising, it is seen that the majority of the species do not permit any definite comparisons being made; the environment in which the Kaawa faunule existed has, so far as is yet known, no

counterpart in the Waitotaran elsewhere, so that the faunule is an isolated one with a large number of species peculiar to the beds. For this reason fossils characteristic of the Waitotaran are practically absent, and one cannot rely on a study of the percentage of Recent species. The genus Heligmope, which occurs in these beds and which also has been recorded from the beds at Hawera by Powell (1931, p. 91), is an important fossil. It is a short-lived genus, and its significance in correlating widely separated faunules has been ably discussed by Finlay (1931). Heligmope along with Miltha neozelanica and Olivella neozelanica are strong links connecting the Kaawa beds with those of Waitotaran age in Taranaki. Heligmope is one of the pelagic Janthinidae, so that its presence in deposits of different facies is not surprising, whilst the Miltha and Olivella are evidently forms with a sufficient range in depth to enable them to pass from one community to another, provided the character of the bottom is suitable. The persistence of certain Miocene genera and species into the Waitotaran at both Hawera and Waipipi, remarked on by Powell (loc. cit., p. 90), has already been referred to in this paper. The similar though rather more pronounced Miocene element that characterises the faunule at Kaawa Creek provides further data by which correlation can be attempted. Henderson and Grange (1926, p. 60) state that Marwick considers that the Kaawa beds correspond with or are slightly older than the Waitotaran Stage of the Wanganuian, and, as far as our knowledge of Waitotaran faunules permits, the present investigation seems to substantiate Dr. Marwick's opinion of the age of these beds; but confident correlation cannot be made until a faunule associated with undoubted Waitotaran rocks of like facies in the Taranaki section has been discovered. Character and Depth of Habitat. The majority of the species indicate that the beds were deposited at a depth ranging somewhere between 10 and 15 fathoms. There are present, however, certain forms that are inhabitants of the intertidal zone, viz., Lepsia and Nerita, while Trochus is commonly found low along the strand, though Mr Powell has informed the writer that Trochus tiaratus is not uncommon down to 13 fathoms in Auckland Harbour. On the other hand, Cuna, Benthocardiella, Uberella barrierensis and particularly Cratis are typical of a deeper-water habitat. This assemblage of genera makes it difficult to assign a depth within reasonably narrow limits. The simplest explanation would appear to be that many shallow-water species have been washed down and have thus become associated with species typical of deep-water stations; but this scarcely fits the facts, for the majority of the fossils and the lithology of the beds both indicate shallow waters. Throughout the 40 feet of beds constituting the fossiliferous outcrop there are several intercalated narrow bands* It has not been possible to locate these bands in the actual outcrop, for it is inaccessible from the beach. Collecting has of necessity been confined to examination of fallen blocks lying at the base of a talus. containing water-worn

pebbles of greywacke, and in these the fossils, even those of large size, are inevitably considerably rubbed. This is evidence of nearshore deposition in shallow waters, and would satisfactorily account for the presence of those species whose normal habitat is the strand or thereabouts. Even the soft sandstones making up the bulk of the beds, though they have little admixture of pebbly material, seem to have been subject to current-action, as is shown both by the presence of fine shelly material in small lenses and narrow strings and by the fact that the great majority of the shells, not excepting the minutae, have worn or rubbed surfaces. One might attempt to explain the conditions of lithology by assuming that there have been minor phases of deposition in shallow inshore waters, separating more prolonged phases of sedimentation in depths that were greater, when the deep-water species became entombed; but it seems more likely that the depth remained approximately constant, for the pebble-bands are not well enough defined to justify the conclusion that there have been oscillations of level of the sea-floor, and, further, if this were so, there would be a much larger number of deep-water species. Variation in the strength and direction of currents could satisfactorily account for the minor variation in lithology. One has yet, then, to explain the presence of deep-water forms in the beds. Since it is believed that in general waves are not effective below a depth of 30 fathoms, it does not seem reasonable to suppose that these forms have been washed inshore by wave-action. Where, however, there is a long, gently shelving off-shore sea-bottom, it is conceivable that moderately strong currents setting in towards the shoreline would transport the light shells of small species; and it is significant that all the deep-water species contained in the beds are small, light shells that could well be thus transported. The Climate of Waitotaran Times. Marshall and Murdoch (1921, p. 90) from their study of the Pliocene faunas of Taranaki have shown that during the early Pliocene the climate of the New Zealand region was a great deal warmer than it is at the present day; and Powell (1931, p. 90), in pointing out that the “long-range stragglers from Miocene times” (referred to above) failed to survive the Waitotaran, has suggested that their extinction, along with that of other warm-water types, was due to a gradual change to a colder climate at about that time. The Kaawa faunule supplies evidence that corroborates these conclusions. Miltha and Olivella, which occur in the Waitotaran of Taranaki, are present also at Kaawa, and these genera, which are not known in the New Zealand area after early Pliocene times, are now restricted to warmer seas. Cheilea, not so far recorded from the Taranaki Waitotaran, likewise indicates milder conditions of climate, and is at the present day definitely typical of tropical waters, while Hipponix favours warmer, though not necessarily tropical, seas. Characteristic Species. The order of relative abundance of the characteristic species is as follows:—Marginella hesterna, Cuna cerussata, Glycymeris modesta, Gonimyrtea concinna, Notolepton antipodum, Nuculana

