Art. XLVIII.—A Comparison of the Land Molluscan Faunas of the Kermadec Group and Norfolk Island.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 47, 1914, Page 498

Art. XLVIII.—A Comparison of the Land Molluscan Faunas of the Kermadec Group and Norfolk Island. By Tom Iredale. Communicated by W. R. B. Oliver. [Read before the Auckland Institute, 16th December, 1914.] I. Introduction. In the Proceedings of the Malacological Society of London, vol. x, pp. 364–88, 1913, will be found a detailed account of the land molluscs of the Kermadec Group. I had hoped to review this fauna as a zoogeographical item, and contrast it with that of the other island groups which have been generally named in connection with this group—viz., Lord Howe Island and Norfolk Island. But my own collection consisted mostly of minutiae, and I was some-what surprised to find that the minute molluscan fauna of these groups was unknown. Since then, however, a very large and varied collection of the land molluscs of Norfolk Island has been made by Mr. Roy Bell, my friend and companion collector at the Kermadecs, who has obtained large quantities of minute forms. Still, no general review can be made, as no collections of minute forms are yet known from Lord Howe Island, the Fiji Group, or New Caledonia. As, however, Norfolk Island has always been a favourite study of the Neozelanic zoologist, especially on account of the former existence on that island of species of birds of the Neozelanic, otherwise, endemic, genera Nestor and Hemiphaga, I propose herewith to make a comparison between the land molluscs known from the Kermadecs and from Norfolk Island. II. List of the Kermadec Group Land Mollusca. In my collection made at the Kermadecs only twenty species were represented; previous to my arrival at the island six species were on record, and, as two were not recovered by me, sixteen species were added. All these were minute forms. One of the most interesting results was the discovery that the molluscs were divisible into two groups—one composed of species which lived upon trees, and which were not found on the ground save through adverse circumstances, such as falling leaves; whilst the other group consisted of species which lived under stones, leaves, logs, &c., on the ground, and were never found upon trees. The tree-dwelling molluscs were,— Helicarion kermadecensis (Smith). Pronesopupa senex Iredale. Ptychodon royanus Iredale. Tornatellina sp., a slender form Calymna arboricola Iredale. — sp., near bilamellata Anton. Flammulina miserabilis Iredale. Elasmias inconspicua (Brazier). Charopa pseudanguicula Iredale. To the ground were confined,— Fanulum expositum (Mousson). Paralaoma raoulensis Iredale. Kieconcha kermadeci (Pfeiffer). — ambigua Iredale. Ptychodon pseutes Iredale. Tornatellina novoseelandica Pfeiffer (? = subperforata Suter). — amandus Iredale. Charopa macgillivrayana Iredale. — sp., a conoid form. — exquisita Iredale.

Whilst the twentieth was a semi-amphibious creature recognized as “Barleeia”! chrysomela Melvill and Standen, described from the Lifu Group. The reason the Tornatellina spp. are unnamed is that just after I had examined the species I noted that Dr. Pilsbry, of the Philadelphia Academy of Sciences, had in preparation a monograph of this difficult group, and I therefore at once sent him my material to incorporate in his work. This has not yet appeared, but it will do so in the near future. III. List of the Norfolk Island Land Mollusca. The Norfolk Island land Mollusca were studied by Sykes from collections at the British Museum, chiefly made by John Macgillivray in 1855, and the results published in the Proc. Mal. Soc. (Lond.), vol. iv, p. 139 et seq., 1900. Twenty-five species were recorded, and the conclusion arrived at was, “As pointed out by Professor Tate and others, the faunal relationship of Norfolk Island lies rather with New Zealand than with the Australian Continent.” This sentence would imply the recognition of characteristic New Zealand forms in the Norfolk Island fauna, but the list here given, with Sykes's nomenclature, does not show any such. Sykes recorded,— Microcystis nux Sykes. Charopa? depsta