question. Since few detailed accounts of its occurrence have appeared, and as its index value is of considerable importance to local and even outside workers, a complete list of its known New Zealand occurrences is here presented. The variety lata Cushman and Jarvis (1928, p. 93; pl. 13, figs, 11a, b) described from the Upper Cretaceous of Trinidad has not been separable so far in New Zealand. It was stated to be more rounded and much larger than the typical Velasco form, and New Zealand specimens can certainly be matched with the description and figure, but in these features such enormous variation is obvious at any locality yielding a large enough suite that no good purpose is seen here at present in separating any variation from epigona. New Zealand specimens exactly agree with Rzehak's original figure. Variation does not seem to have any stratigraphic use and occurs at all horizons; two forms are found at most localities, a thick one and a smaller angular and very thin one (microspheric?) with more coils. No specimens in New Zealand, however, approach the thickness of Silicosigmoilina Cushman and Church (1929, p. 502, pl. 36, figs. 10, 11, 12), the extremely similar genus which is common in the Californian Upper Cretaceous Chico, and “occupies the same place in the fauna that Rzehakina does … in Trinidad.”

The earliest records are from the “Puketoro” formation, which is probably equivalent to the Raukumara of the Survey. Mudstone at the top of the “Puketoro” (probably equivalent to the Mangataone) has yielded a few faunas—the only ones in New Zealand with true Globigerina cretacea—; two of these (3201 and 3203, Puketoro Stream) yielded one specimen of Rzehakina, while another (3180, Waitahaia River) did not. None were found in the red beds at the top of the Puketoro (3204), nor in Raukumara beds with Inoceramus in the Dannevirke area (5293 and 5302), but one specimen was again present in mudstone with Inoceramus (5109A, 15 miles south of Cape Kidnappers). In the Raukumara, Rzehakina is thus so rare in records and specimens that it is probably just becoming established in this faunal province.

The next formation, in contrast, provides abundant records and examples. The formational name Rakauroa was used in Poverty Bay for a sequence of beds between Puketoro and undoubted Tertiary; the lower part of this was noted in Hikurangi and Mata survey districts as extensive conglomerates and other coarse sediments—this is quite possibly equivalent to the Tapuwaeroa of the Survey Rzehakina is here abundant, with frequent large, stout specimens of the lata form (e.g., localities 3242 Mata S.D.; 3249 Waipiro S.D.; 3270 Mangaoporo S.D.; also in associated black shales). It is again abundant in coeval deposits from the Dannevirke area (5301, 5321, 5336 and many others), but it is not present in what is evidently a fauna of the same age from the South Island East Grey District (5329, mudstone below Amuri limestone).

The upper, more typical, Rakauroa consists of finer grained, lighter coloured sediments, which are probably in part at least equivalent to the Survey's Whangai. In the Dannevirke area the argillites and chocolate shales of the Whangai have yielded no Rzehakina, but

odd specimens have been found in Poverty Bay samples lithologically referred to the Rakauroa, and it was abundant at locality 3250A (Waipiro S.D., light mottled marls, Tuparoa Stream; above lower Rakauroa conglomerates).