tenellula, Uberella kaawaensis, Estea impressa, “Neojanacus” kaawaensis, Baryspira exsputa, Rhizorus marwicki, Nucula nitidula, Cosa kaawaensis n.sp., Cuna kaawa n.sp., Eulima n.sp.,* Named in Part 2 of the work on the Kaawa faunule. Maorimactra ordinaria, Scalpomactra scalpellum, Glycymeris kaawaensis, Tawera bartrumi, Gomphina maorum, Polinices propeovatus. Acknowledgments. The writer's sincere thanks are due to Professor J. A. Bartrum for his permission to utilise his collection of Kaawa fossils and to describe his new species; also for his assistance in photographing the species. The writer is greatly indebted to Dr. J. Marwick and Mr A. W. B. Powell for various opinions, for examining a number of specimens concerning which the writer was doubtful, and for their readiness in lending type and other material for comparisons. List of Pelecypoda from Kaawa Creek Beds. N. = Records new to the locality. R. = Species still living. Nucula ambrosia Bartrum and Powell. Nucula nitidula A. Adams R. Nucula ngatutura n.sp. N. Nuculana (Saccella) tenellula Bartrum and Powell. Nuculana n.sp. N. Anomia trigonopsis Hutton ?R. Navicula cf. subvelata (Suter). Acar opuraensis Bartrum and Powell. Barbatia novaezelandiae E. A. Smith R. Glycymeris kaawaensis Marwick. Glycymeris (Veletuceta) modesta (Angas) R. Glycymeris (Veletuceta) waipipiensis Marwick. N.gen. et n.sp. of Limopsidae N. Hochstetteria pinctagrina n.sp. N. Hochstetteria tela n.sp. N. Hochstetteria Kaawa n.sp. N. Cosa kaawaensis n.sp. N. Cosa cf. filholi (Bernard) N. Cratis pliocenica n.sp. N. Perrierina sola n.sp. N. Mytilus canaliculus Martyn. (J. Marwick) R. Trichomusculus cf. barbatus (Reeve) N. Dacrydium simulator n.sp. N. Chlamys sp. Pallium (Mesopeplum) convexum (Q. and G.) NR. Lima cf. colorata Hutton. Limatula cf suteri Dall N. Limatula maoria Finlay R. Atrina zelandica (Gray). (J. Marwick) R. Ostrea sp. (juveniles). Cuna cerussata Bartrum and Powell Cuna flctilia n sp. N. Cuna crassicardo n.sp. N. Cuna kaawa n.sp. N. Cardita aoteana Finlay R. Pleuromeris aff. lutea (Hutton). Pleuromeris miniscula Bartrum and Powell. Vencricardia penerectangularis Bartrum and Powell. Venericardia koruahinensis Bartrum and Powell Verticipronus stirps n sp. N. Benthocardiella orbicula Powell NR. Gonimyrtea concinna (Hutton) R Pteromyrtea dispar (Hutton). Divaricella cumingi (Adams and Angas) R. Miltha (Milthoidea) neozelanica Marshall and Murdoch. Zemysia zelandica (Gray) R. Thyasira flexuosa (Montagu) R. Notolepton antipodum (Filhol) NR. Chironia suborbicularis (Montagu) NR. Arthritica elongata n.sp. N. Melliteryx angulata n.sp. N. Myllitclla fragilis n.sp. N. Zemyllita n.sp. N. Semeloidea donaciformis Bartrum and Powell. Mysella koruahina n.sp. N. Virmysella cf. hounselli Powell N. Rochefortula kaawaensis Bartrum and Powell. Finlayella subtriquetra (Bartrum and Powell). Angulus spencert (Suter) R. Angulus (Peronidia) edgari (Iredale) R. Angulus gaimardi (Iredale) R. Maoritellina huttoni (Smith) R. Leptomya retiaria (Hutton) NR. Mactra discors Gray R. Maorimactra ordinaria (Smith) R. Scalpomactra scalpellum (Reeve) R. Spisula gilberti Bartrum. Zenatia acinaces (Q. and G.) NR. Resania exoptata Bartrum & Powell. Dosinia lambata (Gould) R. Dosinia (Austrodosinia) kaawaensis Marwick. Dosinia (Raina) bartrumi Laws N. Dosinia (Phacosoma) maoriana Oliver. Dosinia cf. subrosea (Gray). Dosinula marwicki n.sp. N. Tawera bartrumi Marwick. Bassina parva Marwick N. Eumarcia kaawaensis Marwick. Gomphina (Gomphinella) maorum Smith R. Novirus caudex n.sp. N. Nemocardium finlayi Bartrum and Powell. Gari lineolata (Gray) R. Gari stangeri (Gray) R. Ascitellina urinatoria (Suter) R. Notocorbula zelandica (Q. and G.) R. Panope zelandica (Q. and G.) NR. Anchomasa similis (Gray) R. Pholadidea finlayi n.sp. N. Eximiothracia pusillior n.sp. N. Myadora waitotarana Powell N. Hiatella australis Lam NR.