Cox. — castaneocincta Sykes. — ? quintali Cox. Trochonanina platysoma Sykes. — ? patescens Cox. Fretum phillipii Gray. Endodonta norfolkensis Hedley. — suteri Sykes. Succinea norfolkensis Sykes. — grayi Sykes. Vertigo norfolkensis Sykes. Rotula campbellii Gray. Omphalotropis brenchleyi Sykes. Medyla insculpta Pfeiffer. — albocarinata Mousson. — imitatrix Sykes. — cerea Pfeiffer. Sitala macgillivrayi Sykes. — navigatorum Pfeiffer. Carthaea stoddarti Gray (incl. C. flosculus Cox). — suteri Sykes. Diplommatina coxi H. Adams. Charopa exagitans Cox. Paludestrina norfolkensis Sykes. Accepting the generic names used by Sykes, out of fourteen genera admitted only five have been included in the Neozelanic fauna. Two of these five have only scare and doubtful records from the extreme north of New Zealand, two others are the Polynesian Charopa and Endodonta, whilst the fifth is the Neozelanic endemic Carthaea. In this last case there is not a very close resemblance to the New Zealand shell, and the generic location is apparently quite wrong. It will at once be seen that there is no close relationship between the Kermadec and Norfolk Island faunas when my list and Sykes's are placed side by side. However, as the greater proportion of my forms were minutiae, and as such minutiae were not represented in Sykes's Norfolk Island collection, no critical comparison could be made. What was certain was that Norfolk Island was inhabited by quite a different and more vigorous shell fauna, but the sequel is altogether beyond the imagination of the most sanguine conchologist. Mr. Roy Bell's collection, which certainly does not include all the forms living on the Norfolk Island group, shows that on the main island over forty distinct species live, whilst some five or six are known as subfossil

only; from Nepean Island, a rock three-quarters of a mile away from the main island, some half-dozen subfossil forms were obtained; whilst the Phillip Island Mollusca, through adverse weather, were not collected, and four species are already known from that island. I herewith give the Norfolk Island land molluscan fauna as known from the results of Mr. Roy Bell's collection, worked out by H. B. Preston— Main Island Living Molluscs. Dendrolamellaria mathewsi Preston. Pittoconcha concinna Preston. Advena campbellii (Gray). Nitor retinaculum Preston. — var. charon Preston. Macgillivrayella crystallina Preston. Greenwoodoconcha nux (Sykes). Cryptocharopa atlantoididea Preston. — castaneocincta (Sykes). Norfolcioconcha norfolkensis Hedley. — tomi Preston. Charopa mathewsi Preston. Allenoconcha basispiralis Preston. — sororcula Preston. — belli Preston. Paralaoma orestias Preston. — mathewsi Preston. — perminuta Preston. — monspittensis Preston. — depressior Preston. — perdepressa Preston. Succinea (Tapada) norfolkensis Sykes. — royana Preston. Nesopupa norfolkensis (Sykes). — congener Preston. Omphalotropis albocarinata Mousson. Roybellia platysoma (Sykes). — var. brenchleyi Sykes. — depressa Preston. — suteri Sykes. Belloconcha norfolkensis Preston. — cerea Pfeiffer. — suteri (Sykes). Palaina coxi (H. Adams). Fanulum insculptum (Pfeiffer). — norfolkensis Preston. — imitatrix (Sykes). — belli Preston. — testudo Preston. Tornatellina norfolkensis norfolkensis Preston. Quintalia stoddarti floscula (Cox). Mathewsoconcha beli Preston. Tornatellina norfolkensis moohuensis Preston. Iredaleoconcha inopina Preston. — caloraphe Preston. Paludestrina norfolkensis Sykes. Collected by J. Macgillivray; not obtained by R. Bell. Johannesoconcha mutivolva Preston. Sitala macgillivrayi Sykes. — pusillior Preston. Found Subfossil, but not Living, on Main Island. Fretum microstriatum Preston. Johannesoconcha minuscula Preston. Mathewsoconcha albocincta Preston. Norfolcioconcha iota Preston. — vexillum Preston. Found Living on Nepean Island. Tornatellina norfolkensis nepeanensis Preston. Tornatellina duplicilamellata Preston. Found Subfossil, but not Living, on Nepean Island. Advena campbellii nepeansis Preston. Quntalia stoddarti nepeansis Preston. Belloconcha elevata Preston. Belloconcha compacta Preston. Succinea nepeanensis Preston. — humerosa Preston.