The succeeding formations are Tertiary. The lowest of these is the Te Hua series in Poverty Bay, equivalent to the Survey's Wanstead in Dannevirke. No true Wanstead or younger fauna has ever yielded Rzehakina, nor have South Island beds of equal age. It is missing in the whole Hampden section from the Cretaceous Katiki and Moeraki series through to the Bortonian (equals Wanstead) Kakaho marls. The red shales, however, provide the biggest problem, especially with regard to the possible Tertiary occurrence of this genus. No other beds in this series (chalky limestone, Te Hua marls, etc.) contain it, nor is it present in the whole series of cores from Waitangi No. 1 well, which seemingly pass through all the varieties of red and green shales and bentonitic earths met with on the surface; the faunas here are all typical Bortonian. But at two surface outcrops of red shales in the Waitangi fault zone (4012 and 1006) faunas were found which somewhat resemble the well cores, except in lacking calcareous species and in possessing abundant Rzehakina and Bolivinopsis spectabilis (Grzyb.). The latter species almost invariably accompanies Rzehakina and has much the same range, but seems to be more widespread. I can see no appreciable difference between this species, as figured from Burdwood Bank by Macfadyen (1933, p. 7, Fig. 1c, d, i, j), and the Trinidad shells identified as the later clotho (Grzyb) by Cushman and Jarvis (1932, p. 43, Pl. 13, Figs 5, 6), though Cushman (1934, p. 42, 43) left them separate. Both are very large and have the same accompanying fauna, and the New Zealand forms are certainly the same as the Burdwood and Trinidad ones. The other elements in the Waitangi red shales could easily be of Rakauroa age, so that either these and Rzehakina range into the Tertiary if the outcrops are correlated with the definitely Bortonian well cores, or else there is more than one “red shale.”

The variability of Bolivinopsis in New Zealand is so great that the forms are difficult to deal with. In the chocolate shales at the very top of the Rakauroa (above the White Argillites) and in the Moeraki beds below the Hampden Bortonian there are irregular and distorted specimens with slightly shorter chambers, apparently exactly the same as the Californian Upper Cretaceous B. californica Cush. and Campbell (Contrib. Cush. Lab., vol. 10, pt. 3, p. 70). But some of the New Zealand shells also recall “Spiroplectammina” mexiaensis Lalicker (l.c., vol. 11, pt. 2, p. 43), which is surely congeneric. And the separation from megalospheric spectabilis, even of B. eocenica Cushman and Barksdale, lately recorded by Parr from the West Australian Eocene, does not seem at all sure. A form quite like the Moeraki specimens, and also spectabilis, has been figured by Nuttall (Journ. Pal., vol. 9, pt. 2, p. 125) from the Eocene of Venezuela.

Since this was written I had an opportunity to collect samples from two further bentonitic surface outcrops; 5392 (grey shale directly below basal Wheao conglomerate, Mangatu River) and 5391

(red shale below road level, Armstrong road, Waitangi district). Besides the usual facies background of such forms as Haplophragmoides coronata, Rhabdammina and Hyperammina fragments, Glomospira corona, Cyclammina, Ammodiscus, etc., Bolivinopsis spectabilis and Rzehakina occur in both, though the latter is rare in 5392; an important addition in both is a number of calcareous species, which are too indefinite to refer positively at present to Cretaceous or Eocene, but which are not the well-core species, and show that the faunal differences are not just due to leaching out of the limy forms. One surface outcrop, however, has been seen (1015; Branch D of Red Shale Stream, Waitangi) which contained a Bortonian fauna as in the well cores; Eocene species of Globorotalia, Pulvinulinella and Rotaliatina were present, but no Rzehakina or Bolivinopsis. In the Puketoro Stream section is 1000 ft. of marl and red shale (3209) with a fauna poor in extent but undoubtedly equivalent to those of the well cores, and like them possessing B. spectabilis, but lacking Rzehakina; it lies definitely above 800 ft. of Rakauroa argillite containing Rzehakina (3206). The existence of two red shale horizons thus seems very probable.

Two other localities yielded Tertiary faunas accompanied by Rzehakina. At Pouawa Dome (3286) a fauna was obtained from below another fauna which was Oligocene. The Oligocene fauna (3287) was slightly contaminated with specimens from the older bed; the latter had a predominately Wanstead fauna, but included three or four species such as Rzehakina, B. spectabilis, etc., which differed in preservation and were almost certainly derived from a still lower horizon in this complicated fault structure. The other locality (4019) was at Te Hua Stream, Waitangi, where the Wanstead fauna also showed Rzehakina and B. spectabilis but, as red shales were closely associated, it seems likely that again derivation occurred. Three further localities, 3284A (Pouawa Dome), 3290 (Whangara uplift), and 4013 (Waitangi uplift) yielded Rzehakina with calcareous species which are possibly Bortonian, but may be Cretaceous; they are all in disturbed areas, and are unlikely to yield pure faunas.

