Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 82, 1954-55
This text is also available in PDF
(2 MB) Opens in new window
– 137 –

The Land Mollusca of Stewart and Solander Islands

[Read before Biology section, Wellington Branch, August 12, 1953; received by Editor, August 24, 1953.]

Abstract

The land snail faunas of Stewart Island and the Solanders are listed. Thirty-three species are recorded from Stewart Island and three from the Solanders. Three new species, Ptychodon gadus, charopa (Pseudegestula) smithae and Phrixgnathus rakiura, and three new subspecies Ptychodon monoplax hiarara, Phrixgnathus flemingi stewartensis, and P. rakiura solanderi are described from the area while a new subgenus Charopa (Pseudegestula) and a new subspecies Charopa (Pseudegestula) transenna brookesi are described from the South Island. The origin and relationships of the Stewart Island land snail fauna are discussed.

I Stewart Island

Although the two large endemic land snails of Stewart Island, Phelussa fulminata (Hutton) and Rhytida australis Hutton were early collected and named, growth of our knowledge of the smaller forms has been slow. In 1897 Suter listed eleven species, describing one as a new variety. In the Manual of New Zealand Mollusca in 1913 an additional two species were added. David (1934) described a new Allodiscus, the subsequent recognition of which has caused some difficulty and recorded Thalassohelix zelandiae (Gray). The figure and dimensions of this latter show that David really had a species of Phrixgnathus (almost certainly P. liratulus Suter). Powell (1939) listed 20 species, describing as new Fectola (Sub-fectola) rakiura. Dell (1952, A and B) described Flammoconcha stewartensis and Obanella allanae. In 1948 the writer collected on Stewart Island and on Codfish Island, and since then has received rich collections from Mrs. C. Smith and the late Miss O. Allan. The O'Connor Collection of land snails, now in the Dominion Museum, contains a quantity of Stewart Island material, much of it collected originally by Miss S. Traill. As a result of this revision of the Stewart Island fauna 33 species and subspecies are recorded.

There are still wide areas from which no land snails are known, especially south of the Rakeahua River. The faunas of many of the off-shore islands are unknown and these islands with their peculiar ecological conditions will certainly have rich Phrixanathus populations. A comparative study of the faunas of these small islands would almost certainly provide valuable results.

To simplify the presentation of tabular results the following abbreviation and indices are used:—

  • D.—the maximum diameter in mm.

  • H.—the height in mm.

  • S.—the height of the Spire.

  • U.—the maximum diameter of the umbilicus in mm.

  • R.—the number of riblets on the body whorl.

  • H.I.—the Height Index; the height expressed as a percentage of the diameter.

– 138 –
  • U.I.—the Umbilical Index: the diameter of the umbilicus expressed as a percentage of the diameter.

  • S.I.—the Spire Index; the height of the spire expressed as a percentage of the total height.

  • R.I.—the Riblet Index; obtained by dividing the number of riblets on the body whorl by the diameter in mm.

Systematics

Murdochia chiltoni (Suter)

This species has not hitherto been recorded from Stewart Island. It is known to the writer from: Horseshoe Bay, N. Gardner, 1/12/47; Raroa Reserve, Thule, R. K. Dell, 1/11/48; Sealer's Bay, Codfish Island, R. K. D., 7/11/48; Herekopere Island, O. Allan,—/5/51. The species is known from several South Island localities.

Phelussa fulminata (Hutton)

This large endemic species occurs very plentifully as a subfossil shell in sand dunes on Native Island, Paterson Inlet, and is known to the writer as a living species from Ferny Gully, Halfmoon Bay, O. Allan, 12/1/48, and from Halfmoon Bay, Suter Collection. To judge from the abundant remains on Native Island, this species was much commoner in the past than it is to-day. It is in the living state a rare shell and together with Rhytida australis forms a remarkable endemic element of large shells. Suter's reference to his costata as a subspecies of fulminata is a surprising combination as the type of costata is much more like henryi than fulminata. The types of both costata and henryi come from Resolution Island and both are known from remarkably few specimens. This is a question that will be considered later in this paper. The matter is raised here to show that the morphological isolation of fulminata is greater than Suter's grouping would indicate.

Therasia thaisa Hutton.

The habit of this species on Stewart Island of living under coastal Phormium and other scrub apparently accounts for its almost universal appearance in dune deposits. It is known to the writer from coastal dunes at the Neck, Paterson Inlet; Native Island; Mason's Bay; Sealer's Bay; Codfish Island and as living shell from under Phormium from Codfish Island and from Halfmoon Bay, Stewart Island, O'Connor Collection (Coll. J. H. Sorensen). The species is common throughout the South Island.

Thalassohelix obnubila (Reeve).

There has been considerable confusion with regard to the Stewart Island records of Therasia and Thalassohelix. Suter recorded Thalassohelix obnubila (Reeve) and Therasia antipoda chathamensis Suter from Stewart Island and Powell (1939) has recorded the same two species. The labelled specimens of both species in the Suter Collection actually belong to the same species and do not agree with the type series of Therasia antipoda chathamensis Suter. The writer has seen no specimens from Stewart Island that do agree with chathamensis and believes that all this type of shell belong to Thalassohelix (although Therasia thaisa does occur). The name that should be applied to the Stewart Island Thalassohelix is difficult to determine, especially as the true identity of obnubila (Reeve) is still in doubt. Determination of the true identity of this form must

– 139 –

await the disentanglement of the Therasia-Thalassohelix complex in New Zealand as a whole. In the meantime records of Therasia antipoda chathamensis Suter from Stewart Island should be regarded with suspicion.

Localities. Living: Halfmoon Bay and Horseshoe Bay, several localities; Freshwater Creek, S. Traill,—/4/42; Herekopere Island, S. Traill,—/11/42 et al.; coastal scrub, Codfish Island, R. K. D.,—/11/48. Subfossil: Native Island, Paterson Inlet, several collectors; consolidated dunes, Sealer's Bay, Codfish Island, R. K. D., 4/11/48.

Phenacohelix pilula. (Reeve)

This is an unexpected element in the Stewart Island fauna. Elsewhere it seems to occur most commonly in the Auckland area, and though it has been recorded from two localities in the southern part of the North Island, it is as yet unknown from the South Island. The closely related P. chordata (Pfeiffer), which differs in the almost closed umbilicus and wider spaced riblets, has a wider distribution, both in the North and South Islands. The Stewart Island specimens of P. pilula are not separable from a series from Auckland except that the colour pattern is more marked in the former. The well-marked colour pattern does occur in other North Island populations, so even this tenuous difference is not constant.

The species is known to the writer from several localities in Halfmoon Bay, Native Island, and Sailor's Rest, south side of Paterson Inlet.