Phillip Island Records. Belloconcha phillipii (Gray). Advena campbelii campbellii (Gray). —grayi (Sykes). Quintalia stoddarti stoddarti (Gray). Indeterminate Records. Charopa exagitans (Cox). Charopa ? quintali (Cox). —? depsta (Cox). —? patescens (Cox). Introduced Species Living on Main Island and Nepean Island. Vallonia excentrica Sterki (recorded by Sykes as V. pulchella). As a key to Sykes's records the following generic identities must be carefully noted:— Microcystis Sykes = Greenwoodoconcha. Charopa = intermediate spp. Trochonanina = Roybellia. Endodonta = Norfolcioconcha. Fretum = Belloconcha. Succinea = Succinea. Rotula = Advena. Vertigo = Nesopupa. Medyla = Fanulum. Omphalotropis = Omphalotropis. Sitala = Sitala. Diplommatina = Palaina. Carthaea = Quintalia. Paludestrina = “Paludestrina.” I have had the pleasure of critically examining the whole of the material Mr. Roy Bell collected, and I here propose to discuss the affinities of the molluscs. Like my own Kermadec list, the Norfolk Island list abounds in new generic names, and it is remarkable how extraordinary the difference is between the two lists: in the former “Endodonts” predominate, in the latter they are almost a negligible feature. Zonitoids practically do not exist in New Zealand; two species occurred at the Kermadecs, while they are omnipresent in the Norfolk Island fauna. IV. Digest of the Kermadec Land Mollusca. A digest of my Kermadec collection must be first made, then a similar one of the Norfolk Island one, and the results compared. Thus, at the Kermadecs the largest snail-shell was that of Helicarion kermadecensis (Smith). As the animal in that genus is large, it was obviously the largest snail. It only occurred to me in one patch at the highest point of the island, though both to Macgillivray and Graeffe it appeared commonly. The shell agrees quite closely in character with that of H. cuvieri (Férussac), the Australian shell, which is the type of the genus, and the generic location seems correct This is mentioned as the genus-name. Helicarion has been used to include any glassy, loosely coiled shell, and these have been recently found to cover many diverse genera of snails. I do not admit the occurrence of Helicarion kermadecensis (Smith) at Hobson's Glen, Auckland, as there are too many doubtful factors in this record. It would therefore appear that Helicarion kermadecensis (Smith) arrived at the Kermadecs from the north. It was the only member of the family Helicarionidae which I consider absent otherwise from the New Zealand zoological region. The next two largest species are referable to the family Zonitidae, as used at the present time. This is a wide group, and future systematic work

will materially restrict its limits. One species was introduced as Trochonanina exposita by Mousson, the other as Helix kermandeci (recte kermadeci) by Pfeiffer. For the former I introduced the genus Fanulum; for the latter I propose the new genus Kieconcha. The peculiar features of the former may be gauged by the fact that it was recently classed in Medyla (the correct name of this genus is Otesia), the species of which differ entirely in shell characters. It will probably be found to have little relationship with the species hitherto associated with it when the anatomy of the snail is carefully considered. I might add that a report upon the animals of these Kermadec snails is now under consideration, and will probably later appear in these Transactions. The genus Kieconcha seemed a necessity, as previously the species had been classed first in Microcystis and then in Macrochlamys. The first was easily seen to differ entirely in shell characters, whilst the latter has been shown to be restricted to India. Kieconcha somewhat recalls the north Australian Nitor. If this suggestion should be confirmed it may have been a companion with Helicarion from the north. Nothing like Kieconcha or Fanulum otherwise occur in the New Zealand biological region. Ten species were obtained which may be classed in the family Endodontidae, as that term was used by Pilsbry. I have already given my reasons for objecting to such a group, and so far I have received little but encouragement in the course I adopted. I am now making a conchological survey of the Australasian “Endodonts,” so will defer my further remarks until a future date. I included three species in Ptychodon, pointing out that the usage of Thaumatodon, unless restricted genera were utilized, was incorrect. My largest species was of the aspect of Thaumatodon, whilst the two lesser were of quite a different nature. Three species were ranked under Charopa, but here again three different styles of shell were included, and the main reason for so classing them was the unarmed mouth. For one flattened widely umbilicated shell I introduced the subgeneric name Discocharopa, as I found a similarly formed species in every detail occurred in so far away a locality as Bass Strait. I have since recognized the same subgeneric form in Queensland. The Ptychodon which I would consider of Thaumatodon aspect is a Polynesian form of shell, whilst the other Ptychodon are nearly related to New Zealand species. All significance of the latter fact is minimized by our entire ignorance of the minute forms which must abound in Polynesia. Research may show this latter Ptychodon to be quite common in the northern island groups. The same remarks apply to the Charopoid forms, and even more forcibly to the two tree-dwelling species I classed in the “Flammulinoid” section of the Endodontidae. I suggested that this section should be given family rank as the family Flammulinidae, but so far apparently no anatomical features are known whereby a diagnosis can be prepared. For two species I introduced a new generic name, Paralaoma. Since my paper was completed I have studied the Norfolk Island molluscs, and, making comparisons, I believe that three species of this genus may be distinguished at the Kermadecs. However, I include this genus with misgiving in the family Endodontidae, and have noted its wide range throughout eastern Australia, where it seems to be well represented. According to one authority, the species I would allot to Paralaoma could be classed in Phrixgnathus, whilst another claims their near relationship to Flammulina Such diversity of location sufficiently excuses a new generic term.