Description and Discussion of Species. Genus Nucula Lamarck. Type (by monotypy): Area nucleus Linné. Nucula ambrosia Bartrum and Powell. Trans. N.Z. Inst., vol. 59, p. 152, 1929. Individuals of this species show tremendous variation in many directions, to such an extent, in fact, that it has been altogether impossible to fix on any particular constant variational trend. Shell-outline, degree of convexity, number and heaviness of teeth, and strength of ornamentation are all subject to considerable variation. Similar specific instability amongst certain other forms is commented upon elsewhere in this paper. N. nitidula A. Adams, found at Kaawa in even greater numbers than ambrosia, exhibits no variation whatever in specific characters. Nucula ngatutura n.sp. (Figs. 1, 4). Shell very small, resembling N. vestigia in shape, but more definitely trigonal. Beaks about at posterior third. Lunule not apparent; escutcheon broad and long. Posterior margin straight, descending rapidly; postero-ventral corner of valve sharply rounded. Hinge with 6 anterior and 4 posterior teeth, which are blunt and heavy for size of shell; resilium small, triangular, oblique. Internal margin finely regularly crenate. Sculpture of fine concentric corrugations, less regular towards posterior, where several die out; there are numerous microscopic radial threads in depressions between concentric folds. Height, 1.1 mm. length, 1.5 mm. Type (a single right valve) in writer's collection. The posterior truncation and notably triangular shape make this species distinctive. It has not the irregular criss-cross pattern of vestigia. Genus Nuculana Link. Type (by monotypy): Arca rostrata Linné. Nuculana n.sp. There is a single left valve, badly worn and differing in shape from all other Neozelanic species. The posterior end is not much drawn out and the shell is not sharply rostrate. The beak is about at the middle. Anterior end broadly rounded. Antero-dorsal margin practically straight; postero-dorsal margin very lightly concave; basal margin broadly and evenly rounded. Surface badly rubbed. Height, 4.8 mm. length, 6.3mm. Specimen in collection of A.U.C.* Abbreviation for Auckland University College. This shell is somewhat after the style of N. nugax Marwick, a fossil in the Miocene beds of Gisborne District. Nuculana tenellula Bartrum and Powell is an exceedingly common fossil in the Kaawa beds.

Genus Navicula Blainville. Type (s.d., Children, 1823): Arca noae Navicula cf. subvelata (Suter). An adult right valve and three juvenile valves have been collected. These shells are extremely close to the Awamoan species. The most important difference is in the shape of the valve, which in the Kaawa specimens has the anterior end more rounded and descending not so rapidly posteriorly; the posterior of shell is straight and more quickly descending. The ligamental area of Kaawa specimens is longer relatively to length of shell, and the posterior flattened area above angulation has finer radials and weaker growth-lines. The surface of the rest of the valve is badly worn. Subvelata is a direct ancestor. These shells probably represent a distinct species, but the only adult specimen available would hardly make a good type, and better material is sure to be collected. Specimens in collections of A.U.C. and of the writer. Acar opuraensis Bartrum and Powell. Trans. N.Z. Inst., vol. 59, p. 154, 1929. A number of further specimens has been collected since the species was described. There is one small shell with taxodont dentition that has proved rather puzzling and could not at first be generically located. Though very small, it has extremely thick and heavy build, dentition and sculpture. It was submitted to Mr Powell, who has satisfactorily located it as a “gerontic juvenile” of the above species. He states that he has similar very small specimens which fairly obviously are Acar sandersonae Powell, yet have all the characters of senility, being extremely thick and heavy, and points out that they are not adults, for, apart from the small size, the hinge is apparently not fully developed numerically. Barbatia novaezelandiae Smith. The specimens, of which quite a number have been obtained, are all small, but otherwise indistinguishable from novaezelandiae. Limopsidae. New genus et new species. This n.gen. is not here named, as Mr Powell has it described in MS. His shell comes from Big King Island in 260 m. Another species has been found in the Kaawa beds, while there is a third at Nukumaru. The Kaawa species will be described in a subsequent paper.

Genus Hochstetteria Vélain. Type: H. aviculoides Vélain. Hochstetteria pinctagrina n.sp. (Figs 2, 3, 6). This form is allied to both pinctada Finlay and meleagrina (Bernard), but it seems nearer pinctada. The slight convexity of valve allies it to the former rather than to the latter. It is more drawn down in a postero-ventral direction than either of these species, but in shape has the characters of both. Surface of valves entirely smooth, there being no indication of the radial riblets present in the two Recent species. Prodissoconch small but distinct, further forward than that of meleagrina, but not terminal as in pinctada. Height, 3.5 mm. length, 3.5 mm. Types (odd right and left valves) in writer's collection. Hochstetteria tela n.sp. (Figs. 5, 7, 8). Shell very small, well inflated, valve protruding below and in front of beaks after the manner of Modiolus. Prodissoconch distinct, bounded by a circular fold and rising into a low cone in the centre of this. Dorsal margin straight throughout most of its length, slightly dipping in front of prodissoconch and insensibly curving to form the broadly and regularly convex posterior margin; basal margin short, flatly convex; anterior margin straight, rising very steeply. Valve-margins crenate over the postero-dorsal and antero-ventral portions, where internal radial furrows meet the edge. Externally there are about 10 fine radial threads with very wide, flat interspaces; fine, regular concentric threads ornament the inter-radial spaces, and these are slightly convex towards the beak. The pattern thus formed is not unlike a portion of the web of the garden spider. Immediately below the middle part of the prodissoconch there is a ligamental depression in the hinge; below and at either side of the prodissoconch there is a pad of vertically disposed taxodont teeth. Height, 1.7 mm. length, 1.8 mm. Type in writer's collection. Two right valves were collected. Hochstetteria kaawa n.sp. (Figs. 10, 11). This is a form closely related to the Chatham Island fossil Philobrya galerita Marwick, but it can readily be distinguished by its much more circular outline and by the presence of radials over the whole surface. The prodissoconch is bounded by a circular fold, and rises into a low dome within this; it is placed almost at the middle of the dorsal margin, perhaps slightly nearer the anterior end. There are 11 thin, thread-like radials, separated by very wide interspaces. Concentric sculpture, other than a few obscure growth-striae, is not present. Internal furrows radiate to meet both the anterior and posterior margins, causing faint marginal crenulations. Hinge and teeth similar to those of tela n.sp. (described above), with the exception that a small posterior lateral arises at the extremity of hinge. Height, 1.5 mm. length, 1.5 mm. Type (a single right valve) in writer's collection.