D. H. R. H.I. R.I.
Halfmoon Bay 3.05 2.59 66 85 22
Halfmoon Bay 3.18 2.72 62 85 19
Halfmoon Bay 3.04 2.68 69 88 23
Titirangi, Auckland 3.18 2.90 64 91 20
Titirangi, Auckland 2.90 2.63 55 91 19
Titirangi, Auckland 3.04 2.63 60 85 28

Thermia cressida (Hutton)

This species is moderately well represented in collections from Stewart Island, but the exact localities are seldom indicated. At the same time it would appear to be local and seldom common in any locality. Outside Stewart Island it has been recorded from numerous localities in the South Island south of Greymouth, though at no time common. Ulva Island, S. Traill,—/8/42, and Mason's Bay, O. Allan,—/3/49, are the only two localities from which the writer has specimens with well authenticated locality data, although Halfmoon Bay is quoted in some of the old collections.

Allodiscus planulatus (Hutton)

Small specimens from Sealer's Bay, Codfish Island, and Thule, Stewart Island, match perfectly small shells of this species from the Auckland Islands and from Long Beach, Dunedin. This is a very wide ranging species throughout New Zealand and has been recorded from the Auckland Islands.

Allodiscus cf. smithi Suter

Three specimens of an Allodiscus from Sealer's Bay, Codfish Island, agree very well with Allodiscus smithi Suter, known only from Mt. Somers, except that the shell is somewhat more elevated and the riblets are considerably more numerous. The differences are best shown in the table of measurements and indices below.

– 140 –
Picture icon

Fig. 1—Phrixgnathus rakiura n.sp. Holotype. 2.27 × 1.23 mm. Fig. 2.—Phrixgnathus rakiura solanderi n. subsp. Holotype. 2.36 × 1.45 mm. Fig. 3—Phrixgnathus campbellica (Filhol). Campbell Island. 2.27 × 1.27 mm. Fig. 4—Charopa (Pseudogestula) smithae n.sp. Holotype 1.77 × 1·0 mm. Fig. 5—Phrixgnathus flemingi stewartensis n. subsp. Holotype. 3.18 × 2.05 mm. Fig. 6—Phrixgnathus liratulus Suter. Lectotype. 4.18 × 3.0 mm. Fig. 7—Phrixgnathus phrynia Hutton. Lectotype. (Syntype of P. phrynia major Suter) 3.5 × 2.41 mm. Fig. 8—Phrixgnathus phrynia (Hutton). Halfmoon Bay. 2.73 × 1.67. Fig. 9—Ptychodon monoplax hiarara n.subsp Holotype. 2.35 × 2.23 mm. Fig. 10—Ptychodon monoplax monoplax Suter. Holotype. 2.32 × 1.19. Fig. 11—Ptychodon gadus n.sp Holotype. 1.54 × 0.93 mm. Fig. 12—Ptychodon gadus n.sp. Details of aperture.

– 141 –

These differences are sufficient to warrant subspecific differentiation, but until the species of Allodiscus are thoroughly revised it will not be named.

D. H U. R. HI. U.I. R.I.
Sealer's Bay Codfish Island 2.27 1.23 0.23 143 54 10.1 63
Sealer's Bay Codfish Island 2.59 1.45 0.25 150 56 9.6 58
Sealer's Bay Codfish Island 2.5 1.35 0.27 160 54 10.4 64
A. smithi Suter (type) 2.41 1.22 0.23 108 51 10 45

Allodiscus mossi stewartensis David, 1934

The recognition of this species has occasioned some difficulty since no Allodiscus had been collected from Stewart Island by New Zealand workers, and although David's photograph indicated an Allodiscus his description did not fit any known form. In addition he compared his stewartensis with A. mossi and A. ponsonbyi, and the latter species has never been included in this genus although Suter (1913, p. 643) did note that the shell of A. mossi resembled that of Phenaco-helix ponsonbyi Suter. However, three species of Allodiscus are now known from Stewart Island, and one of these fits David's description and figure quite well. As David's original description is not readily available in this country a rather free translation is given below: “Shell slightly conical. The umbilicus is fairly wide, perspective. Colour horny with more or less regularly spaced brownish flecks together with radial streaks. Sculpture consists of sharp, fine ribs, the interstices twice as wide as the ribs. Whorls 5 ½. Spire nearly half as high as the aperture. Aperture oblique, credent shaped. This form stands between A. mossi Murdoch and A. ponsonbyi Suter. Maj. diam. 5 mm., min. diam., 4·5 mm., height 3 mm. Paterson Inlet, Stewart Island.”

Some details may be added to David's account. The protoconch is smooth in all the specimens seen Interstices between radial ribs with several minute radial threads crossed by very fine spirals. Radials numerous, between 105 and 135 on the body whorl. Umbilicus comparatively wide, between one-sixth and one seventh the major diameter. The form is therefore very close to A mossi, from which it may be distinguished only by the comparatively wide umbilicus. For this reason the writer advocates the use of stewartensis in a subspecific sense, mossi being the nominate form of the polytypic species concerned Allodiscus mossi mossi Murdoch ranges from Auckland south in the North Island, and has been recorded from as far south as Southland in the South Island.

Localities. Native Island, Paterson Inlet; subfossil in consolidated dunes, and living under scrub.

D. H. R. U. H.I. R.I. U.I.
A. mossi stewartensis (fide David) 5.0 3.0
Native Island, alive 4.9 29 105 0.72 60 22 14
Native Island, alive 5.0 3.2 119 0.81 65 24 16
Native Island, subfossil 5.0 3.1 135 0.81 64 27 16
Native Island, subfossil 4.4 2.7 111 0.72 61 24 16
A mossi mossi Muıdoch (cotype) 5.0 30 118 0.45 00 24 9

Flammulina zebra Le Guillou

The only record for the area is from Sealer's Bay, Codfish Island, R. K. D., 9/11/48. A widely distributed species in New Zealand, it has been recorded from many North and South island localities, south of Mt. Pirongia. The type locality is the Auckland Islands and the writer has collected it in Fiordland and on the Antipodes Islands It is therefore surprising that it has only been found on

– 142 –

an off shore island on the north-west coast of Stewart Island in this area. It is, however, always a very local species, and may yet prove to be more widely distributed on Stewart Island.

Flammulina perdita (Hutton)

This is the commonest Stewart Island Flammulina, although often local in occurrence. One of the most usual habitats is beneath leaves at the bases of tree trunks. The species is widely spread throughout New Zealand.

Localities : Leask's Bay, I. Worthy, —/11/47; Sailor's Rest, Paterson Inlet, E. N. Gardner, —/11/47; Mason's Bay, O. Allan, 0/9/48; Codfish Island, R. K. D., 7/11/48; Herekopere Island, O. Allan, —/5/51.

Flammulina feredayi (Suter)

Ferny Gully, Halfmoon Bay, O. Allan, —/8/49. This is another species of somewhat sporadic distribution throughout the southern part of the North Island and the South Island, which also occurs on Stewart Island.

Obanella allanae Dell.

This shell occurs not uncommonly around Halfmoon Bay. It has been found elsewhere at Dusky Sound and at Governor's Bush, Mt. Cook.

Flammoconcha stewartensis Dell.