A minute tree-dwelling pupoid form would have been classed in Nesopupa save for the fact that the mouth was unarmed. Almost facsimiles, but with fully armed aperture, have been examined from Tahiti. I drew attention to the lack of arming in the aperture by creating the new genus Pronesopupa, thereby suggesting the greater antiquity of my shell. I did this as I noted that juvenile Tahitian shells were similarly unarmed. Nesopupa and its allies are typically Polynesian, no form whatever being known from New Zealand. The Tornattellinids are unnamed, for reasons given; but I studied them a little, and I can state that of the five species three were typically Polynesian, whilst Suter has recorded one as occurring at the Kermadecs and north New Zealand. In the latter case it may have only recently reached New Zealand, as undoubtedly the fifth, which is referable to a very distinct genus, Elasmias, is at present being carried about the Pacific. I strongly feel that it is a comparatively recent settler on the Kermadecs, and, as it is very rare at Lord Howe Island, may have also recently arrived there. This review will indicate that the Kermadec land Mollusca tell us very little, save that they apparently show their source to have been northern in preference to southern. When contrasted with the Norfolk Island molluscs much more will be seen. V. Digest of the Norfolk Island Land Mollusca. The Norfolk Island main-island living forms are of most interest in this connection; and, firstly, Dendrolamellaria matthewsi is of peculiar value, as this is a species perhaps referable to the family Helicarionidae, but of such peculiarities in the shell alone that Dendrolamellaria would appear justifiable as a genus. It recalls the marine genus Lamellaria in form, and is quite unlike the Kermadec Helicarion in every detail. Its ancestry at the present moment is quite obscure. Advena campbellii is what has been known as Rotula campbellii. The shell differs in character from that of Rotula, which, moreover, is a genus confined to Mauritius. It is a fine handsome shell, and its character can be estimated from the fact that its nearest relative was ever supposed to be the Mauritius shell. Anatomical study may later give us some clue to its affinity. The three species classed in Greenwoodoconcha are such as are often referred to Microcystis, and one recalls that genus, which was first described from Pitcairn Island, in its colour-markings. These shells are, however, somewhat solid. The half-dozen species of Allenoconcha would probably also have been placed under Microcystis, but they differ appreciably in shell characters, and the shells are very fragile. One species of Microcystine shell has been classed under the north Australian genus Nitor, as the shell features agree quite closely with that genus. None of these Microcystine genera closely resemble the Kermadec Kieconcha, the Nitor having the most likeness. A most beautiful species was called Trochonanina platysoma by Sykes. Another species has been discovered by Mr. Roy Bell, and for these the genus Roybellia has been formed. The shell, in its depressed shape, recalls Fametesta mirabilis Pilsbry, from the Bonin Group, but upon comparison the Norfolk Island species is found to differ in texture entirely, being a much more beautiful and delicate shell As animals have been preserved, we shall later learn of the affinities of this remarkable form. It vaguely suggests itself as a relation of Fanulum.

