bolivina. The aperture, too, remains incompletely developed—not symmetrical, closed and central as in Vulvulina, but depressed and open at one end where the base of the chamber was. Its development from the typical Textularian aperture of the young shell (a simple gap at the base of the chamber, longest across the thickness) does not seem to be the simple gradual lengthening in the direction of width seen in Vulvulina, but proceeds by perforation; the upper lip of the early aperture remains as a bridge behind which one or more gaps open in the base of the terminal depression; the bar or bars thus formed between openings are finally more or less absorbed.

The distinctness of shell habit has already been noted by Cushman (1932, p. 82) in discussing V. pectinata Hantken (Lower Oligocene of Hungary) and V. mexicana Nuttall (Lower Eocene of Mexico), both almost certainly members of the new subgenus. The low angle and but slight curvature of the sutures is much more characteristic of this group than of true Vulvulina.

Vulvulina (Semivulvulina) capitata Stache.

1864. Tewtilaria capitata Stache: Reise de Novara, Pal., vol. 1, p. 270, pl. 24, figs. 10a, b, c.

With this species must be synonymised two other forms described at the same time and from the same locality by Stache; T. carinata d'Orb. var. antipodum (I.c., p. 272, figs. 21a, b, c—maorica used in explanation of plate) and robusta (l.c., p. 272, figs. 22a, b, c—inflata used on plate). These are all individual forms of a not very variable species and have already been united by Chapman (1926, p. 32), though under the wrong names Spiroplecta carinata d'Orb.

The species is characterised by very sharply carinate sides, with a narrow but distinct, opaque, and frequently serrate flange; angle of expansion about 60°; more or less strongly limbate sutures; little coarse grained matter in the test, which is mostly of fine cement; and the presence of a more or less sharp medial angulation.

This is one of the most useful and reliable key species of the Oligocene. It is not always present in coeval samples, but tolerates a wide range of sediments, from glauconitie marl to limestone. It is sometimes replaced by typical Vulvulina and at a few localities the two occur together, notably in the Cobden limestone, where they are both abundant at locality 5361 (Runanga road, several hundred feet above base).

Its time range when abundant is strictly Upper Ototaran, though odd specimens may be found up to Lower Hutchinsonian; in this range it is widely spread over New Zealand. It is absent from all Waiarekan and Lower Ototaran localities, such as Lorne, basal Kaiata mudstone, and the rest of the Kaiata sequence, including the lower Point Elizabeth beds; Cormacks limestone just above diatomaceous earth; and lower Kakanui limestones (typical Lower Ototaran). It appears in the following Upper Ototaran localities: 5181 (Upper Kakanui limestone), 5359 (Upper Point Elizabeth), 5360 and 5361 (Cobden limestone; quite typical), many Whaingaroa localities, and G.S. loc. 1910,

Swinburn River (but not in the Lower Ototaran G.S. 2101 from same district). It is very uncommon in the Poverty Bay Oligocene, but is present at 1146A (Mangapapa Stream, Motu S.D., just below Ihungia). As regards the Duntroonian, it is not in the Wharekuri lower greensands of Fisherman's Rock, nor the middle greensands (though both these seem Upper Ototaran on foraminiferal evidence), but occurs at every locality in the upper greensands (Kekenodon beds); it is not uncommon and quite typical at G.S. loc. 1897 (Wendon Valley, Waikaka subdivision), a facies of the Chatton beds, though it has not been seen at Chatton. It is present at G.S. loc. 2108 (near Otorohanga, in the Te Kuiti series), which is in other respects very like Duntroonian. In the Porangahau district it is present in beds of about the same age (5388). Occasional specimens have also been found at four localities in the Poverty Bay Wheao beds (just above Oligocene); three of these are from the Whangaroa lowest dark mudstones (3143, 3144, 3017), the fourth from the Mangatu River (5394).

An interesting occurrence, together with other Upper Ototaran forms, is in the limestone at Tioriori, Chatham Is., where the only identifiable fossil apart from foraminifera is the abundant mollusc Notostrea tarda, a form simulating Gryphaea.

The line extends into the Waitakian (as the next described species) and a few Hutchinsonian localities, but is absent from every subsequent horizon. Lithologically there is a suitable facies at localities 5363 and 5364 (just above the Cobden limestone near the village), which appear to grade from the limestone and to have a Duntroonian and a Wheao fauna; typical Vulvulina continues on in abundance, but Semivulvulina disappears. There is a superficially similar form occasionally found in the Hampden (5178), Waitangi (4005), and Hurunui (5570) Bortonian, but adults from the last locality show several uniserial chambers and a different apertural development from capitata; it is in fact so close in details to V. advena Cushman (1932, p. 84, Pl. 10, f. 14), from the Jackson Upper Eocene of America, that it may be referred to that species. Also in the Hurunui Upper Bortonian (5573) is a smooth typical Vulvulina strongly resembling the Mexican Eocene V. colei Cushman (l.c. p. 84, Pl. 10, f. 21, 22). in the Awamoan are occasionally seen specimens like the early stages of Semivulvulina, but adults show that they really represent a Spiroplectammina allied to the Recent pseudocarinata Cushman.