This endemic species appears to be very rare, as the unique holotype from Ferny Gully, Halfmoon Bay is the only specimen known.

Ptychodon gadus n.sp. Plate 7, Figs. 11, 12.

Shell minute, depressed, umbilicate, finely radially striate. Sculpture consisting of fine, fairly closely spaced, somewhat protractive, radial riblets, 61 to 69 on the body whorl (69 on the holotype). Interstices with very fine radial threads crossed by dense microscopic spirals. Protoconch of 1 ½ whorls, sculptured except for the extreme apex with close-spaced radial ribs. Whorls 4 ½, including the protoconch. Spire slightly raised, suture somewhat impressed Aperture obliquely lunate, with 14 lamellae arranged as follows: Two teeth occur on the parietal wall, one about the middle emarginate at the top, the other, stouter, situated on the lower parietal wall. Two long, stout teeth arise from the columellar lip. Ten comparatively equidistant lamellae on the outer wall, the lower seven being alternately small and large. Umbilicus deep, cylindrical, about one-fifth the major diameter.

Locality. Sealer's Bay, Codfish Island, R. K. Dell, —/11/48. Holotype (M. 5709) and many paratypes (M. 5710) in the Dominion Museum.

P. gadus appears to be closest to P. microundulata (Suter) from which it may be readily distinguished by the narrower umbilicus, by the additional lower tooth on the parietal wall, the alternation in size of the teeth on the lower part of the outer lip and by the fewer ribs.

H. D U R H I U.I. R
Holotype 0.93 1.54 0.27 69 60 19 45
Paratype 0.96 1.5 0.29 63 64 19 42
Paratype 0.93 1.5 0.27 64 62 18 42
Paratype 0.90 1.5 0.28 61 60 18 41
Paratype 0.93 1.45 0.27 67 64 19 46
– 143 –
Ranges and Means of Indices
H.I 60–64 (62)
U.I. 18–19 (18.5)
R.I. 41–46 (43)

Ptychodon monoplax hiarara n. subsp. Plate 7, Fig. 9.

A series of shells from Codfish Island agrees closely with the type series of P. monoplax Suter from Bisbee Bay, Preservation Inlet, except that the axial ribs are much more numerous, 104 to 134 on the body whorl compared with 72 to 87 in monoplax. For this reason the Codfish specimens are here separated as a new subspecies. The name is derived from the Maori words for “many ribs”.

Major diameter, 2.35 mm., minimum diameter, 2.23 mm.; height. 1.32 mm.; major diameter of umbilicus, 0.82 mm.; riblets on body whorl, 123.

Holotype (M. 5711) and many paratypes (M. 5712) in Dominion Museum.

Locality. West Coast, Codfish Island, R. K. Dell, 6/11/48; consolidated sand dunes, Sealer's Bay, Codfish Island, R. K. Dell, 4/11/48 (subfossil).

Ranges and Means of Riblet Indices
Ptychodon monoplax monoplux Suter 34–37 (36)
P. monoplax hiarara Dell 45–54 (48)

Genus Charopa Albers, 1860

Die Heliceen, ed. 2, p. 87.

Type (o.d.) Helix coma Gray.

It is very doubtful if all the small species such as anguiculus (Reeve), montivaga Suter, benhami Suter, bianca Hutton, and montana Suter are congeneric with coma (Gray), the genotype of Charopa. Iredale's genus Discocharopa proposed for the Kermadec Island exquisita Iredale may be available for this group, but certainty on the matter must await more detailed comparisons and analysis These minute flattened. widely umbilicate shells with numerous riblets are a puzzling group. Suter's key (1913. p. 701) to these forms breaks down immediately, since all the species appear to have spirals in the interstices. Examination of the Suter Collection land snails soon shows that in two respects Suter's work was faulty. His measurements are in many cases wrong, and his descriptions of microscopic detail (e.g., the fine spirals in interstices and the fine sculpture on protoconchs) is often faulty. The microscopic equipment he used must have been inadequate. Preliminary examination of these small Charopid shells shows that wide variation exists although geographic series appear fairly constant. There are at the moment three possibilities. There may be very wide” variation, there may be many more species than have been recognised in the past, or there may be quite extensive geographic variation within the species. The Stewart Island fauna is difficult to determine, since the variation in the rest of New Zealand has not yet been thoroughly analysed and the number of specimens available from Stewart Island is small. Elucidation of the true relationships must wait until the mainland species are monographed.

Charopa bianca (Hutton).

Suter recorded this species from Halfmoon Bay and specimens from Codfish Island and Paterson Inlet agree comparatively well with Suter's specimen. As a group these Stewart Island forms do not, however, agree with series from nearby

– 144 –

mainland areas. Unfortunately Hutton's land snail types cannot be located with certainty in the Canterbury Museum and no topotypes are available to the writer With this proviso these Stewart Island specimens are best located in this species for the time being.

D H. U R
Codfish Island 1.98 0.93 0.61 106
Ferny Gully, Halfmoon Bay 1.77 0.93 0.64 99
Top of Paterson Inlet 1.93 0.98 0.52 104
Halfmoon Bay (Suter Coll.) 1.77 0.88 0.45 88
Ranges and Means of Indices
H.I. U.I R.I
Stewart Island 47–53 (51) 21–31 (29) 49–56 (53)
The Nuggets (Suter Coll) 46–53 (49) 24–28 (26) 43–48 (46)

Charopa sp.

Two specimens from Codfish Island, while agreeing in the main with the type series of Charopa anguiculus fncosa Suter (which may merit specific rank as the riblet counts are very different) have a much narrower umbilicus and are much higher.

D H U R
Sealer's Bay Codfish Island 1.72 1.0 0.5 60
Sealer's Bay Codfish Island 1.68 0.91 0.54 65
Ranges and Means of Indices
H I. U I R. I
Charopa sp. (Codfish) 54–58 (56) 29–32 (30.5) 35–38 (36.5)
Charopa anguiculus furosa Suter (Type series 25–26 (25) 48–52 (49) 32–35 (34)

Subgenus Pseudegestula nov.

Subgenotype (o.d.) Endodonta (Charopa) transenna Suter, 1904.

Iredale in 1915 supplied generic names for Suter's groupings of the smaller land snails without, however, examining specimens. Most of these groupings have been used equally uncritically since that time. Endodonta transenna has been accommodated in Egestula. This species is, however, much more closely related to the small Charopa species than it is to egesta Gray, the genotype of Egestula. Iredale in his treatment of the land mollusca of Lord Howe Island, Australia and Norfolk Island has proposed many new Charopid genera, and until the type species of these forms has been examined the differentiation of the smaller Charopid forms in New Zealand is difficult. The characters of transenna and number of related forms are so marked that they are here named. The major distinguishing features are the small size, flattened spire, riblets slightly retractive at the suture, and the presence of quite strong spiral sculpture on the base. This spiral sculpture is much finer than in Egestula, where the spirals are prominent raised ridges and are well developed on the upper surface. Protoconch appears smooth under normal magnifications, but minutely pitted under high powers. The habitat of the genus appears to be under dead bark.