As above noted, I introduced this genus-name for a smooth Kermadec shell, and at the present time three Norfolk Island shells are associated with it under my genus-name. Two of these are heavily sculptured, but the third is smooth. I now doubt that these are congeneric with Fanulum, as they all have a toothed and callused columella, though varying in shape and sculpture. The report upon the anatomy of the species will settle this point. Sykes proposed Fretum to cover a series of Norfolk Island shells, but selected as type a Fijian species. This is not considered congeneric, so that Belloconcha is made use of to indicate these Norfolk Island species, whose generic affinity, however, would seem to be the Fijian Fretum Quintalia has been introduced as a genus-name for the species Sykes classed under the New Zealand endemic genus Carthaea I cannot see any reason, from a study of shell characters, for the association of these two. Carthaea, moreover, is placed in the Endodontidae, whereas I would not so class Quintalia. Three extraordinary genera are represented in Mathewsoconcha, Iredaleoconcha, and Pittoconcha. Few specimens have yet been obtained, but they all show quaint features. Mathewsoconcha recalls Belloconcha in form, but it has a callused semi-toothed columella. Iredaleoconcha recalls Philippine shells in general appearance, being somewhat flattened, keeled with convex base, but it has a deeply channelled suture. A shell of very similar appearance has been lately obtained near Obi, in the Moluccas. Pitto-concha is a small, somewhat conical, heavily sculptured shell, suggestive of the style of sculpture seen in the Lord Howe species howinsulae. Macgillivrayella is a crystalline flattened Zonitoid of ordinary appearance, yet quite unlike anything yet known. The preceding completes the Zonitoids as at present known living on Norfolk Island, but Macgillivray collected three very minute forms which Mr. Roy Bell did not obtain. One of these seems a typical Sitala as that genus is represented in Fiji, whilst the other two are called Johannesoconcha, and are quite remarkable minute forms, glossy and many-whorled; nothing has yet been collected from the north with which Johannesoconcha can be compared. A shell of a Charopoid facies, but which plasters itself over with mud, has been called Cryptocharopa. It is somewhat doubtful what its affinities are, but some molluscs in New Caledonia indulge in the same mud-plastering habits. The “Endodonts” are minute and very sparsely developed; they number six if Paralaoma be considered such, but three only otherwise. One which has an armed aperture, and would be broadly classed in Ptychodon, has been generically termed Norfolcioconcha, as it does not fit well in with either Ptychodon (typical), Thaumatodon, Nesophila, or any other section. The other two are flat discoidal unarmed forms which are classed in Charopa. It is somewhat remarkable that these three do not come at all close to the Kermadec “Endodonts”; but here again ignorance of northern minutiae forbids investigation into their affinities. My genus Paralaoma, instituted for Kermadec species, is represented by three forms, but as this genus is somewhat commonly distributed in eastern Australia no great stress can be laid upon this fact. I suggest that its discovery in northern groups may be confidently anticipated. It seems a well-defined and easily recognizable genus, and I look forward to the investigation of its anatomy with great interest.

A species of Succinea lives on Norfolk Island; this genus is absent from the Kermadecs and New Zealand, though present in Australia and the northern groups. The Polynesian genus Nesopupa is represented by a large sinistral species with the mouth well armed, and quite unlike the Kermadec dextral unarmed form I have called Pronesopupa. No relation of either form occurs in New Zealand. The Polynesian Omphalotropis is well developed, three species and one variety being admitted; these species are closely allied to Fijian forms. No member of this genus was found at the Kermadecs, though one species has been recorded from the extreme north of New Zealand; but the record is somewhat doubtful. The family Diplommatinidae is also credited with three representatives, all belonging to the genus Palaina. This is purely a Polynesian group; a solitary species appears on the New Zealand list, but will be almost certainly omitted in the near future, and no member of the family occurs at the Kermadecs. Two subspecies only of Tornatellina have been recognized, one of which is confined to the Moohu Stone, a rocky islet off the north coast. The peculiar item of note is that this species is not close to any of the five distinct species I obtained at the Kermadecs, but recalls a Fijian species. A “Paludestrina” concludes the list of living molluscs obtained on Norfolk Island. I am unable to give any idea of its relations, save that it differs obviously from the New Zealand Potamopyrgus, and no freshwater univalve was obtained at the Kermadecs. Of as great, or even greater, interest were the shells collected as subfossil from the limestone quarry at the south of the island, near Kingston. Associated with living forms such as Succinea norfolkensis Sykes, Omphalotropis albocarinata Mousson, Quintalia stoddarti floscula (Cox), Fanulum insculptum (Pfeiffer), Fanulum imitatrix (Sykes), Advena campbellii (Gray), Greenwoodoconcha castaneocincta, (Sykes), were numerous shells apparently now quite extinct. The largest shells are the next in size to Advena campbellii (Gray), and are of remarkable interest, as being, from a very close criticism of shell characters, strictly referable to the Fijian genus Fretum. Smaller shells are apparently typical Mathewsoconcha; and in this connection it is worthy of note that the living species of this genus is, as far as is at present known, exceedingly rare. From sand contained in these subfossils two minute species were sorted out—-one a Johannesoconcha, the other a Norfolcioconcha—both of which differ from the living species. Still more peculiar is a large quantity of dead shells collected on Nepean Island. This island is only three-quarters of a mile distant, and is apparently devoid of vegetation. Three living species were obtained—two species of Tornatellina, and the third Vallonia excentrica Sterki, an introduced European form. This latter species has acclimatized itself on Norfolk Island, and by some means has reached this inhabitable islet, and is now flourishing there. The two species of Tornatellina are interesting, as one of them is only sub-specifically distinct from the Norfolk Island species, whilst the other is quite different, yet has not been discovered on Norfolk Island. It does not come near the Kermadec species. The dead shells are, however, of an extraordinary nature; the large bulk is composed of two species of Belloconcha, differing appreciably from the main-island species.