Vulvulina (Semivulvulina) waitakia n.sp. (Plate 68, Figs, 1a, b).

Directly descended from capitata and with the same terminal chamber, aperture and initial development. The barring across the aperture seems to be more persistent, and a developed terminal chamber with closed in aperture even rarer. A particular feature is the test formation; the bulk of the shell is fine cement, but a row of coarse sand grains is incorporated just above the sutures, leaving upper two-thirds of chambers quite smooth, the lower third and the medial join roughened. This takes the place of sutural limbation and the joins of chambers are otherwise linear and not well marked. The form differs at sight in its slenderness, the angle of expansion

less sharp spine on keel at lower two-thirds jutting outwards and downwards at same angle as chambers. Sutures more distinct on later half of shell, there rather broadly and shallowly excavate; initially faintly limbate, inclined at about 20° to horizontal; medial sutures sub-linear and less marked. Sides strongly carinate, but keel interrupted by the sharp spines with broad bases, except initially, where it is sharply serrate. Aperture the normal creseentic slit, rather widely open, rounded at the sides and rather short, except in senile shells. Test typically consisting of sand grains of varying sizes, few of them large, except in terminal chambers, set in abundance of cement, the whole surface smoothly finished.

Height of holotype, 2 mm.; width, 1.7 mm.; thickness, about three-quarters width.

Holotype from Pukeuri, 4 miles North of Oamaru, Awamoan (middle Miocene). Identical forms are at Awamoa Creek (type of Awamoan), Mount Harris, Target Gully, Ardgowan, but not at other South Island Awamoan localities, such as Rifle Butts, All Day Bay, etc., where the next species is common. It also occurs at several localities in the type Tutamoe beds of Muddy and Island Creek, Poverty Bay, though much less commonly.

The full range of this species cannot be given yet, but it seems to be characteristically an Awamoan form, doubtfully occurring in the Hutchinsonian or Taranakian. These spined Textularias are an extremely common feature of the New Zealand Miocene, but after the description and figures for the above species had been prepared, it was discovered that there are two rather similar species, apparently quite distinct and of different range, though occurring together at some localities. The second species is briefly described below in order to make it available for use by local stratigraphers. It is much commoner, has a considerably longer range, and will be more fully dealt with and illustrated in a later paper.

Textularia miozea n.sp.

Shell closely resembling the previous species, but in every way coarser and stouter in build. General details of chamber formation and shape, sharp lateral keels, aperture and the distinctions between microspheric and megalospheric forms are extremely similar. The spines are less well developed; often absent, usually restricted to blunt serrations or occasional fistulose projections and only exceptionally strongly projecting. The test is regularly built of much coarser material (this is usually one of the best features of distinction) and though the cement is just as abundant the sand grains preferred are mostly larger and frequently dark coloured; the surface is sometimes smooth but more often roughened by the coarse grains. The species reaches a much larger size, though having about the same number of chambers. It is relatively wider, with a greater angle of expansion and considerably stouter at all stages, normally almost as thick as wide.

Height reaching 2.5 mm.; width, about 1.3 mm.

Holotype from Geological Survey locality 1342, Poverty Bay, Patutahi S.D., Waikura Stream, about 2 miles 60 chains E.S.E. of

concave, whereas in reussi their convexity is always more or less apparent. The microspheric form of the new species is stouter and wider and the initial expansion of its angles considerably greater, about 40° to horizontal, instead of to vertical, as in reussi. The megalospheric form differs at sight in the absence of unangled terminal chambers. The aperture is on the whole smaller, though the shell is larger. The suture lines between chambers are usually less distinct and excavated. After this species was distinguished, much better specimens were found at Hampden, and the holotype chosen will be figured later.

Height, up to 3.5 mm.; width, up to 2.5 mm. (microspheric form). Megalospherie form about three-quarters this size or smaller.

Holotype from Hampden, 1.½ miles North of Kakaho Creek, upper-blue clays, about 5 ft. below top. Collected by J. Park. This isnear the top of the Bortonian.