This subgenus had also been separated out independently as new by Mr. A. E. Brookes, who had recognised the distinctness of the South Island form. Mr. Brookes has generously allowed the writer to name these forms.

– 145 –
Key to Species of Pseudegestula
1 Umbilicus one-third major diameter, body wlorl flattened, Riblet Index 32–56 transenna
(a) Riblet Index 41–56 (average 47) transenna transenna (Suter)
(b) Riblet Index 35–43 (average 36) transenna brookesi n. subsp.
Umbilicus one-fifth to one-quarter the major diameter, body whorl swollen, Riblet Index 66–72 (average 68) smithae n.sp

Charopa (Pseudegestula) transenna transenna (Suter)

The nominate form is known only from the Waitakerei Range, Auckland, at Henderson and Swanson. It may well be that it is much more widely spread but has not been collected. The umbilicus is very wide, about one-third the major diameter. The feature which distinguishes the nominate form from the South Island subspecies named below is the number of riblets, which, though variable, is always high (84 to 142) The spire is flattened, often level with the body whorl, although this feature is variable in both subspecies and accounts for the fairly wide range in the height/diameter ratios.

Charopa (Pseudegestula) transenna brookesi n. subsp.

This South Island subspecies is distinguished by the fewer ribs (81 to 108 in 9 specimens measured) giving a considerably lower Riblet Index. This difference is best shown in the table given below. This is a comparatively small difference, but the characters of these small Charopids seem remarkably constant, and taken in conjunction with the close geographical correlation, the difference seems significant. This form is known from several South Island localities, and is comparatively common wherever collected. It will probably be found to occur in intermediate localities as well.

Localities. Mt. Balloon, Mt Arthur tableland at 4,100 feet (type), R. K. Dell, 26/1/1948; several localities on the Mt. Arthur tableland from 3,600 to 4,000 feet; Bainham, A. C. O'Connor, —/11/46; foot of Sealey Range. Hooker Valley. A. E. Brookes; Notornis Valley, Te Anau, J. H. Sorensen, 11/1/1949.

Holotype (M. 4091) and 8 paratypes (M 5702) in the Dominion Museum.

D H U HI UI R R.I.
transenna transenna (type) 2.86 1.35 1.0 47 35 126 44
transenna transenna (type series) 2.04 0.77 0.73 37 36 84 41
transenna transenna 2.5 1.17 0.86 46 34 120 48
transenna transenna 2.54 1.18 0.81 46 32 142 56
transenna brookesi Mt. Balloon (Type) 2.77 1.18 1.09 43 38 102 36
Mt Balloon (paratype) 2.0 1.0 0.64 50 32 76 38
Mt Balloon (paratype) 2.73 1.32 1.04 48 38 109 39
Mt Balloon (paratype) 2.72 1.27 1.0 46 37 108 39
Mt Balloon 2.27 0.95 0.77 42 34 73 32
Bainham 2.11 0.91 0.73 43 34 81 38
Sealey Range 1.86 0.91 0.54 49 29 81 43
Sealey Range 2.64 1.32 0.92 50 35 99 38
Sealey Range 2.5 1.20 0.9 48 36 98 35
Notoinis Valley 2.68 100 37
Ranges and Means of Indices
Diam. H I. U.I. R I.
transenna transenna 2.04–2.86 37–47 (44) 32–36 (34) 41–56 (47)
transenna brookesi 1.86–2.77 42–50 (46) 29–38 (35) 32–43 (36)
– 146 –

Charopa (Pseudegestula) smithae n.sp. Plate 7, Fig. 4.

This species is best diagnosed in comparison with transenna. It does not appear to be as large as either form of transenna, the largest diameter seen in a shell of 4½ whorls being 1.77 mm. in comparison with 2.86 mm. for transenna transenna and 2.77 mm. for transenna brookesi. The umbilicus is considerably narrower in smithae, the Umbilical Index being 22 as compared with 34 and 35 for the two subspecies of transenna. The radial riblets are much more numerous than in transenna as he ranges and means show.

Holotype (M. 5703) and 3 paratypes (M. 5704) in Dominion Museum.

Locality. Top of Paterson Inlet, Stewart Island, O. Allan (holotype), —/5/50; Native Island, Paterson Inlet, alive, O. Allan, —/6/50.

D. H. U. H.I. U.I. R. R.I.
Holotype 1.77 10 0.40 56 22 119 67
Paratype 1.54 0.82 0.37 57 24 111 72
1.54 0.82 0.32 57 21 102 66

Pseudegestula smithae differs in so many details from the two subspecies of transenna that it is given full specific rank. It is named for Mrs. C. Smith, of Halfmoon Bay in appreciation of much assistance to the writer and of her efforts to make known the molluscan fauna of Stewart Island.

Ranges and Means of Indices
Diam. H.I. U.I. R.I.
P. Smithae. 1.54–1.77 mm. 56–57 (56.5) 21–24 (22) 66–72 (68)

Fectola reeftonensis Suter.

This widely spread species is well represented on Stewart Island, being recorded from numerous localities around Halfmoon Bay and Paterson Inlet and from Codfish Island.

Fectola roseveari Suter.

The same remarks apply to this species except that it was not collected from Codfish Island, although it was represented in a collection made from Herekopere Island by Miss O. Allan in May, 1951.

Fectola tapirina (Hutton)

Suter recorded this species from Halfmoon Bay, but the writer has seen it only from Codfish Island, where it is widespread, and also occurs in the consolidated dunes at Sealer's Bay.

Fectola (Subfectola) rakiura Powell.

This small white form may be easily overlooked, but it appears to be rather local.

Localities. The Neck, Paterson Inlet (type); Native Island, Paterson Inlet, alive, O. Allan, —/6/50; Herekopere Island, O. Allan, —/5/51.

Phrixgnathus liratulus Suter. Plate 7, fig. 6.

In the original description of this species the locality was given as “New Zealand. exact locality unknown Collected-by the late Mr. C. Traill.” The actual type series in the Suter Collection is labelled. “Stewart. Island”. The type series consists of three shells. Suter in his original description gave the measurements as 4.2 × 3 mm. The largest specimen is 4.18 × 3 mm. and this specimen is here selected as lectotype.

– 147 –

This species is the common keeled Phrixgnathus of Stewart Island. Powell recorded celia from two Halfmoon Bay localities, and Suter had specimens from the same locality. The writer cannot separate these latter specimens from young liratulus Suter. The relationships between liratulus and celia, has yet to be finally determined, but the solution of the problem must await the completion of a revision of the genus in New Zealand. The writer has seen no specimens referable to celia from Stewart Island.

In the Museum collections there are a few specimens of a keeled Phrixgnathus with a relatively high spire and very narrow umbilicus, but its status has not a yet been determined. Phrixgnathus liratulus is obviously the species recorded as Thalassohelix zelandiae (Gray) by David (1934).