Genus Cosa Finlay. Type (o.d.): Hochstetteria costata Bernard. Cosa kaawaensis n.sp. (Figs. 9, 12, 16). This species has not the concave anterior side of trigonopsis, nor the sharply serrated ribs, and the adults are much smaller. It differs from wanganuica, its closest relative, in the nature of the serrations on the costae, for these are not sharp and mounted on a median ridge on the ribs, but are low and wide, and extend across the whole width of the rib. Also the intercostal concentric threads are coarser and more widely spaced than those of wanganuica. Kaawaensis and wanganuica approach each other closely in shape. The new species cannot be confused with filholi, for that species has wide, flat-floored intercostal spaces and a median thread on each rib. Costata is separable by means of the median groove on each of the costae. Serratocostata is similar to kaawaensis, but the concentric threads are much finer, fewer in number, and a good deal more distant one from another. Radial costae 11-in number. Dimensions: Dorso-ventral, 2.1 mm. antero-posterior, 1.6 mm. (holotype). Corresponding dimensions of a paratype: 2.8 mm., 2.05 mm. Type and about 30 paratypes in writer's collection. Cosa cf. filholi (Bernard). There are a half-dozen shells, and these have the ribbing and squarish shape of filholi, but differ in having slightly fewer ribs. The development of concentric threads is similar in both. There is no distinction from filholi other than that relating to the radial costae. Genus Cratis Hedley. Type (o.d.): C. progressa Hedley. Cratis pliocenica n.sp. (Figs. 13, 14, 15). Shell very small, much higher than long, somewhat quadrate in shape. Prodissoconch almost at anterior end, rising prominently above dorsal margin, surrounded by concentric rim, central dome well defined. Dorsal margin short, straight; anterior margin cut down straight and vertically for two-thirds of its distance, then retreating in a gentle curve to merge into the lightly convex ventral edge; posterior margin broadly convex; postero-ventral corner much more sharply rounded than antero-ventral one. Posterior wing of valve crossed by about 6 fine radials; the centre of valve having radials (8 in number) much stronger, coarser and widely spaced, about three times own width apart; towards anterior the radials become obsolete. Concentric sculpture consists of numerous fine threads in the spaces between the radials. Internally there are 2 or 3 strong corrugations radiating to intersect the postero-ventral edge of shell. Ventral margin lightly crenate; postero-ventral margin with several strong crenulations. Ligament-pit just below prodisso-conch, vertical. One or two strong teeth on hinge below beak, and 2 horizontal ones out at posterior end of hinge. In addition, rudimentary taxodont teeth are present, though apparently no longer

functioning, in a horizontal furrow between prodissoconch and adult hinge. It thus appears that as the adult stage is reached a new hinge with different dentition forms below the early attachment with its taxodont tooth-pads. Height, 2.2 mm. length, 2.0 mm. (holotype). Type (odd right and left valves) and 3 paratypes in writer's collection. This species is very like one already described in MS. by Mr A. W. B. Powell, a Recent form taken in deep water off Big King Island. Cratis miocenica n.sp. (Figs. 18, 20). It may be of interest to record here a further species of Cratis occurring in Awamoan beds at Clifden, Southland. The only species recorded from New Zealand prior to the present two is an Awamoan one from Gisborne District (C. ovata Marwick). The Clifden form described below has closer affinity with this than with either the Pliocene or the Recent species. Shell small, circular except for the straight dorsal edge. Prodissoconch at about anterior fourth, small, distinct. Dorsal margin long, perfectly straight; posterior, ventral and anterior margins almost form part of the circumference of a circle; there is a small concavity below prodissoconch. Ornamentation consists of distinct concentric corrugations and 8 or 9 radials, a little stronger than the concentric ridges, and on the anterior third of the valve. Internally the edge of valve is broadly bevelled, much as in Limopsis. Hinge-plate broad and strong, much wider in front than behind, where it gradually merges with the dorsal margin. There is 1 strong tooth immediately below prodissoconch, and trace of another very weak tooth. The wide anterior part of hinge bears 3 strong transverse teeth, disposed less horizontally than those of ovata Marwick. Height, 2.1 mm. length, 2.5 mm. Locality: Clifden, Southland, band 8 (Awamoan). Type (a single right valve) in writer's collection. The shape at once distinguishes this species from ovata; the limiting of radial sculpture to the anterior third of valve is also characteristic. Trichomusculus cf. barbatus (Reeve). Shell small, quadrate, its dorsal and ventral margins almost parallel; the anterior margin broadly and regularly convex; posterior margin lightly convex above and strongly convex below. Beak very near anterior end, high up, not full and prominent. Sculpture as in barbatus, and the disposition of crenate and smooth margins as in that species. Umbonal angulation rather narrow. Differs from barbatus in having a straighter dorsal margin curving downwards only a little in front. The anterior end is more broadly convex. Specimen (a single right valve) in writer's collection.