This is one of the most useful species for distinguishing the Bortonian from the overlying beds and Oligocene. It is an abundant and characteristic form throughout the Upper Hampden section from the top of the Kakahoian marls (including the “Marly Clay” with Zeauvigerina) down to and including the greensands, i.e., the whole Bortonian. It is replaced by quite definite reussi in the very topmost Hampden sample from a locality at the extreme north of the section and exposed only at very low tide; for this and other reasons the bed is regarded as Tahuian. It is present in the Bortonian of Waihao Downs and in the McCullogh's Bridge lower greensands below the phosphatic band, but is clearly replaced by reussi in the Upper morefossiliferous greensands at the latter locality—the type Tahuian. This is one of the few Foraminiferal distinctions available for separating Bortonian and Tahuian, but it appears constant and valid. The foraminiferal fauna of the Burnside marl which has been variously classed as Eocene or Oligocene is definitely referable to the Tahuian on this and other counts. From this horizon G. reussi extends upwards commonly through the whole Ototaran, but is not known to me-from any Duntroonian or Waitakian facies. Towards the end of its range its presence is sometimes useful in distinguishing an Upper Ototaran from a Duntroonian fauna. The ancestral G. proreussi indicates by its presence the Bortonian age of such other beds as the Pahi greensands, the Waitangi marls of Poverty Bay, the chalk marls below the Amuri limestone and the Asterocyclina bed of White's Creek, North Canterbury, which is the same horizon as the Eyre River type locality of Discocyclina speighti Chapman (1932, page 483); these orbitoids were classed by Chapman as middle Eocene, which, from other viewpoints, also seems to be the best age for the Bortonian.

As regards a lower limit, the bed of greensands which forms a middle stage of the Moeraki-Hampden section is extremely poor in fossils, but has yielded a few specimens of proreussi and some calcareous species of the upper marls, so that this bed should be included in the Kakahoian Bortonian. The Moeraki glauconitic marls below this (including the “Moeraki boulders”) have a quite distinct series of faunules, consisting of only arenaceous species, but not including

false appearance of striation, on the later whorls an alternating effect—a pitted suprasutural area followed by an upper oblique translucent band. Sutures linear, fairly well marked, chambers slightly convex and about two-thirds as high as wide.

Height, up to 0.9 mm.; width, 0.12 mm.

Holotype from loc. 5371, type Wanstead (Motuotaria S.D., main road 1 mile south of Wanstead Hotel).

This is an important and abundant index species of the Wanstead and Bortonian. It is almost invariably present in the many Wanstead samples seen from the Dannevirke area, and occurs also at several places in the Eocene Waitangi marls of Poverty Bay and in the Pahi marls. At all these localities it is associated with the genus described elsewhere as Zeauvigerina. It is present, though somewhat rarely, throughout the Kakaho marls at Hampden, thus covering the whole Bortonian. It is not known outside of this middle Eocenerange, being conspicuously absent in the abundant Tahuian fauna from the Burnside marl.

The new species is evidently closely allied to Bolivina applinae Plummer, which has been referred to Loxostomum by Nuttall (1930, p. 285, pl. 24, fig. 45), but which evidently is better placed in Rectobolivina. Nuttall notes that applinae has fine longitudinal surface striae and crenulate indentations along the base of the chambers, the biserial portion being over half the shell—these details allow of immediate separation from bortonica. Another ally appears to be the Lower Oligocene Bifarina vicksburgensis Cushman (1929, p. 45) from the Byram marl; it has the somewhat staggered chambers, but the early periphery is serrate and the coarse perforations are differently distributed.

The use of the generic name raises the question of related genera. The present position of species variously referred to Sagrina Loxostomum, Bifarina, Rectobolivina, Siphogenerina and Tubulogenerina seems to me in a state of some confusion, and often anomalous. The groupings adopted illustrate well certain defects of the system of genus-making and systematic location by means of theoretical definition, as opposed to lineage grouping round the type species. The former method holds complete sway at present in the Foraminifera but has not been very satisfactory in other groups and, while apparently simple, leads to a loose and contradictory use of genera, as seen in many families of the Foraminifera. It is absolutely contrary to the International Rules to regard a genus as a wordy concept. Article 25 makes it clear that a genus name is invalid unless “the author has applied the principle of binary nomenclature,” and a genus without species does not conform to the principles of binary nomenclature. Genera conceived only by definition cannot have any but a personal claim to acceptance; opinion 2 unequivocally states that “names based upon hypothetical forms, have no status in nomenclature.” It seems to me that some concepts. in the treatment of Foraminifera need modification if these ideas are logically applied.

species with small retral processes. The diverse type of shell formation and sculpture makes Polystomellina an unhappy location for the Austral group, which seems to be associated more with the heavily and coarsely sculptured species of Elphidium.

Attention should be drawn to the aperture of the new genus. All of the descriptions and most of the figures quoted give this incorrectly. It has been variously described as “The normal arched fissure at the inner margin” (Brady), “An arched opening at the base” (Chapman, Parr and Collins), “A low opening at the peripheral margin, extending slightly to the ventral side” (Cushman). These are all misleading; I have never seen a true apertural opening in this group. A slight hollowing out may occur along contact with base and last chamber, but it does not continue inwards; the whole apertural face is a solid wall overrun by the heavy sculpture-ridges. The meeting of these with the base has somewhat the effect of producing cribrations, but in perfectly preserved shells there is no trace of a real opening. It is possibly unnecessary, as the whole shell is so porous and the terminal face so extremely thin and perforate. This complete absence of true aperture is regarded as highly diagnostic, and of course at once prevents any connection with Rotalia. It is extremely easy to get a wrong impression of this aperture if the shell is in the slightest degree broken or worn. A number of the strongly spirally sculptured species of Elphidium show the same aperture, while others have a row of pores in the terminal face.