Localities. Numerous localities around Halfmoon Bay; The Neck, Paterson Inlet, O. Allan, 8/5/49; Ulva Island, Paterson Inlet, I. Worthy, —/11/47; Native Island, subfossil, O. Allan, 7/1/48; Mason's Bay, A. W. B. Powell and O. Allan; Sealer's Bay, Codfish Island, R. K. Dell, 7/11/48.

The measurements of the type series and indices for a number of Stewari Island populations are given below.

D. H. S. U. H.I. S.I. U.I.
Lectotype 4.18 3.0 1.42 0.5 72 47 12
Syntype 3.77 2.36 0.91 0.5 63 38 13
3.91 2.82 1.09 0.54 72 38 14
Ranges and Means of Indices
H I. S.I U.I.
Type series (3 specimens) 63–72 (69) 38–47 (41) 12–14 (13)
Mason's Bay (3) 75–84 (81) 48–52 (49) 13–17 (15)
Raroa Reserve (2) 74–76 (75) 45 (45) 13 (13)
Sealer's Bay (5) 67–79 (74) 40–49 (45) 8–17 (13)

Phrixgnathus phrynia Hutton. Plate 7, Figs. 7, 8

Phrixgnathus phrynia Hutton, 1883. N. Z. Journ Sci., vol. 1. p. 476.

Phrixgnathus acanthinulops Suter, 1891. Trans. N.Z. Inst., vol. 23, p. 92.

Phrixgnathus phrynia var. major Suter. 1897. Proc. Mal. Soc., vol. 2, p. 259.

Suter described Phrixgnathus phrynia var. major in 1897 from Halfmoon Bay as being much larger than P. phrynia (diameter 3.5 as against 2.5 mm.) and with less distinct radial riblets. This form was listed as a variety by Powell (1939) but is omitted from his Checklist (1946). Examination of Suter's type series of major shows that these specimens are probably gerontic specimens of phrynia Hutton in which the outlines have become softened and the riblets obsolete. The normal form of the species occurs commonly on Stewart Island and agrees very well with South Island specimens of phrynia. There do not appear to be any shells in collections which are referable to var major apart from the type series although extensive collections of phrynia Hutton are available It must therefore be concluded that the sample upon which var major was based was taken from a population which exhibited gerontic features. The name major cannot be utilised in a subspecific sense as the Stewart Island populations do not appear to differ appreciably from South Island series.

Localities Many localities in the vicinity of Halfmoon Bay; top of Paterson Inlet, O. Allan, 8/5/50; The Neck, Paterson Inlet, O. Allan, —/5/49; Ulva Island. Paterson Inlet; Herekopere Island, O. Allan, —/5/51; Native Island. O. Allan. —/6/51.

– 148 –

The measurements and indices of a series of Phrixgnathus phrynia Hutton from Halfmoon Bay are given below:—

D. H. S. U. R. H.I. S.I. U.I. R.I.
2.91 1.82 0.73 0.62 36 62 40 21 12
2.68 1.77 0.77 0.59 worn 66 .43 22
2.68 1.95 0.91 0.51 38 73 46 19 14
2.77 1.86 0.82 0.57 38 67 44 21 14
2.73 1.77 0.82 0.62 33 65 46 23 12
Ranges and Means of Indices
H.I. S.I. U.I. R.I.
P. phrynia Hutton, Halfmoon Bay 62–73 (67) 40–46 (44) 19–23 (21) 12–14 (13)
P. phryna major, Lectotype 69 41 20
p. acanthinulops Suter, Type 68 48 19 13

Phrixgnathus regularis Pfeiffer.

This species which is widely distributed in the North and South Islands, is represented in collections from Halfmoon Bay, Native Island and Codfish Island.

Phrixgnathus flemingi stewartensis n. subsp. Plate 7, Fig. 5.

A series of Phrixgnathus from Stewart Island prove to be subspecifically separable from the Snares Island P. flemingi Dell. In the original description of flemingi (Dell, 1950) the dimensions of the holotype as stated were incorrect. The correct measurements are given below. The Stewart Island form proves to be constantly higher in comparison with the diameter though the spire heights are much the same. For this reason the Stewart Island shell is separated as P. flemingi stewartensis. The Stewart Island form has a stronger colour pattern, though this is not a good diagnostic feature. The height differences are best shown in the accompanying table of measurements and indices. Holotype (M. 571 3) and many paratypes (M. 5714) in the Dominion Museum.

Localities. Sealer's Bay, Codfish Island, R. K. Dell. 7/11/48, (type locality); West cost of Codfish Island (very large specimens); top of Goat Island, Mason's Bay, O. Allan, —/3/50; Native Island, Paterson Inlet, J. H. Sorensen, 17/7/39 (O'Connor Collection); Tomihau Island, Paterson Inlet, J. H. Sorensen, 15/4/41. This species is also recorded from Solander Island later in this paper.

D. H. S. U. H.I. S.I. U.I.
P. flemingi Dell
Holotype 3.24 2.04 0.77 0.64 61 38 19
Paratype 3.73 1.91 0.73 0.73 51 37 19
P. flemingi stewartensis Dell
Holotype 3.18 2.05 0.86 0.64 63 42 20
Paratype 3.32 2.0 0.73 0.59 60 37 15
Paratype 2.95 1.86 0.68 0.54 63 37 18
Paratype 3.35 2.27 1.0 0.70 68 44 21
3.0 1.86 0.68 0.45 62 36 15
Ranges and Means for Populations of flemingi
H.I. S.I. U.I.
P. flemingi flemingi
Snares Islands 51–61 (56) 37–38 (37.5) 19 (19)
P. flemingi stewartensis
Goat Island (Type series) 60–68 (63) 36–44 (39) 15–21 (18)
West coast, Codfish 57–69 (63) 32–54 (44) 17–22 (19)
Sealer's Bay, Codfish 56–67 (63) 34–40 (38) 14–21 (17)
– 149 –

Phrixgnathus rakiura n. sp. Plate 7, Fig. 1.

Shell small (up to 2.5 mm. major diameter), depressed-turbinate, umbilicus comparatively wide, thin, shining. Spire low (spire Index 31–47), suture impressed. Whorls 4 ½, including a protoconch of 1 ½ whorls. Protoconch comparatively large, last half whorl with very fine spirals which appear to consist of spiral rows of microscopic raised pustules. Post-nuclear sculpture consisting of fine growth lines with a few widely spaced irregular growth wrinkles. Whole surface covered with a microscopic reticulation. Umbilicus deep, comparatively wide (Umbilical Index 21–25). Aperture obliquely lunate, with thin outer lip.

Holotype (M. 5707) and numerous paratypes (M. 5708) in Dominion Museum collection.

Major diameter. 2.27 mm.; height. 1.23 mm.; diameter of umbilicus, 0.64 mm.; height of spire. 0.45 mm. (holotype).

Localities Native Island, Paterson Inlet, under scrub (type locality), also subfossil; many localities around Halfmoon Bay and Paterson Inlet; Codfish Island, living and subfossil.