Genus Dacrydium Torell. Type (fide Suter): Modiola (?) vitrea Moller. Dacrydium simulator n.sp. (Figs. 17). In shape and external appearance this species simulates Nucula very closely indeed. The external surface is ornamented by fine, regular, concentric sculpture. Shell moderately inflated. Anterior side short, cut down almost vertically at first, then retreating ventrally in a broad gentle curve; slightly concave immediately under beak. Basal margin convex; posterior margin broadly and regularly convex. Hinge poorly developed; teeth not distinct, but hinge is thickened and raised for some distance behind the beak, and there is a weak tooth-like swelling close in front. Chrondrophore just below beak. Dimensions: greatest diameter, 4.7 mm. least diameter, 3.1 mm. Type (a single right valve) in collection of A.U.C. The writer is indebted to Mr. A. W. B. Powell for locating this species. The outline at once distinguishes it from other recorded Neozelanic forms. It is the first fossil species to be recorded in New Zealand. Chlamys sp. The record of Pecten williamsoni Zittel from Kaawa Creek is based on identification of juvenile shells by Suter. It is safer to regard the species as indeterminable until better material is available. Pallium (Mesopeplum) convexum (Q. and G.) There is a single left valve, which agrees well with the above species and fits no other described form. Specimen in writer's collection. Limatula cf. suteri Dall. There are a half-dozen valves. The external suleus and internal medial rib is generally not well defined. The ribbing is quite that of suteri. This is a new record from the Kaawa beds. Specimens in collections of A.U.C. and of the writer. Genus Cuna Hedley. Type (o.d.): C. concentrica Hedley. Cuna fictilia n.sp. (Figs. 19, 23). Shell similar to C. delta (Tate and May), but the anterior and posterior ends of the convex ventral border extend further dorsally, and merge insensibly into the straight dorsal portions of the anterior and posterior margins, there being no sub-angulation as in the Australian species. Externally the valves are very obscurely ribbed by low, wide, almost obsolete corrugations, seen only towards ventral margin. Growth-striae and plications are also present.

Lunule and escutcheon long, narrow, depressed. Prodissoconch small, smooth, rounded. Hinge heavy and teeth large, massive; dentition that typical of the genus. Posterior adductor scar considerably more elongate than anterior one. Basal margin lightly and finely crenate within. Height, 2.0 mm. width, 1.5 mm. (holotype). Type and about 100 paratypes in writer's collection. As has already been pointed out, the present species shows close affinity with the Tasmanian C. delta (Tate and May). There is another n.sp. in the Kaawa beds (described below), and this, too, shows distinct alliance with another Tasmanian species, C. concentrica Hedley. It is the only Neozelanic form with concentric sculpture so far recorded. Cuna crassicardo n.sp. (Figs. 21, 22). This shell has very much in common with C. concentrica Hedley. In shape the two species agree closely, but crassicardo is more drawn up towards beaks, and has the junction of the straight postero-dorsal margin with the semicircular ventral edge higher up. The upper portion of the shell is more inflated, and the beaks are considerably curved in. There are 16 concentric folds in both species, but those of crassicardo are separated by interstices equal to their own width, not narrower as in concentrica. Internally there are about 8 weak, widely-spaced, shallow, radial grooves separating broad interspaces, and visible only on the ventral part of valve. The lunule and escutcheon are fairly wide and considerably excavated. Prodisso-conch small, rounded, polished. A long, thin, vertical cardinal tooth bisects the hinge and rises strongly from it. Height, 2.1 mm. width, 1.7 mm. Type (a single right valve) in writer's collection. Cuna kaawa n. sp. (Figs. 24, 27). Closely related to otagoensis Powell, but differing in outline and strength of hinge. It is more broadly convex ventrally than otagoensis and not so high in relation to width. The distance across from extremities of hinge is greater, so that the anterior and posterior margins meet above in a wider angle. The posterior dorsal edge is more arched and the anterior one more excavated. The hinge and teeth are considerably lighter in build, and the crenulations within the ventral margin broader and less numerous. The greatest width in an antero-posterior direction is higher up than that of otagoensis. Both species have very weak radial sculpture on external surface. The adult shell is larger than otagoensis. Also the cardinal tooth of right valve is shorter, small and not as oblique as that of Powell's species, and the posterior socket is wider and shorter. Height, 2.2 mm. length, 2.05 mm. (holotype). Corresponding dimensions of a paratype: 2.9 mm., 2.6 mm. Type and 32 paratypes in writer's collection.

Cardita aoteana Finlay. Trans. N.Z. Inst., vol. 57, p. 459, 1927. The Kaawa shells referred here are all much smaller than the Recent form, nor are they of such heavy build. The hinge and teeth are much lighter. The number of ribs is the same as in the Recent shells. Genus Verticipronus Hedley. Type (o.d.): V. mytilus Hedley. Verticipronus stirps n.sp. (Figs. 25, 26). Shell very small, mytiliform. Prodissoconch not prominent, bounded by a low concentric fold, its surface microscopically radially striated. Anterior end regularly lightly convex, not concave below beak; posterior end sub-angled at middle, straight above angle, evenly convex below; basal margin narrowly convex; shell narrowly and evenly rounded above. There is no sculpture other than ill-defined growth-lines. Muscle-scars and pallial line not visible. Cardinal tooth present as a rounded knob (probably worn), not so prominent as that of V. mytilus. There are two oblique lateral teeth close above posterior ungulation. Height, 1.4 mm. length, 1.1 mm. Type (a single left valve) in writer's collection. This fossil is readily distinguished from the Recent genotype, for the beaks are blunter and more rounded, the anterior side quite regularly rounded, and the posterior angulation not so ventrally placed. Genus Perrierina Bernard. Type: P. taxodonta Bernard. Perrierina sola n.sp. (Fig. 29). Shell very small, ovate, inequilateral. Beaks prominent. Prodissoconch surrounded by a high, thick rim and situated nearer the anterior end. Weak radials appear towards the ventral border, and growth-stages are present here and there; surface of valve is rather worn. Valve margins smooth. Anterior cardinal short and heavy proximally, with a narrow depression in front of it and a triangular one behind, separating it from a short, thick, slightly oblique tooth just below beak. Then follows the resilium, posterior to which is an interval before the posterior cardinal (which is only moderately long and narrow and rises abruptly from ventral edge of hinge) makes its appearance. Horizontal laterals not well preserved. A broad fold runs externally from umbo to postero-ventral margin of shell. Height, 2.1 mm. width, 2.5 mm. Type in writer's collection. In shape this species simulates P. ovata Marwick, but it is much larger. The hinge is lighter for the size of the shell, and the horizontal laterals are not so well developed. Ovata has the beaks central.