Such a species as Elphidium imperatrix Brady (1884; pl. 110, figs. 13–15) must be very closely connected with the Notorotalia line; it shows the same development of peripheral spines as the Middle Tertiary New Zealand forms, has the same aperture and the same type of strongly ornamented proloculum. Apart from its perfect symmetry, it is in fact a Notorotalia. Another prominently spinous Indo-Pacific form is Rotalia trispinosa Thalmann (pulchella d'Orb., auct.), but the aperture and development show plainly that this is unrelated to Notorotalia, and is a true Rotalia of the beccarii type.

Notorotalia zelandica n.sp.

This species is best described by comparison with Rotalia clatthrata Brady. That species was described from “Pour Stations between Australia and New Zealand” and “Two Stations on the West Coast of Patagonia.” Brady noted that the latter were different, as is shown by his figure 9 (from Station 305, West Coast of Patagonia). His figure 8 was from Station 162, 38–40 f., off East Moncoeur Island, Bass Strait, and this must be taken as the typical form; if it has not been done before I now choose as lectotype of Brady's species the Australian specimen which was the original of his figures 8a, b, c.

Forms of this kind are extremely abundant in shallow water all around New Zealand at the present day, and a long ancestral line is known. The common Recent New Zealand form does not appear to be the same, however, as the Australian. I have not seen Recent samples from there; it is not present in my samples from 7 fathoms Derwent Estuary, Tasmania, and it is not in any of the large Peronian or Solanderian faunas I have available. It seems to be

more at home in a colder climate, though Heron-Allen and Earland have noted that it dies out as the sub-Antarctic is approached. Specimens I have from the Pliocene of the Abbatoirs Bore, Adelaide, (recently referred to the Lower Miocene howchini by Howchin and Parr, but not the same as any Muddy Creek lower bed species I have seen) agree in general with Brady's figure and, using this as a basis, one notes many differences in the New Zealand Recent form. The latter grows much larger (up to 1.4 mm.), while Brady's type was only 0–8 mm. The New Zealand shell has a relatively blunt periphery, the medial keel being masked by nearly similar ridges above and below it, and the sides not shelving steeply to meet it as in clathrata. The Australian shell is mostly domed above and fairly flattish below, but zelandica is nearly always flatly depressed above and more convex basally, especially just inside periphery. The central part of the base is notably depressed or excavated, with an umbilical pitted area as in Elphidium; clathrata has a more evenly flattened base and smaller pitted area. The sculpture of clathrata between the limbate sutures consists of relatively few and distant irregular ridges, the spirals on the ventral side unwinding slowly; zelandica has much heavier and closer intra-sutural ridging, the ridges on the upper two-thirds of each chamber pointing steeply outwards at about 80° to spiral suture, but on the lower two-thirds of chamber suddenly altering slope to about 45°, evenly covering whole surface between radial sutures. On the basal surface the spirals unwind much more rapidly and are frequently discontinuous at sharp angles from chamber to chamber. Microspheric form regularly spiral; megalospheric form with a large proloculum minutely papillate all over.

Diameter of adults, 1 to 1.4 mm. Height somewhat variable, about half diameter in adults and somewhat less in young.

Holotype from upper part of Castlecliff beds, Uppermost Pliocene; chosen in preference to a Recent specimen, as the latter are mostly worn and therefore show weaker sculpture. Figures of this and aperture details will appear later.

This Recent and Upper Pliocene species is characterised chiefly by its large size, rounded or bluntly sub-angled periphery, with a greater convexity below it than above. It occurs also in the underlying Nukumaruan (mid-Pliocene) but is not known as yet from the Waitotaran or Opoiti stages (Lower Pliocene).

Numerous Tertiary species are known from all horizons back to the Bortonian, where very rare and tiny forms may indicate the beginning of the lineage, but have all essential characters, such as aperture, perfectly developed.

Notorotalia stachei n.sp. (Plate 69, Figs. 3a, b).

Shell small, depressed biconvex, but often flattened above or below. About 2.½ whorls, with eight to nine chambers in last one, distinct ventrally, obscure dorsally. Whole dorsal surface covered with short, strong, blunt ridges, masking sutures in most cases; where clearly developed they are seen to consist of 5–6 subparallel cords with subequal interstices, as strong as or stronger than the limbate radial sutures and completely crossing the chambers between

The discovery of abundant specimens in the New Zealand Eocene of a species closely related to the Mexican Eocene Eponides trümpyi is very interesting. The Mexican form was recorded by Nuttall as frequent in the Chapapote (upper) and Aragon (lower) members of the Eocene, but not outside this age. In New Zealand, specimens are common in the Upper Bortonian and apparently less so in the Lower Bortonian. It has not occurred in the Tahuian (Uppermost Eocene) but a closely related form is present in the lower and middle Rakauroa (Upper Cretaceous). The New Zealand Eocene form is best described by comparison with trümpyi as follows.