This small, featureless shell appears to belong to a group of shells with a specifically southern distribution. They have much the facies of a Paralaoma in the depressed-turbinate outline, the comparatively wide umbilicus and the type of protoconch but are consistently larger and lack the well marked axial sculpture. Anatomical examination may show their true relationship. and they may well be related to such mainland forms as P. viridula. Other members of the group are P. hamiltoni Suter from Macquarie Island, a form allied to rakiura from Solander Isand (described later in this paper) and apparently P. campbellica (Filhol) from Campbell Island. This latter species has never been figured, and the original description was somewhat inadequate. The only form known from Campbell Island referable to this species is figured here (Fig. 3) for comparison with rakiura. and the basic measurements and indices are given below. It differs from rakiura in the much finer growth lines, smoother shell, more definite growth wrinkles, a much lower spire and a narrower umbilicus. It agrees in general outline and appearance and the microscopic reticulation, and is undoubtedly closely allied. The group may well be separated eventually from Phrixgnathus.

D. H. S. U. H.I. S.I. U.I.
W. coast, Codfish Island 2.41 1.45 0.64 0.5 60 44 21
W. coast, Codfish Island 2.18 1.25 0.45 0.45 57 36 21
Native Island 2.29 1.27 0.43 0.5 55 34 22
Native Island 2.32 1.27 0.45 0.59 55 35 25
Native Island 2.27 1.41 0.54 0.54 62 38 24
Native Island 1.91 1.11 0.37 0.41 58 33 21
Native Island 2.23 1.14 0.36 0.5 51 31 22
Native Island 2.36 1.5 0.64 0.5 63 42 21
Halfmoon Bay 2.04 1.16 0.41 0.45 57 35 22
Halfmoon Bay 2.5 1.64 0.77 0.59 65 47 24
Ranges and Means of Indices
H.I. S.I. U.I.
P. rakiura n. sp. 51–65 (58) 31–47 (37) 21–25 (22)
P. campbellica 53–62 (56) 22–28 (25) 15–18 (16)
– 150 –

Genus Paralaoma Iredale

The distribution of the species of this genus in New Zealand is poorly known as they have not been extensively collected in the past although they are very common in leaf-mould. Two species which occur practically throughout New Zealand have been collected from Stewart Island.

Paralaoma lateumbilicata (Suter, 1890)

Numerous localities in Halfmoon Bay and Paterson Inlet; Native Island, sub-fossil, O. Allan, 7/1/48; Sealer's Bay, Codfish Island, R. K. D., 7/11/48.

Paralaoma sericata (Suter, 1890)

Numerous localities in Halfmoon Bay.

Rhytida australis Hutton

There is considerable variation in size amongst living populations. On the whole living shells are smaller than subfossil specimens. The colour pattern also varies from light to dark brown. The Codfish Island specimens lack the well-defined umbilical dark patches, and some of the shells are much yellower. Apart from this there seems to be no major consistent differences between any of the existing populations nor between living and fossil shells.

Localities. Subfossil: Native Island, Paterson Inlet. Living: Numerous localities around Halfmoon and Horseshoe Bays; Bravo Island, Paterson Inlet, S. Traill, —/12/45; Ulva Island, Paterson Inlet, S. Traill, 1943; Mason's Bay, S. Traill; Freshwater River, Paterson Inlet, S. Traill,—/6/42; Sealer's Bay, Codfish Island, R. K. Dell, 9/11/48.

Ranges and Means of Measurement
Diameter Height
Native Island (subfossil) 14.4–17 (15.6) mm. 8.9–10.7 (9.3) mm.
Sealer's Bay, Codfish 11.0–13.3 (12.3) mm. 6.3–7.6 (68) mm.
Ferny Gully, Halfmoon Bay 13.1–15.0 (138) mm. 7.4–8.7 (81) mm.

Records thus exist for the coastal areas north of a line joining the southern portion of Paterson Inlet and Mason's Bay. There do not appear to be any records for the southern part of the Island.

Dentition. Formula 14.1.14. Central small, about half the length of inner lateral. Laterals increasing gradually in size to the 12th, which is very large. The 13th lateral is very long but very thin and needle-like, the outermost very small and vestigial. The largest lateral bears a very slight longitudinal ridge, much weaker than that seen in members of the patula series such as meesoni perampla Powell. In comparing the radulae of species of Rhytida the number of small, vestigial outer laterals should be indicated separately from the main series. The known radular formulae for the New Zealand species of Rhytida are as follows:—

Group A. patula Hutton 18.1. (17.1); citrina Hutton 16.1. (15.1), meesoni perampla Powell 14.1.14, specimen from Mt. Arthur 15.1. (14.1), otagoenis Powell 14.1.14; hadfieldi Powell 16.1. (15.1) and 18.1. (17.1).

Group B. greenwoodi greenwoodi (Gray) 12.1. (10.2) and 11.1. (10.1); greenwoodi webbi Powell 11.1. (10.1); stephenensis Powell 13.1. (12.1) and 10.1 (9.1); dunniae (Gray) 18.1. (13.5); duplicata vivens Powell 15.1. (11.4), pycrofti Powell 13.1. (10.3); tarangaensis Powell 15.1. (11.4).

– 151 –
Picture icon

Text-Fig. 1.—Rhytida australis Hutton. A. Radula; B, C, D, E, subfossil shells, Native Island; F, Recent shell, Halfmoon Bay.

Group C. australis Hutton 14.1. (12.2).

The figures in parentheses show the division between primary and vestigial laterals. The species in group A have a strong longitudinal ridge along the largest lateral, those in group B lack this ridge. As has been mentioned, australis has a rudimentary ridge which links it to the patula series. The small number of major laterals (12) links it more to the greenwoodi series (9–13) than to the patula series (14–17). In these respects australis is intermediate between the two groups, although the evidence linking it to the patula series is slight. Rhytida australis Hutton appears to have been isolated for a considerable time, at least in comparison with the life of the genus in New Zealand. The nearest species geographically is Rhytida otagoensis Powell, which has a well-marked ridge on the largest lateral tooth, 14 laterals with no vestigial outer series and no central, and is thus well-differentiated morphologically.

Atharocophorus cf. bitentaculatus (Q. and G.)

Four specimens from Raroa Reserve and one from Codfish appear to belong to Atharocophorus and are close to bitentaculatus (Q. & G.). In our present state of knowledge of these sings accurate identification is impossible.