Benthocardiella orbicula Powell. Trans. N.Z. Inst., vol. 61, p. 538, 1930. There is a single right valve. This species is also known from the Nukumaruan and Castlecliffian, as well as in the Recent fauna. Specimen in writer's collection. Notolepton antipodum (Filhol). Suter, Man. N.Z. Moll., p. 924, 1913. 45 valves have been collected, many of them immature. Adults can in no way be distinguished from specimens of antipodum taken off Otago Heads. Specimens in collections of A.U.C. and of the writer. Chironia suborbicularis (Montagu). Suter, Man. N.Z. Moll., p. 923, 1913. A single left valve, which constitutes a new record for these beds. Specimen in collection of A.U.C. Genus Arthritica Finlay. Type (o.d) : Kellia bifurca Webster. Arthritica elongata n.sp. (Fig. 28). This has a more elongate valve than bifurca (Webster). The beaks also are more central, and the greatest length of valve is lower down in bifurca. The hinge and teeth are lighter for the size of the shell. The single cardinal tooth is short, not so dependent as that of bifurca and oblique to hinge (vertical in bifurca). The resilium also is longer and narrower. Growth-striae present externally; the punctate sculpture is very much coarser and deeper than that on the type species. Height, 3.5 mm. length, 4.8 mm. (holotype). Type (odd right and left valves) and a number of paratypes in writer's collection. Paratypes also in collection of A.U.C. The internal limey patches, a feature frequently of bifurca, seem not to be present. In general build and character of hinge elongata is closer to the genotype than to crassiformis Powell. From crassiformis it can be distinguished at sight by different shape (the former is less elongate and subrotund) and by much lighter build and less heavy hinge and teeth. Genus Melliteryx Finlay. Type (o.d.): Erycina parva (Desh.). Melliteryx angulata n.sp. (Figs. 32, 33). Shell of moderate size for the genus, triangular in outline; beaks almost central, not prominent, directed forward. Anterior dorsal side practically straight; posterior lightly arched up; both descending at same angle. Posterior and anterior ends narrowly convex, the former more sharply rounded than the latter. Ventral margin broadly convex. Resilium posterior, elongate. Hinge light.

In the left valve there is a thin, sharp, claw-like cardinal, a long, narrow posterior lateral, and a short, fairly thick, anterior lateral; cardinal of right valve not preserved, anterior and posterior laterals and sockets for reception of laterals of left valve well marked. Sculpture of concentric growth-stages and punctuations. Height, 4.0 mm. length, 5.7 mm. Type in writer's collection. The sharply convex ends and more elongate form readily distinguish this species. Genus Myllitella Finlay. Type (o.d.): Myllitella vivens Finlay. Myllitella fragilis n.sp. (Figs. 34, 35, 37). Shell of average size for the genus, almost circular, inflated, beaks central. Sculpture typical, 5 ribs per mm. Dentition exactly as in M. pinguis, but teeth very much smaller and lighter. The posterior muscular impression is somewhat the larger. In adults there is a very heavy growth of limey material standing up from the original scars as two large tubercular masses, and these are most striking features when the shell is viewed internally. So fragile is the valve that these incrustations can be seen as dull white patches on the external surface. Height, 4.5 mm. length, 4.5 mm. (right valve). Types (odd right and left valves) in writer's collection. This species is of extremely delicate texture, and has the hinge and teeth very considerably lighter than those of any other species. Pinguis and finlayi, both of Marwick, are Nukumaruan fossils; vivens Finlay is a Recent form. There is an undescribed species in the beds at Castlecliff, distinguished by finer sculpture (6 ribs per mm., though this varies somewhat), less circular shell, heavier hinge and teeth. Zemyllita n.sp. This is represented by a single immature left valve. It is closely related to stowei (Hutton). The punctuations on the outside are coarser than those on stowei, the hinge heavier, and the teeth rather different. Specimen in writer's collection. Genus Mysella Angas. Type (by monotypy): M. anomala Angas. Mysella n.sp. Shell very small, of solid build, internally thickened. Beaks at about posterior third, small, but distinctly standing up. Anterodorsal margin long, straight for the most part, slowly descending; postero-dorsal margin short, very lightly convex, rapidly descending; basal margin broadly and evenly convex; anterior end of valve less