Nuttallides subtrumpyi n.sp.

Shell of entirely the same build and general features as trümpyi, with similar dorsal and ventral sutures. The flange is much sharper; Nuttall described and figured “a rounded peripheral flange,” but New Zealand specimens have a fairly acute, bevelled, translucent peripheral border. They are distinctly planoconvex, the dorsal side being mostly flat or slightly raised medially, while trümpyi is apparently biconvex. This feature varies somewhat, but the New Zealand shell as a whole presents a recognisably different shape from Nuttall's figures. An immediate distinction is the character of the ventral boss; this, in trümpyi, is mentioned and figured as rounded, and does not disturb the basal outline; in subtrümpyi it is almost always quite flat over its whole area, the basal outline being thus sharply truncated. The dorsal surface and to some extent the ventral has a heavy layer of translucent shell material evenly spread all over, through which the sutures are visible, the central ones indistinctly where the layer becomes thickest. This feature is not mentioned by Nuttall, who figures the central portion as clearly as the rim.

Diameter, up to 1 mm. (trümpyi averages 0.7 mm.); width, two-thirds to three-quarters diameter.

The figures prepared for this description did not show the essential details clearly, and illustrations will be given in a later paper.

Type from locality 3310, greasy marl, Matakohe Survey District, Auckland, 1 mile north-west of Pahi, Paparoa Arm. This is interbedded with greensands of the Pahi series, contains also Zeauvigerina and Hantkenina and is definitely Upper Bortonian. Elsewhere the species extends throughout the Bortonian (but not above), with a few rare specimens, possbily a distinct species, from the uppermost Cretaceous.

The reference of this type of shell to Eponides can hardly be upheld. Nautilus repandus, the genotype of Eponides, is a common Indo-Pacific shell, and its lineage is continued well back in our Tertiary. Fairly typical specimens extend down to the Lower Miocene, but in the Lower (and perhaps Upper) Ototaran a distinct ancestor occurs. This differs in having a less-pointed base, a distinctly open umbilical area formed by the great spread of the aperture ventrally, and especially in having fewer chambers; regularly five to a whorl, instead of seven to ten as in repandus. This form is especially common in the Waiarekan tuffs of Lorne (G.S. locality 1100) and

may take the name of Eponides lornensis, new species. Chapman, Parr and Collins (1934, p. 565, pl. 9, figs. 18a–c) have recorded and figured repandus from the Australian Tertiary as far down as the Oligocene Balcombian; as they say, Muddy Creek examples seem inseparable from typical Recent shells, and the same applies to N.Z. examples from the Miocene onwards. Waitematan specimens occasionally look like lornensis, but adults always have six or more chambers.

All these true representatives of Eponides show the typical generic aperture, a widely open space entirely confined to the straight contact between the last whorl and base; its outer margin notably convex and with a wider gape ventrally before the umbilical union of the sutures is reached. In the ancestral lornensis it spreads into this area; in all forms there is a definite end at the periphery with no tendency to invade that area at a new angle. The apertural face in typical Eponides is usually more or less cribrate. Other species of Eponides, such as the American Tertiary guayabalensis Cole and the European Miocene to Recent schreibersii d'Orb., keep this type of aperture with not a great transition in shell features.

This is totally unlike the apertural features of trümpyi and its allies, with the angled lateral extension into the peripheral face, and as several diverse species are known with this type of aperture it is here made the principle basis for generic segregation, though general habit of shell and type of chambers are not without significance also.

I suggest as a congeneric species Discorbina alata Marsson, which has been figured and described as a Pulvinulinella by Cushman (1931, p. 311, pl. 36, figs. 5a–c). Specimens extremely similar to those there figured from the Saratoga chalk occur at locality 3250A, Tuparoa Stream, Waipiro S.D., Poverty Bay, in marls of mid-Rakauroa age. The aperture in Cushman's figures is not altogether distinct, but the New Zealand specimens show exactly the Nuttallides aperture. The basal knob of clear callus is prominent only in juveniles, adult examples tending to surround this below with a basal ridge, forming a pseudo-umbilicus. The later figures of this species given by Cushman and Jarvis (1932, p. 48, pl. 15, figs. 1, 2) are less like the Tuparoa and Saratoga specimens and more resemble in shape another species occurring in the New Zealand Rakauroa.

The devolpment of the basal characteristics in alata lead me to suggest that the characteristic Upper Cretaceous species recorded by authors as Gyroidina or Globorotalia micheliniana d'Orb. also belongs to Nuttalides. Cushman's figures of Annona chalk and Antigua specimens show clearly the sinuous suture and flange development and the characteristic aperture. Also the pseudoumbilicus, as in alata, is quite different from the pervious one of Gyroidina. Two other Cretaceous species certainly to be grouped with micheliniana are Gyroidina alabamensis Sandidge (Alabama Ripley) and Globorotalia subconica Morrow (Kansas).