Discussion

The land snail fauna of the Stewart Island area, including Codfish and the other outlying islands, except for the Solanders, now numbers 33 species. Of these, eleven species are endemic to the area (this includes Phrixgnathus flemingi stewartensis, which also occurs on Solander Island). That one-third of the land mollusca should be endemic is all the more surprising when one considers that Stewart Island is connected to the mainland by a broad shelf nowhere covered by a depth of more than 20 fathoms (N.Z. Chart No. 14, 1952), and under 15 miles across. It is to be presumed that it would be joined to the mainland during

– 152 –

glacial periods in the Pleistocene and isolated again during inter-glacial rises in sea-level. Since glacial conditions did not descend to sea-level in this area, a wide pathway suitable for colonisation would have existed. Willett (1951, p. 33. fig. 6) includes Stewart Island in the periglacial zone in the Pleistocene. He (1951, p. 31) defines this zone, after Zeuner, as “the zone surrounding an ice-sheet, in which the cooling effect of the ice, produced a frost climate, and the zone in which the climate favoured permanent frozen subsoil”. Climatically it would coincide with the area above the upper limits of the bush-line on mountains to-day. Judging by what is known of the temperature tolerance of New Zealand land mollusca to-day such conditions, if general over the island, would have eliminated most, but not all of the molluscan fauna of the island. Several species of Phrixgnathus occur commonly above the bush-line, and in the writer's experience some Rhytida species (e.g., R. meesoni perampla Powell on Mt. Arthur) and small Flammulina (such as F. feredayi Suter) will also survive in appreciable numbers above the bush-line. It is therefore feasible to expect that the Rhytida. most of the Phrixgnathus species and Flammulina feredayi or their ancestors could have survived the glacial periods on Stewart Island. That is if Willett's picture of all-over tundra conditions on Stewart Island is correct and no pockets of bush cover remained. In view of the strongly differentiated nature of the Rhytida and the presence of three endemic species of Phrixgnathus it is highly probable that these four forms or their ancestors were already present on Stewart Island at the onset of glacial conditions. In view of the known distribution of Flammulina feredayi and the fact that it has not differentiated on Stewart Island, it seems probable that it represents a much later faunal invasion. From the fact that the Phelussa has no close relations and that the other large forms of the genus in the south of the South Island, P. henryi and P. costata have a scattered peripheral distribution, it may be presumed to be an ancient form that has achieved its morphological peculiarities through long isolation. The Flammoconcha is another strongly differentiated form. The only other species in the genus, F. cumberi (Powell) has been collected by the writer from patches of stunted beech above the main bush-line on Mt. Arthur tableland. It may be presumed, therefore, to be tolerant of low temperatures so that Flammoconcha stewartensis. as well as Phelussa fulminata may be supposed to be old elements in the Stewart Island fauna—i.e., they may represent the pre-glacial elements in the fauna.

Pseudegestula smithae and Subfectola rakiura are obviously related to South Island forms but have not changed appreciably on Stewart Island. Allodiscus mossi stewartensis and Ptychodon monoplax hiarara are allopatric forms of South Island species showing less appreciable differences.

A considerable number of forms are found on Stewart Island and in the South Island without appreciable differentiation—e.g., Murdochia chiltoni, Therasia thaisa, Thermia cressida, Allodiscus planulatus, Flammulina zebra, F. perdita, F. feredayi, Obanella allanae, Charopa bianca, Fectola reeftonensis, F. tapirina, F. roseveari, Phrixgnathus phrynia, P. regularis, Paralaoma lateumbilicata and P. sericata. On the grounds of morphological relationship, geographical distribution and degree of differentiation a basic, old endemic element and three progressive faunal additions can be postulated. It is conceivable that these three additions corresponded with three glacial periods.

– 153 –

An interesting group comprising Phrixgnathus flemingi stewartensis and P. rakiura shows subantarctic affinity though in this case the “subantarctic” forms may well have been derived originally from the New Zealand mainland. A species with an anomalous distribution pattern is Phenacohelix pilula.

The outline postulated above would not have been necessarily modified if the climatic conditions were not as severe as discussed above. Willett (1951, p. 31) suggests, “It would seem possible for some of the marginal forest of the greater than 30 to 35 inch rainfall zone to have survived up to 500 feet above sea level in such areas as Stewart Island, south-east Otago, north-west Nelson and north Westland. These survivals could act as centres of colonization during interglacial and post-Pleistocene times.” This would allow a greater chance for the old endemics to survive on Stewart Island throughout the Pleistocene glaciation. In any case it is doubtful if on Stewart Island, the periglacial tundra zone was bounded below by a loess zone with a dry, cold steppe climate as Willett postulates for the east coast of the South Island, It is more probable that the comparatively slight development of an ice sheet and the ameliorating influence of the surrounding sea would allow scrub and forest to remain around the periphery of the periglacial tundra zone.

Consideration of the internal distribution of the land mollusca within the Stewart Island area brings out a number of differences. Unfortunately data is available for only two of the off-shore islands, Codfish and Herekopere. Codfish is comparatively large (3 ½ miles long), is little affected by sea birds, has a varied plant cover including a considerable area of rimu-kamahi-rata forest, and is separated from Stewart Island by about 2 miles of sea. Herekopere Island is smaller (½ mile long) and has a dense Olearia scrub cover up to 20 or 25 feet high over much of its area. It is separated by about 3½ miles of sea from Stewart Island itself, and is considerably affected by the large numbers of petrels that nest there. The total number of species, the number of Stewart Island endemics, the number of local endemics, and the number of species present but not occurring elsewhere in the Stewart Island area are listed below for these throe localities:

Stewart Island
Mainland Codfish Herekopere
Total number of species 26 19 6
No. of Stewart Island area endemics 9 4 1
No. of local endemics 4 2
Other species not found elsewhere in Stewait Island area 7 2

The numbers in all headings decreases with a decrease in area, and this apparent correlation may well be an actual one. The fact that Codfish should have two locally endemic forms and should also have two other species not found elsewhere in the Stewart Island area is of interest as an indication that some at least of the off-shore islands have had a different faunal history, at least in part.

Land molluscs have been collected from four consolidated dune systems on Stewart Island, Native Island, and the Neck in Paterson Inlet, Mason's Bay and Sealer's Bay on Codfish Island. One species is found in all four dune areas—i.e; Therasia thaisa, and this is the only species recorded from the Neck and Mason's Bay. Additional species may well be found at both localities. Nine species have been recorded from Native Island and five from Codfish.

– 154 –

Five of the species from Native Island are Stewart Island endemics, and two of them, Rhytida australis and Phelussa fulminata are much commoner as sub-fossils than they have proved to be as living shells. Two of the Codfish Island species are Stewart Island endemics, and one of these is a locally endemic form. Therasia and Thalassohelix are well represented in almost all the dune faunas and representatives of these two genera are generally found in dune deposits throughout New Zealand. In many ways both appear to be very tolerant of changing conditions, and representatives of both genera elsewhere in the New Zealand area are often the last elements of the local land snail fauna to persist when forest is cleared.

II. Solander Island

No land snails have hitherto been recorded from the isolated Solanders, to the west of Foveaux Strait. From leaf-mould collected by Dr. R. A. Falla and Mr. C. L. Lindsay, of the Dominion Museum on the 9th December, 1947, and on 20th July, 1948, three species have been sorted out.