The next most numerous shell is a large form of Quintalia stoddarti, intermediate between Q. s. stoddarti (Gray) from Phillip Island and Q s. floscula (Cox) from the main island. A rarer shell is a fine large variety of Advena campbellii called Advena campbelli nepeansis. It is much larger and more flattened that either the Phillip Island or main-island forms. Succcinea is well represented, and at the present time it is impossible to decide the truth regarding these. Two species are separated, a very large slender elongate form called Succinea nepeanensis, and a more common one with shouldered whorls named Succinea humerosa. This might be best considered as the representative subspecies of the mainland S. norfolkensis Sykes, than which it averages larger in size. Omphalotropis is also fairly well represented, but the variation in sculpture and form known in this genus has prohibited the differentiation of the Nepean Island subfossils. One of the most interesting items to me was the discovery of Fanulum insculptum (Pfeiffer), occurring rarely; whilst one specimen of Fanulum imitatrix (Sykes) was observed. Altogether a number of the F. insculptum (Pfeiffer) were separated, and, though they have not been named, they average smaller and more conical than the living main-island form. The reason for the extinction of these Nepean Island molluscs is at present unknown, but the facts seem to show that the species formerly living were of larger size than the present existing species, whilst the Fanulum seems to show the reverse—that is, the examination of the available material, which is not small, tends to the conclusion that whereas Advena, Quintalia, and Mathewsoconcha (probably also Succinea) have decreased in size or become extinct, Fanulum has increased in size and numbers, but whether at the expense of the other genera is not evident Mr. Roy Bell was unable to collect on Phillip Island, but four species have been described from that locality—viz., Belloconcha phillipii (Gray), Belloconcha grayi (Sykes), Advena campbellii (Gray), and Quintalia stoddarti (Gray). The latter two are still represented on both Nepean Island and the main island, but the exact relationships of these are not known The type set of Quintalia stoddarti (Gray) are much larger than the very long series from Nepean Island, which again greatly exceed in size the living main-island shells, which also differ in their more conical form. These three are ranked as subspecies only, and there seems little doubt as to the accuracy of this determination The type, however, of Advena campbellii (Gray) is a small shell not much larger than the living main-island form. As the subfossil Nepean Island shells are much larger and more depressed, the suggestion at once occurs that the type from Phillip Island may not be representative of the form there living. The species Belloconcha phillipii (Gray) is the giant of the genus, whilst Belloconcha grayi (Sykes) is also a large species. These both surpass the subfossil Nepean Island species, whilst on the main island only one small species has been found alive, and so far with very restricted habitat, though two other small species have been found subfossil. I hope later to be able to record the results of a new search for Phillip Island molluscs, and note their relationships. It remains to note that four species recorded by Sykes were not obtained by Roy Bell—viz., Charopa exagitans (Cox), C.? depsta (Cox), C.? quintali (Cox), and C.? patescens (Cox). From the published de-

scriptions and figures, none of these can be referable to Charopa as now understood, and they can only be considered as very doubtful constituents of the Norfolk Island fauna. VI. Review and Conclusions. The conclusions to be drawn from these digests can now be summarized, and strongly indicate the different origin of the land molluscan faunas of the two groups. Whereas the Kermadec series alone did not tell much, when contrasted with the Norfolk Island collection quite a deal can be determined. The apparent source of the Kermadec molluscs is Polynesia, the Neozelanic element being quite probably also of that origin, having arrived from the north. It might here be observed that, whilst the Endodontidae have been sometimes claimed as an Antarctic group, equally strong claims have been made for their northern origin. Two masters in the science of drawing deductions from observations made on the distribution of land molluscs have differed on this point. I would suggest that correlation might be brought about were restricted groups utilized. As far as my own observations have yet proceeded, I find groups giving an Antarctic range, whilst other groups as surely show a northern origin. Thus in the Kermadecs the “Thaumatodon” group emphasizes