Another genus which must necessarily be brought into the discussion is Pulvinulinella Cushman. The genotype of this is P. subperuviana Cushman, the description and figures of which show an

elongate aperture just below the periphery and entirely in the plane of coiling, without prolongation along the basal margin towards umbilicus. The aperture was originally described as “somewhat loop shaped,” with the remark that the type species “does not show the apertural characters as definitely as do the Recent species.” This was evidently overlooked by Chapman, Parr and Collins (1934, p. 569, pl. 9, figs. 19a–c) when describing the Australian Oligocene tenuimarginata as doubtfully belonging to this genus. In shell features it is quite like the genotype except for the straight aperture, which, however, is seen in Recent species such as pacifica Cushman. The Balcombian form, which can be taken as gencrically typical, is abundant also in New Zealand, where is has an extended range from at least Lower Oligocene to Upper Miocene.

A series of such typical species shows that the Pulvinulinella aperture is well defined and not really much like that of Nuttallides, where the lateral extension, though definite, is merely an adjunct to the main aperture, and it is stretching the limits of the genus too much to include alata and similar types. The recently described P. gyroidinaformis Cushman and Gaudkoff (1938, p. 2, pl 1, figs. 1, 2) has some unusual shell features but apparently a fairly typical aperture. Another type of aperture, however, which is decidedly atypical is that shown by Planorbulina culter Parker and Jones. This has been very generally referred to Pulvinulinella but some writers have questioned this, and Nuttall (1930, p. 293) has stated that the appertures are very different. The good figures given by Brady (Chall., pl. 96, fig. 3c) and Schwager (1866, pl. 7, fig. 111; as bengalensis) clearly show that the striking feature of this aperture is an oblique slit in the end face, well below the periphery, opening out from the contact slit between base and last chamber, and descending well into the face at an angle about midway between basal and peripheral margins. A number of similar forms are known, including mexicana Cole, velascoensis Cushman and probably the lately described texana Cushman (1938, p. 49), which, though compared with alata, seems to have more the culter type of aperture and shell coiling. This kind of shell is common practically throughout the Tertiary succession in New Zealand, and the lineage extends into the Upper Cretaceous. The apertural and shell distinctions from Pulvinulinella and Nuttallides are here emphasised by the proposal of the new genus Parrella, with Anomalina begalensis Schwager as genotype. This form is chosen for the type instead of Parker and Jones' culter on account of the ambiguity associated with the latter. Chapman and Parr (1937, p. 119) have noted that the true culter differs from Brady's interpretation, which probably was equivalent to Schwager's species; since my remarks on apertures have been based on shells like Brady's and Schwager's figures, it is plain that bengalensis must typify the genus. These remarks were written before Chapman and Parr's were seen; their independent suggestion that a new genus is probably needed is much wiser than their unsatisfactory location in Cibicides.

The distinction of the calcar group from typical Rotalia has been recognised by both Cushman and Galloway, the former placing it in the separate family Calcarinidae, while the latter used the older name Tinoporus as family basis. But in each case the sundering was made on the complex shell features exhibited by the more specialised genera, and little heed was paid to the aperture. Cushman wrote “The family in its simplest forms is close to Rotalia, in fact Calcarina calcar may be a Rotalia,” while Galloway stated that “if the family consisted only of the genera Calcarina and Tinoporus it could well be left in the Rotaliidae.” To the rather slight distinctions Galloway gives between Calcarina and Rotalia I would now add as a primary feature the distinctive aperture, and though I do not accept the complete groupings of either author, the development of the more complex genera from the simpler calcar line seems clear, and separate family rank justified. Cushman includes more genera than Galloway; whatever may be the objections to some of them, there can be little doubt of the affinities of the Cretaceous Siderolites Lamarck, 1801—the oldest generic name in the family. It is the practice of many outstanding workers on mollusca and other groups to take the earliest genus name as automatically fixing that of the family, but this position has not been taken in Foraminifera, and in view of the numerous alterations it would cause in classification, I hesitate to adopt it. Yet I personally feel that this is the safest and most logical method, for obvious reasons, and that a family Siderolitidae in the present case will some day be used. The question of authorship of family names is immaterial and leads to much useless personal controversy.