broadly convex than posterior. Hinge heavy; chondrophore wide and with a cardinal-like tooth on either side, the anterior one larger and heavier than posterior one; beyond each of these teeth there is a low lateral ridge with a linear groove immediately dorsal to it. Muscle scars strongly impressed. Valve margins entire. Sculpture absent; fine growth-striae only. Height, 1.0 mm. length, 1.2 mm. A single left valve was collected, but has been irreparably broken. Though small, this shell is evidently adult, as shown by the strong thickening within and by the very well developed hinge and teeth. Virmysella aff. hounselli Powell. There are three small shells, probably not adult. They show even lighter development of teeth and hinge than does hounselli, and in this respect approach closely to certain undescribed forms from Awamoan beds in North Otago. Powell (1931, p. 111) has used Virmysella to cover tellinula Odhner and his own species hounselli. Both of these have a lighter build and less massive development of hinge and teeth than do shells like M. unidentata Odhner and M. bidentifera Powell. Specimens in writer's collection. Finlayella subtriquetra (Bartrum and Powell). Trans. N.Z. Inst., vol. 59, p. 157, 1929. This species is extremely close to the Awamoan F. sinuaris Laws. A ready means of separation, however, is to be had in the more inflated left valve of sinuaris. Also the hinge-plate of sinuaris is wider and the anterior and posterior dorsal margins not so straight as those of the Kaawa species. In features of pallial sinus the two species are practically identical. Maoritellina huttoni sterrha (Suter). This form, identified by Suter, is listed by Bartrum (1919, p. 105), but the shell proves to be a small specimen of Finlayella subtriquetra (Bartrum and Powell). Leptomya retiaria (Hutton). Trans. N.Z. Inst., vol. 17, p. 322, 1885. There is a pair of valves, complete in every respect. These constitute a new record from the locality. Specimens in writer's collection.

Zenatica acinaces Q. and G. Suter, Man. N.Z. Moll., p. 971, 1913. Several more or less fragmentary valves have been collected. These make a record new to the locality. Specimens in collection of A.U.C. Dosinia (Raina) bartrumi Laws. Trans. N.Z. Inst., vol. 61, p. 548, 1930. At the time of description only a single decorticated specimen was available. It was forwarded to Dr. Marwick, who referred it to Raina. Subsequent collecting has substantiated Dr. Marwick's location, for the sculpture consists of the same elements as those of D. bensoni, the type species. Genus Dosinula Finlay. Type (o.d): Dosinia zelandica Gray. Dosinula marwicki n.sp. (Figs. 36, 38). The hinge and dentition are almost a replica of those of zelandica Gray, but the posterior lateral tooth, nymph and ligamental depression are relatively shorter, and the escutcheon is better marked. The shell is much shorter relative to height, so that the posterior dorsal edge is shorter and the anterior descends much more rapidly. Pallial sinus acute as in zelandica, but a little deeper. The sculpture, which consists of coarse, wide, rather bevelled concentric ridges, somewhat irregular in strength and here and there having several thin, sharp lamellae crowded into spaces between them, is somewhat reminiscent of that of D. crebra, but the new species has the outline more circular and the beak less prominent. Height, 40.0 mm. length, 44.0 mm. Type (a single left valve) in collection of Auckland University College. Collected by Miss E. King-Mason. A Dosinula from Starborough Creek, Marlborough, in the New Zealand Geological Survey collection, was kindly sent by Dr. Marwick for comparison. It is close to the present species, but is more elongately oval, and has the sculpture not so heavy, though coarser than that of zelandica. Bassina parva Marwick. Trans. N.Z. Inst., vol. 57, p. 619, 1927. There are 4 specimens, all of which are left valves. They are slightly more elongate than topotypes, and the hinge is somewhat lighter, but otherwise agree well with parva. Specimens in collections of A.U.C. and of the writer.

Genus Notirus Finlay. Type (o.d.): Venerupis reflexa Gray. Notirus caudex n.sp. (Fig. 39). Shell of only moderate size for the genus, elongate; beaks near anterior end as in reflexa. Lunule very narrow, scarcely differentiated. Disposition of teeth as for the genotype, but the hinge and teeth are much lighter; teeth project strongly below hinge, the median cardinal more oblique and the posterior cardinal not so much projecting, compared with reflexa. The nymph is relatively broader than that of the Recent species. Entire surface ornamented with fine concentric threads crossed by very fine radial linear grooves. This condition is typical of very young specimens of reflexa, but the concentric threads of caudex are rather finer and the radial grooves better differentiated and further apart. Internal characters of valve as for reflexa. Height, 9.0 mm. length, 14.0 mm. Type (a single left valve) in writer's collection. Finlay (Trans. N.Z. Inst., vol. 59, p. 278, 1928) has provided Notirus for Irona Finlay. Marwick (1927, p. 634) anticipated that Irona would prove to be an offshoot from ancestors of Paphirus, and the early Pliocene form described above affords strong evidence that this is correct. Caudex is directly ancestral to the Recent reflexa. It has the same type of hinge and similar strongly projecting teeth; but the sculpture over the whole shell is akin to that of Paphirus, and is not that of adult Irona. Marwick (loc. cit.) has shown that juvenile I. reflexa of 5 mm. length can scarcely if at all be distinguished from P. largillierti of the same size, and that the sculpture of both is practically identical. This new Pliocene species shows, then, that, in early stages of divergence of Irona from Paphirus, modification of the dentition was more quickly assumed than new sculptural characters, for the Ironid form of shell and dentition was apparently evolved before the sculpture of adult ancestral Ironids began to vary much from that of Paphirus. Aloidis zelandica (Q. and G.). Suter, Man. N.Z. Moll., p. 1010, 1913. 8 valves have been obtained since the fauna was originally described. This is a new record from these beds. Specimens in collection of A.U.C. and of the writer. Genus Pholadidea Goodall. Type (by monotypy): P. loscombiana Goodall. Pholadidea finlayi n.sp. (Figs. 41, 42). Obviously a very close ally of the Recent P. tridens Gray, to which it is almost certainly directly ancestral. The most ready means of separation is to be had in the outline, for the fossil is much less elongate behind, the line of junction with the siphonoplax