a northern source, whilst just as surely the “Discocharopa” group suggests a southern origin. I hope to deal with such questions in more detail later, giving facts in support of these suggestions. The most casual examination of the Norfolk Island land Mollusca prohibits any idea of the attachment of this fauna to that of New Zealand. The preponderance of Zonitoids and the scarcity of “Endodonts” are equally assertive, whilst the nature of the Zonitoids suggests at once Fiji as a source of supply. The lack of Placostylus is somewhat extraordinary, whilst, though no species on the Fijian Group recalls Advena, it must be remembered that the Fijian Group has also suffered such vicissitudes as would easily account for the extinction of its relations, if any ever existed there. The presence of a subfossil inseparable in shell characters from the Fijian genus Fretum is noteworthy, especially as there appears little doubt that Belloconcha is a closely allied genus. This genus would seem to have been the predominant Norfolk Island mollusc, though now Fanulum appears to have taken this place. The extraordinary Roybellia, I have suggested, may be allied to Fanulum. The exact relationships of the other Zonitoids is difficult to suggest, as so little is known of the Polynesian” minutiae. The Microcystine genera Greenwoodoconcha and Allenoconcha are certainly closely allied to “Microcystis” and the varied Polynesian shells commonly known under that genus-name. The Omphalotropis, Palaina, and Nesopupa are very closely allied to Fijian and other Polynesian species. Indeed, I suggest that longer series from other localities would cause their degradation to the rank of subspecies only. Though Norfolk Island is situated somewhat midway between New Caledonia and New Zealand, and in recent charts soundings show the presence of a submerged ridge connecting the three, the land molluscs prohibit any direct connection between the three, whilst they just as surely point to a junction in some way or other with the Fijian Group. I make this reservation as, though Placostylus has reached that group, it has not been found Recent or fossil on Norfolk Island. Though I have not introduced any mention of the Lord Howe Island land Mollusca, I would simply note that Hedley pointed out these were

of New Caledonian origin, and my own studies amply confirm this. I hope to supply a detailed account of these later, but would state that they differ totally from those of the Kermadecs, Norfolk Island, or New Zealand, whilst they agree in much detail with New Caledonian land molluscs. My results are,— The Kermadec land Mollusca are few in species and very minute. The few Zonitoids and comparative abundance of the “Endodonts,” the absence of Omphalotropis and Palaina whilst Tornatellinids are fairly abundant, indicate the lack of any recent land connection, but suggest the recent peopling of the group by means of drift. The nature of the molluscs incline the direction of the drift to be from the north. The conclusion that the present land molluscs would force the inclusion of the Kermadec Group under the term “Oceanic islands” would also suggest its inclusion in whatever biological region the other groups favoured. The northward-drift tendency at once suggests Polynesia, but the combined results of other studies favour New Zealand. If the latter be accepted, the separation I have indicated elsewhere of the group as a distinct province is amply confirmed by the criticism of the land Mollusca. The Norfolk Island land Mollusca are numerous, varied in species and genera; they strongly indicate recent and practically undisturbed land connection. The nature of the land molluscs, however, at once derides any suggestion that this has been with either New Zealand or New Caledonia, whilst it just as strongly formulates an immensely strong claim for this to have been with the Fijian Group. The Lord Howe land Mollusca are fairly numerous in species and genera, and many of large size; they confirm recent land connection, and leave no doubt as to its former attachment to New Caledonia, whilst they prohibit any land connection thereof with New Zealand. They will be fully discussed later; but, as far as land molluscs are concerned, there is not the slightest reason for their quotation in favour of a Neozelanic connection. I note this as they have been so affirmed by writers ignorant of the two land molluscan faunas, and who have simply based their conjecture upon the occurrence of Placostylus on both islands, though every other land mollusc plainly indicates the contrary. I am forwarding with this paper a series of land molluscs from Norfolk Island to be deposited in the Auckland Museum, where those interested can study them and criticize my assertions.