It is to be noted that Cushman and Galloway have used the name Calcarina in different senses, and the validity of its interpretation must be considered. Cushman's interpretation is based on his Philippine Report (1919, pp. 363–368, and 1921, p. 351), where he considered that “The type species of Calcarina is Nautilus spengleri Gmelin. This species is definitely taken as the genotype by d'Orbigny”; at the same time (p. 354) he recognised and figured Montfort's baculatus as a valid species, though suppressing his generic name Tinoporus on the ground that his diagnosis did not agree entirely with his figure and evidently included specimens of Baculogypsina sphaerulata. Galloway (1933, p. 312) disagreed with this, regarding Tinoporus as a valid name, based on the figure, and giving the genotype of Calcarina as calcar “by absolute tautonymy and designated by Parker, Jones and Brady, 1865.” As regards the former contention, it seems obvious that since baculatus is recognisable, Tinoporus, whatever the diagnosis, was introduced with one valid species, and only one, and is therefore a monotypic genus; as it represents the earliest name for the spengleri-defrancii-mayori group it must be used. As for the second contention, Cushman later (1933, p. 312) pointed out that the complete statement of Parker and Jones was not free from ambiguity, and continued to regard spengleri as the type. But d'Orbigny did not definitely designate types, and the question of tautonymy was passed over by Cushman. Article 30 of the International Rules, however, gives a number of cases gov-

erning type designation, to be applied in definite order of precedence—cases 1a, b, and c do not apply in the present instance, but 1d states that “If a genus, without originally designated or indicated type, contains amongst its original species one possessing the generic name as its specific name or subspecific name, either as valid name or synonym, that species or subspecies becomes ipso facto type of the genus.” This adequately covers the case of Calcarina calcar d'Orb., 1826, whatever other species he included, and Calcarina is thus the valid name for the calcar-venusta group, as distinct from Tinoporus for the heavily spiked, secondarily thickened species.

The discussion of this family and generic name is necessary here, since Calcarina has been credited with a New Zealand Tertiary species, which is quite unrelated, while an important index species which truly belongs here has been obscured by a wrong location elsewhere.

Tinoporus hispidus (Brady).

This, as Calcarina hispida, has been reported from the New Zealand Oligocene by Chapman (1926, p. 87; pl. 17, fig. 9) as an initial fossil record of the species. All reference to the record and to the species in New Zealand should be expunged; nothing remotely like this Recent tropical shallow water species occurs in this country, Recent or fossil. I have examined Chapman's type and slides and much topotype material from Flat Top Hill, and what Chapman had is quite evidently a species of Notorotalia, very abundant here. Similar shells covered with minute spines have been seen only at Landon Creek, Everett's Quarry, and the uppermost limestone at Kakanui Point—all Upper Ototaran horizons. It is difficult to see the details exactly on account of bad preservation, the deposits having been subjected to re-crystallisation by percolating water, but it is almost certain that the “spines” are secondary calcite crystals and have nothing to do with the original shell structure. Similar spines appear on parts of other species in these beds, though this genus seems peculiarly liable to develop them; specimens in the Bridge Point green tuffs are sometimes finely spinose, sometimes partly and irregularly coated, sometimes smooth. I have already (in Allan, 1938, p. 90) drawn attention to this species, which will be described later from better preserved material; apart from the pseudo-spines, the Duntroonian form is very different, and Dr. Allan's argument is not affected. Tinoporus, however, can be completely dispensed with in New Zealand.

While on this subject it may be mentioned also that all reference to and records of “Rotalia papillosa var. compressiuscula Brady” must be deleted from the New Zealand area. There is nothing in this country recalling this obvious relative of schroeteriana, a true Rotalia, which genus is represented in New Zealand only by a form of beccarii. Heron-Allen and Earland recorded the species from “a few fossil specimens” in the “Terra Nova” Report, and Chapman (1926, p. 86) has identified it from the Weka Creek Grey Marls (Basal Miocene). Since on his type slides the same form is variously identified as Pulvinulina elegans and P. karsteni, and mixed

indiscriminately with at least two other genera, not much weight can be allowed to this record. His specimens are once again a species of Notorotalia, close to his own spinosa.

Calcarina mackayi (Karrer).

1864. Rosalina mackayi Karrer; Novara Exp., Pal., vol. 1, p. 82, pl. 16, fig. 14.

Described from the Lower Miocene Waitemata beds of Orakei Bay, this species has been misunderstood ever since. Chapman (1926, p. 85) referred it to Rotalia, “as closely related to Rotalia beccarii,” but there is not even family resemblance. Karrer's figure is somewhat idealised, but good specimens for a new figure are hard to find as the form is very restricted. The aperture, not shown by him, is distinctive. It is of the same general style as already described for Calcarina, but differs in some details; the slit along contact of base and last whorl is absent or filled in and the large slit in the terminal face is considerably lower down than in calcar. This gives the appearance of an isolated opening slightly removed from and at a small angle to the contact margin, outlined by a strong rim all round. This development seems not a far step from the typical, and probably well-preserved juveniles would show gradation; indeed there is some evidence of it in Australian specimens.