Phelussa costata (Suter)

One dead but complete specimen and one broken fragment were obtained. Suter's type series from Resolution Island are all broken shells and only two other complete specimens have been traced; a single specimen from the Nuggets in the Nuggets Collection and the Solander Island shell. These specimens agree in all particulars except that the umbilicus is considerably wider in the Solander and the Nuggets specimens. These differences are best seen in a comparison of the indices. Phelussa henryi (Suter) from Resolution Island has a very much lower riblet count.

Measurements and Indices for Phelussa costata (Suter)
D. H. S. U. R. H.I. S.I. U.I. R.I.
Solander Island 8.64 5.0 1.73 1.36 117 58 35 16 13
Holotype 10.7 0.55 130 5 12
Paratype 12.0 0.45 143 4 12
Paratype 10.9 0.64 6
Nuggets 9.2 5.27 1.0 1.0 122 57 19 11 13
Ranges and Means of Indices
H.I. S.I. U.I. R.I.
P, costata Type series 4–6 (5) 12–13 (12.5)
P. costata (Nuggets and Solander) 57–58 (57.5) 19–35 (27) 11–16 (13.5) 12–13 (12.5)
P. henryi 55–57 (56) 23–24 (23.5) 16 (16) 7 (7)

The small number of specimens involved, the occurrence of costata and henryi in the fame locality and the occurrence of two recognizable forms of costata presents a taxonomic puzzle. Until larger series are available the Solander Island and Nuggets specimens are designated P. costata, type A. If the wide umbilicus of the type A series is found to be constant it will probably merit subspecific separation. The two forms of costata have much the same Riblet Index and apparently similar overall dimensions but differ in the diameter of the umbilicus (which is wider in the Solander shell than in the specimen from the Nuggets). P. henryi Suter has much the same dimensions as costata, has the wide umbilicus (the same as the Salander shell) and a much lower Riblet Index. There is thus a series of

– 155 –

independently varying characters involved and larger series will show how valid the above trends may be. Unfortunately the plant cover of Resolution Island has been considerably altered, and Phelussa may no longer occur there.

Phrixgnathus rakiura solanderi n.subsp. Plate 7, Fig. 2.

A long series from leaf-mould from Big Solander Island shows close relationship with P. rakiura, described earlier in this paper. The Solander shells agree with Stewart Island populations except that the shell is higher in relation to diameter, has a comparatively higher spire, and a narrower umbilicus. These characters appear very constant and, particularly as regards Height Index and Umbilical Index, show little overlap. For this reason the Solander series is separated as a subspecies of rakiura.

Holotype (M. 5705) and many paratypes (M. 5706) in the Dominion Museum.

Diameter, 2.36 mm.; height, 1.45 mm.; height of spire, 0.64 mm.; diameter of umbilicus, 0.41 mm. (holotype).

Locality. Big Solander Island, Foveaux Strait, common in leaf-mould collected by R. A. Falla, 9/12/47.

D. H. S. U. H.I. S.I. U.I
2.36 1.54 0.68 0.5 65 44 21
2.27 1.56 0.73 0.41 68 46 18
2.32 1.5 0.64 0.45 64 42 19
2.04 1.41 0.64 0.41 69 45 20
2.23 1.56 0.68 0.45 70 43 20
Ranges and Means of Indices
H.I. S.I. U.I.
64–70 (67) 42–46 (44) 18–21 (19.5)

Phrixgnathus flemingi stewartensis Dell

A single specimen of a Phrixgnathus is indistinguishable from long series of this subspecies described earlier in this paper from Stewart Island.

Discussion

The land snail faunas of small, isolated off-shore islands are always of interest. The number of species involved is often small, but they are often well represented numerically. The land snail fauna of the Solander Islands conforms in part to this pattern. Three species have been collected, and the relationship of these three species and the degree of differentiation present in each varies considerably. The largest shell, a representative of the widely, though sporadically distributed Phelussa costata-henryi complex seems indistinguishable from a specimen from the Nuggets. Phrixgnathus flemingi stewartensis, of which one specimen was obtained, is the same as a comparatively common Stewart Island shell. Phrixgnathus rakiura solanderi is an allopatric form of a Stewart Island species, is comparatively small, and appears to be very common.

An examination of the land snail faunas of isolated islands in the southern part of the New Zealand area—e.g., the Snares, the Antipodes, and the Solanders, has influenced the writer in formulating some general conclusions on the distribution of small land snails. Firstly, in spite of their isolation, the islands have land snails. This is brought home particularly in the case of the Antipodes, where all geological and faunal evidence indicate complete isolation with no

– 156 –

recent land connections. Secondly, these faunas have a general similarity in the types of mollusca present. Phrixgnathus species predominate, though Flammulina may also be present. Unless one is to believe in spontaneous generation of similar or even the same types in various isolated areas one must conclude that these small land snails have some efficient dispersal method, probably quite passive, by which they achieve their distribution pattern. In the case of the southern islands the temptation is very strong to postulate that sea birds are the active agent. The incredible numbers of petrels, albatrosses and penguins that utilize the southern islands as nesting and in some cases moulting grounds profoundly influence the local environments. Without being specific as regards method, it may be postulated that sea birds have assisted the small land snails to achieve their present distribution in the southern islands. In any case it would appear that these small land snails, particularly those living in leaf-mould and in low vegetation, cannot be used in an exact way in biogeography, especially in southern regions.

In the case of the Solanders, the two species of Phrixgnathus fall into this category and should be classed as accidental migrants. The larger Phelussa cannot fall into this category, and its presence on the Solanders is difficult to explain. The present peripheral distribution of the Phelussa costata-henryi complex is probably due in very large part to glacial control. Providing a land connection to the Solanders could have existed during glacial periods there is no difficulty in explaining the presence of a Phelussa on the Solanders. It is difficult to see why the Solander Phelussa has not become more appreciably differentiated from the Nuggets population since then. In any case the Stewart Island Phelussa fulminata must have been isolated at a much earlier geological time than the Solander Island P. costata.

Literature Cited

David, Von L., 1934. Zoologische Ergebnisse der Reissen von Dr. Kohl-Laisen nach den Subantarktischen Inseln bei Neuseeland und nach Sudgeorgien. 9. Senckenbergiana. Bd. 16, p, 126–37.

Dell, R. K., 1952 A. Otoconcha and Its Allies in New Zealand. Dom. Mus. Rec. Zool., vol. 1, no. 7, pp. 59–69.

—— 1952 B. New Species and Genera of New Zealand Land Snails with a Revision of the genus cavellia. Dom. Mus. Rec. Zool., vol. 1, no. 9, pp. 87–97.

Powell, A. W. B., 1939. The Mollusca of Stewart Island. Rec. Auck. Inst. Mus, vol. 2, pp. 211–238.

Suter, H., 1897. The Land Mollusca of Stewart Island. Proc. Mal. Soc., vol. 2, pp. 258–9.

Willett, R. W., 1951. The New Zealand Pleistocene Snow Line, Climatic Conditions, and Suggested Biological Effects. N.Z. Journ. Sci. Tech., Ser. B, vol. 32, pp. 18–48.