farthest removed group: while still closely resembling the Aporrhaidae in the nervous system, they have become highly specialised in the digestive system (Yonge, 1932) and in the female reproductive organs and often grotesquely modified in external features. The aporrhaid-struthiolariid line diverged from the basal stock by the adoption of a burrowing habit, with the formation of siphonal tubes by the elongated proboscis; the pharynx and radula are reduced owing to the mode of feeding on finely divided plant detritus. The Struthio-lariidae have proceeded a good deal further in their evolution than the aporrhaids, most notably in the following features:

Marwick and Finlay (supr. cit.) suggest that Struthioptera gave rise to Conchothyra, which includes the earliest shells assigned to the Struthiolariidae. The next step is represented by Struthiolarella, which would then “appear as an ancient basic member of the Struthiolariidae.” Radular and opercular material first becomes available for the family' with the two species of Perissodonta, the most primitive living members of the group, which show a widely discontinuous geographical distribution, P. mirabilis occurring at Kerguelen Land and P. georgiana at South Georgia. They probably, however, represent true relict forms of an ancient stock, since both agree in the peculiar multiplication of the marginal teeth of the radula. The shell agrees most closely with Struthiolarella, with which genus Steinmann and Wilckens (1908) proposed to incorporate Perissodonta. Marked aporrhaid features are retained, such as the strongly rounded body whorl; the strong, sigmoidally curved axial ribs, especially well developed on the older whorls; the unisinuate outer lip receding at the suture to form a wide posterior sinus; and the straight columella. The radula (Text Fig. 5) is also reminiscent of Aporrhais: the central tooth shows the transition from a row of sharp denticulations to the large, multi-serrate cusp of the modern struthiolariids, while the laterals have the triangular aporrhaid shape as distinct from the rectangular form in Struthiolaria and Pelicaria. However, the increase of the marginals to five pairs is an anomaly which indicates that Perissodonta has diverged some distance from the main stem of the family.

The operculum of Aporrhais pes-pelicani (Text Fig. 5) forms an oval dise, with a callused rim surrounding the muscle scar, and the distal end produced into a wide, squarish plate. The struthiolariid operculum (Text Figs. 6 and 7) is derived by the production of the callus rim to form a sharp distal spine—an adaptation for the use of the operculum in assisting locomotion by thrusting into the substratum, In Perissodonta (Text Fig. 6) the operculum is of the clawed

struthiolariid type, but like the radula diverges from the main evolutionary line, the disc being strongly angled along the axis of the spine.

The animal of Perissodonta is a typical struthiolariid: the gill filaments are elongate and linear, and there is a food groove, suggesting that ciliary feeding takes place. Unfortunately, the four specimens examined were all males, and it is not yet certain whether or not a brood pouch is present. The shell apex is small and planorboid, as in Struthiolaria, suggesting that there is a free-swimming stage of some duration.

Among the modern struthiolariids, Pelicaria is clearly an advanced genus which, according to Marwick (1923), diverged from Struthiolaria in the late Oligocene or early Miocene, is scantily represented in the Miocene, and exhibits a marked speciation in the Pliocene. The salient diagnostic feature is the capuliform apex, and in addition Finlay (1927) has pointed to the well-marked shell characters distinguishing Pelicaria from Struthiolaria: the presence of strong, spiral cinguli giving a bicarinate or tricarinate appearance to the whorls: the frequently rounded body whorl: the canaliculate suture line: the lesser development of the inner lip callus and the almost straight outer lip. The radula (Text Fig. 8) is very close to that of Struthiolaria (Text Fig. 7), but has a highly distinctive feature in the exaggeration of the cusps of the central and lateral teeth. The central cusp is very long and well overlaps the succeeding tooth.

Restriction of range in Pelicaria at an early stage, consequent upon the retention of the veliger, may well have been a contributing factor in the more complete separation of the genus. Ecological isolation possibly played a part in the original divergence of Struthiolaria and Pelicaria, the latter showing at the present day a greater tolerance of muddy conditions, although there is no apparent adaptive difference and the two forms very generally overlap. The effect of delay in liberation of the larvae is clearly seen in the restricted geographical range of the existing struthiolariid genera, as compared with the wide range of Aporrhais in the northern hemisphere. It is significant that Pelicaria vermis is confined to the North Island of New Zealand, while Struthiolaria papulosa extends throughout New Zealand and reaches the Kermadecs. The pronounced Pliocene radiation of species of Pelicaria may similarly be ascribed to geographical isolation following the elimination of the larval stage, while the tendency of the recent Struthiolaria papulosa to break up into local races may be also due in part to the shortening of the free-swimming phase. It may be remarked, however, that the so-called geographical “races” of S. papulosa do not appear to be always clear-cut: for example, by no means all of the Foveaux Strait specimens are of the smooth shouldered “gigas” type, while numerous examples of the short, moderately nodulose shell, of the topotypic Cook Strait form have been collected from Cheltenham, Auckland.

It is interesting to note that Tylospira scutulata, the single living Australian representative of the Struthiolariidae, has a bulbous apex similar to—but rather smaller than—that of Pelicaria. The animal

and life history of Tylospira are unknown, but from consideration of sculpture characters Finlay (1931) places the genus fairly close to Pelicaria. It is certainly a highly advanced, gerontic form, and it would be desirable to examine the radula and operculum—which the writer has not so far been able to procure—in order to determine the exact relation of the genus to the main stem of the family. It is at least a possibility that the scaphelloid apex may be a parallel development in Pelicaria and Tylospira, by the independent loss of the free swimming stages.

The following phylogenetic diagram expresses the relationships of the Struthiolariidae, as suggested from the geological evidence, and the anatomical results presented above.

References To Literature

1. Finlay, H. J., 1927. A Further Commentary on New Zealand Molluscan Systematics. Trans. N.Z. Inst., 57, 496–545.

2.— 1931. On Austrosassia, Austroharpa and Austrolithes, new genera: With some Remarks on the Gasteropod Protoconch. Trans. N.Z. Inst., 62, 7–19.

3. Finlay, H. J., and Marwick, J., 1937. The Wangaloan and Associated Molluscan Faunas of Kaitangata, Green Island Subdivision. New Zealand Geolog. Survey. Pal. Bull., 15. Govt. Printer, Wellington.

4. Fretter, Vera, 1942. The Genital Ducts of some British Stenoglossan Prosobranchs. J. Mar. Biol. Assoc. U.K., 25, 173.

5.— 1940. The Genital Ducts of Theodoxus, Lamellaria and Trivia, and a Discussion on their Evolution in the Prosobranchs. J. Mar. Biol. Assoc. U.K., 26, 312–349.