Apart then from the red shale question, it is evident that Rzehakina in New Zealand, especially when abundant, denotes Rakauroa age (Whangai plus Tapuwaeroa?), that it extends feebly to the underlying Raukumara, and that its appearances in overlying beds are all doubtful.

Outside of the Poverty Bay and Dannevirke areas Rzehakina is very rare. This may be owing to the absence of suitable lithologies. Yet the variety of facies in the Rakauroa carrying abundant Rzehakina is considerable (conglomerates, sandstones, black shales, siliceous marls, bentonitic red shales, etc.). Some of these facies of suitable age certainly occur elsewhere in New Zealand, but the only other locality where an occurrence is known is 1283, near Silverdale, North Auckland. Here it is not uncommon and genuinely part of the fauna, which comes from a hydraulic limestone area and though that formation is Wanstead Eocene, it seems to be from an underlying Cretaceous horizon. The rest of the fauna is rather similar

This seems a suitable place to point out and correct an error regarding the Whaingaroa formation which has been world-wide ever since Stache described its fauna in 1864. Every subsequent author seems to have accepted without question his dictum as to its Eocene age. Cushman and Dusenbury, for example (1934, page 52), identified several Eocene Californian species with these supposedly Eocene New Zealand forms; several other workers have made similar comparisons and as late as 1937 Cushman and Parker (1937, page 66) took the age of topotypic Whaingaroa material as Eocene. Even in Australia this opinion has been prevalent; Chapman, with little relevant material, quoted Upper Eocene in his New Zealand Bulletin, and Parr and Collins (1937, pp. 190–211) quite recently discussed the Whaingaroa Polymorphinidae in terms of this age. In a later note, quoted in a paper by Bartrum (1937, p. 436), Parr has been the only writer to amend this conclusion and suggest a Lower Oligocene age. In 1933 I had written him stratigraphic notes on the succession in Poverty Bay, pointing out that much Tertiary still underlay beds coeval with the Whaingaroa, and his examination of a definite Eocene fauna from North Auckland made him realise that Stache's conclusions must be wrong. Stache's error was pardonable at the time, as he had no New Zealand comparative material and no knowledge of the stratigraphic column; it is only by intensive comparison with standard sections from many places in our Lower Tertiary that the Whaingaroa can be accurately placed. It is, as a matter of fact, Upper Ototaran, which is probably Middle Oligocene. Well defined faunas of the underlying Waiarekan and Lower Ototaran and the overlying Duntroonian and Waitakian (all stages of the Oligocene) are known in sequence at several localities, and the relation of the Whaingaroa fauna to the Eocene is but slight. The evidence for all this will be elaborated elsewhere—but the current misconception of the “Novara” fauna should be rectified.

The species Rotaliatina sulcigera (Stache) (plate 69, figs. 6, 7) is an extremely variable one in size, shape and height of spire, but has a general strong resemblance to mexicana, the genotype. This was from the lower Oligocene Alazan shale. Nuttall (1930, p. 272) has recorded it also as common in the upper Eocene Chapapote and rare in the middle Eocene Guayabal. This roughly corresponds to its New Zealand range. The apertural and sculptural characters appear much the same in the two species, but the New Zealand form is readily distinguishable by its more numerous chambers, having almost always nine (very occasionally eight to ten), instead of seven in a whorl.

Besides the Eocene and Oligocene forms, a species has recently been described from the Pliocene of West Java. This is R. globosa, Yabe and Asano (see 1937, page 124, plate 19, fig. 7). The authors remark that “This is a large species. … The raised and granulated sutures are of peculiar type … .” The size—about 2 mm.—is in fact enormous for the genus and the whole appearance of this Pliocene form shows plainly that it is derived from typical Javan species of Rotalia, such as schroeteriana P and J., and especially alveiformis Thalmann. It is really a form of the latter in which the