C. mackayi is strictly limited to that part of the Hutchinsonian represented by the Waitemata beds, and is very rare in New Zealand. It is a large and characteristic species, but has so far been seen from only four other localities. At the first, 3303 (Whakau Stream, Tutamoe S.D., Poverty Bay), it is not uncommon (though badly preserved) in company with enormous numbers of orbitoids, mostly Nephrolepidina. This is the only locality in New Zealand where free orbitoids are abundant; it is an isolated outcrop, but from other evidence probably marks the base of the Ihungia, and can safely be correlated with the Waitemata series. This is in keeping with Marwick's record of the Waitematan index molluscan genus Paracominia from the Igneous Conglomerates of the basal Ihungian. The two other localities are 5277 (“Blue Bottom” sandy marl, 6 miles S. of Greymouth, 4 miles E. from coast), and 5054 (Trelissick Basin, tuff between limestones, junction of Porter and Thomas Rivers). Both these represent the same Hutchinsonian horizon as the two previous; from similar tuffs at Whitewater Creek, not far from 5054, Thomson has recorded the index Hutchinsonian brachiopod Pachymagas parki. The final locality is from basal Mahoenui, 4 miles N. of Pio Pio (5579), where the species occurs with abundant Semivulvulina waitakia Finlay.

This Calcarina apparently occurs also in Australia; specimens are rather common at Torquay (Lower Miocene) and seem to differ little from the New Zealand ones. The characteristic aperture is quite the same, as is also the case in abundant specimens from the Lower Miocene of Filter Quarry, Batesford, though the latter are highly nodulose and evidently a distinct species. This discounts the references by Heron-Allen and Earland (1924, p. 181) and Chapman (1909) of this form to Rotalia calcar. The aperture and shell ornament are distinguishable at sight and the Batesford species has

by Finlay and Marwick (1937, p. 15). Waiarekan I would eliminate entirely as merely a basal facies of the lower Ototaran; if used at all it should be only in a zonal sense. The new term Duntroonian has been proposed by Allan (1938, p. 89) for a stage directly above the Ototaran, and directly below the Waitakian, which name was proposed long ago by Park for the Waitaki limestone, known now to underly the Hutchinsonian. As pointed out in this paper, though the distinction between Duntroonian and Upper Ototaran is valid as Allan set forth, the separation from Waitakian seems impractical elsewhere in the absence of good molluscan and brachiopod faunas, and the differences may again be merely zonal or even facies. A term is possibly needed for the Lower Hutchinsonian, characterised by less evolved Pachymagas of the hectori-haasti type, leaving the true Hutchinsonian for the faunal horizon of P. parki, Rhizothyris rhizoida, and the Waiparia-Neothyris association. At least one new stage, though of small extent, seems to be needed between the Hutchinsonian and Awamoan, as shown by the Clifden and Poverty Bay sequence; the higher Clifden and Mt. Brown beds with Stethothyris sufflata and Neothyris novara being very probably an Awamoan facies—they are not recognisable elsewhere and the mollusca are close. The Opoiti series has been recognised by Ongley (1928, p. 7) as a necessary unit between the Taranakian and the Waitotaran; it is well marked in the North Island and from the Foraminifera belongs in the Wanganuian System. In order to have a definite stage name available I here propose the Opoitian Stage for the interval of time represented by the depositions of the beds quoted by Ongley as between the Mapiri series and the Waitotaran (represented by his Waihua beds, the Ormond and Tokomaru being of uncertain and varying age and better dropped altogether), together with such periods as may be represented therein by non-deposition or erosion. Faunally, I define this stage as characterised by abundance of Rectobolivina aff. bifrons, the first appearance of Patellinella, Bigenerina aff. nodosaria, Globorotalia inflata, Siphotextularia wairoana, Sigmoilina n.sp., and Plectofrondicularia n.sp., and the last appearance of Haeuslerella and many long-ranging Miocene lineages.

In allotting all these stage names to the broad European divisions, molluscan and brachiopod workers have been hampered by the frequent poverty or absence of faunas from critical horizons. I have now seen large foraminiferal faunas from every stage mentioned, with the exception of Wangaloan, and my conclusions have been reached solely from their study, though I have naturally utilised my experience with the two other groups. I have taken two outside levels and one fundamental hypothesis as guides in drawing up the table here suggested, and its validity depends on theirs. The levels are:—Middle Eocene age for the Bortonian, because of the occurrence of Discocyclina and Asterocyclina, and Lower Miocene age for the Hutchinsonian, because it is almost without exception the horizon of Nephrolepidina spp. and Miogypsina, and correlates with the Australian Janjukian, already referred to that age. The hypothesis is that the major observable faunal changes will most likely be at the dividing planes between Eocene, Oligocene, Miocene and Pliocene; there is definite evidence elsewhere that this is not so, and therefore

Figs. 1a, b—Vulvulina (Semivulvulina) waitakia n.sp.; 1a co-type, shows aperture bars too regular, 1b holotype, X 45.
Figs. 2a, b—Siphotextularut wairoana n.gen, n.sp.; 2a co-type, 2b holotype, X 90.
Figs. 3a, b—Plectofrondicularia pohana n.sp.; holotype, X 60.
Figs. 4a, b—Plectofrondicularia parri n.sp.; Pukeuri paratype, X 60.
Fig. 5—Rectobolivina parvula n.sp.; holotype, X 120.
Fig. 6—Rectobolivina bortonica n.sp.; holotype, X 60.

To face page 532.