is more oblique (directed postero-ventrally), and the basal margin is broadly and regularly convex throughout, not almost straight or concave behind as in tridens. The triangular area bearing the file-like rasping ornamentation is more restricted, its antero-ventral edge ascending more rapidly towards the antero-dorsal margin. The smooth, convex plates that close the front of the shell are thus relatively wider, and more conspicuous. The protoplax rises more prominently before the beaks and is not so flattened on top; its posterior lobes are not preserved. The metaplax is strongly sinuous in section, much more so than that of tridens. Hinge similar to that of tridens. Pallial sinus not entirely visible, but apparently approximating closely to that of the Recent species. Siphonoplax not preserved. Height, 13.0 mm. length, 22.0 mm. (holotype). Type in writer's collection. There are 2 sets of valves, an odd adult left valve and a very young left valve. Named in honour of Dr. H. J. Finlay, of Dunedin. Genus Eximiothracia Iredale. Type (o.d.): Thracia speciosa Angas. Eximiothracia pusillior n.sp. (Fig. 40). Shell small, oblong, very inequilateral, posteriorly truncated. Beaks within posterior fourth, directed backwards, not rising above anterior dorsal margin, which is long and very lightly convex; anterior end regularly convex; basal margin convex over anterior half, thereafter straight to faintly concave; posterior end very rapidly descending, concave just below beak, and then convex below that, meeting basal margin in a sharp convexity or sub-angulation. Sculpture of low growth-folds, seen best traversing the posterior flattened area behind the well-marked fold running from beak to postero-ventral corner. Radial sculpture absent. Hinge poorly differentiated, with a deep triangular cleft below beak, and a single strong lamellar projection rising from its edge just in front of the cleft. Valve margins smooth and sharp. Height, 3.1 mm. length, 5.5 mm. (holotype). Type in collection of Auckland University College. A paratype in writer's collection. Two right valves have been found. The species comes nearest to vegrandis (Marshall and Murdoch), a Pliocene fossil from Castlecliff, to which it is directly ancestral. The difference in outline and in the position of beaks readily separates these two forms. Myadora waitotarana Powell. Rec. Auck. Inst. Mus., vol. 1, no. 2, 1931. Powell (1931, p. 95) has stated that the shells of Myadora from Kaawa Creek, recorded by Bartrum and Powell as M. antipodum Smith, appear to be immature examples of waitotarana.

Figs 1 4— n.sp. holotype × 17 3. Figs. 2, 3, 6.—Hochstetteria pinctagrina n.sp. holotype, × 9 1. Figs. 5, 7, 88—Hochstetteria tela n.sp. holotype, × 17.3. Figs. 10, 11.—Hochstetteria kaaiva n.sp. holotype, × 17.3. Figs. 9, 12, 16.—Cosa kaawacnsis n.sp. holotype, × 17.3. Figs. 13, 14, 13.—Cratis pliocenica n.sp. holotype, × 17.3. Figs. 17.— simulator n.sp. holotype, × 9.0.

Figs. 18, 20.—Cratis miocenica n.sp. holotype, × 17.3. Figs. 19, 23.—Cuna fictilia n.sp. holotype, × 17.3 Figs. 21, 22.—Cuna crassicardo n.sp. holotype, × 17.3. Figs. 24, 27.—Cuna kaawa n.sp. holotype, × 17.3 Figs. 25, 26.—Verticipronus stirps n.sp. holotype, × 17.3. Fig. 29.—Perrieria sola n.sp. holotype, × 17.3. Fig. 28.—Arthritica elongata n.sp. holotype, × 9.0.

Figs. 32, 33.—Melliteiyx angulata n.sp. holotype. × 9.0. Figs. 34, 35, 37.—Myllitella fragilis n.sp.: holotype (Fig. 37), × 9.7. Figs. 36, 38.—Dosinula marwicki n.sp. holotype, × 1.0. Fig. 39.—Notirus caudex n.sp. holotype, × 3.6. Figs. 41, 42.—Pholadidea finlayi n.sp. holotype, × 1.4. Fig 40.—Eximiothracia pusillior n.sp. holotype, × 9 0.

List of Literature. Allan, R. S., 1933. On the System and Stage Names applied to Subdivisions of the Tertiary Strata in New Zealand, Trans. N.Z. Inst., vol. 63, pp. 81–108. Bartrum, J. A., 1919. New Fossil Mollusca, Trans. N.Z. Inst., vol. 51, pp. 96–100. Bartrum, J. A., and Powell, A. W. B., 1928. Mollusca from Kaawa Creek Beds, West Coast, South of Waikato River, Trans. N.Z. Inst., vol. 59, pp. 139–102. Finlay H. J., 1931. On Turbo postulatus Bartrun: Does it indicate a Pliocen. Connection with Australia?, Trans. N.Z. Inst., vol. 62, pp. 1–6. Henderson, J., and Grange, L. I., 1926. The Geology of the Huntly-Kawhia Subdivision, N.Z. Geol. Surv. Bull. no. 28 (n.s.), 112 pp. Marshall, P., and Murdoch, R., 1921. Tertiary Rocks near Hawera, Trans. N.Z. Inst., vol. 53, pp. 80–96. Marwick, J., 1927. The Veneridae of New Zealand, Trans. N.Z. Inst. vol. 57, pp. 567–635. ——1931. The Tertiary Mollusca of the Gisborne District, N.Z. Geol. Surv. Pal. Bull. no. 13, 177 pp. Powell, A. W. B, 1931. Waitotaran Faunules of the Wanganui System, Rec. Auck. Mus., vol. 1, no. 2, pp. 85–112.