The Structure and Adaptations of the New Zealand Vermetidae
Part I. the Genus Serpulorbis

By J. E. Morton,
Department of Zoology, Auckland University College

attached, at first coiled like Turrietella, later irregularly extended.” Moreover, there is an “operculum as large as the aperture.” Serpulorbis is distinguished in being firmly attached along one side to the substratum, and irregularly uncoiled from the outset. The operculum is always absent as in sipho.

Taxonomy

sipho. As regards the juvenile shell just after attachment, South Australian specimens identified by Mr. B. C. Cotton as belonging to sipho are quite dissimilar in sculpture to young Serpulorbis at the same stage from New Zealand localities. Fig. 25 illustrates a juvenile shell dredged off Oamaru (Finlay Collection). In the adult the sculpture of the young shell is usually obscured by the superposition of the adult coils, and it was not possible to assign the dredged material conclusively to its species. Finally, the coloration of the animal of sipho, illustrated by Quoy and Gaimard (1834) under the name Vermetus arenarius, is again unlike that of either of the New Zealand animals studied. It is therefore proposed to designate the Auckland species of Serpulorbis as follows:

1. Serpulorbis zelandicus (Quoy and Gaimard) 1834

The first of the Milford species is clearly entitled to the original name proposed by Quoy and Gaimard for their neozelanic vermetid. The authors gave no description of the shell characters, merely remarking upon what has been pointed out above—the similarity of the shell to related species of Australian vermetids, and proceeding to describe the coloration of the animal. “Ce vermet a tellement de rapports avec V. elegans, que c'est avec doute que nous en faisons une éspèce particulière. On ne peut qu'indiquer ses couleurs. Těte jaunâtre en arrière, brun et ponctuée de rouge en avant. Le pied est seulement jaunâtre avec des taches rouges. Le manteau est largement bordé d'un orange vif. Le tube, contourné sur lui-même, ne nous a point offert de caractère appréciable sur le moment.” Reference to Quoy and Gaimard's atlas of zoological illustrations shows that the two coloured drawings of Vermetus zelandicus quite adequately identify the animal with one of the Milford species under consideration. The use of the term zelandicus was first confused by Suter, who quite unwarrantably annexed to Quoy and Gaimard's colour description an account of the shell of Hutton's Siphonium lamellosum.

Finlay (1930) thereupon rightly considered Hutton's shell to be identical with that described by Suter under Serpulorbis zelandicus, and was led to accord the latter name priority over lamellosum for Hutton's Siphonium. Serpulorbis zelandicus, however, applies validly to a separate shell, prior to and quite distinct from Hutton's species, and a fuller description is now provided.

The shell is moderately large and vermiform, sub-solitary or in small groups of two or three intertwined, seldom forming larger aggregates. The coiling is completely untwisted and the disposition of the whorls irregular or in two or three loosely coiled convolutions attached along the whole of one surface to the substratum, save for the apertural portion which is generally vertical so that the opening faces directly upwards. The aperture is circular in section, the attached sides of the tube flattened or irregularly moulded to the substratum, and the exposed surface regularly convex. The diameter regularly increases, reaching 8–10 mm, across the aperture of a large shell. The sculpture of the free surface is predominantly of longitudinal ridges of somewhat unequal size, consisting of several more prominent cords, separated by three to seven rather smaller riblets. The longitudinal sculpture is crossed at close intervals by small sharp rugae sometimes giving a finely scaled appearance to the living shell, and in the beach worn

blunt denticles laterally. The two pairs of marginals are stout, curved and falciform with small blunt denticulations along the convex edge.

Slide of radula and colour records of living animals are deposited in the Auckland Museum.

2. Serpulorbis aotearoicus sp. nov.

This species resembles zelandicus in adult shell characters, tending to a darker brown or purplish-brown colour, but without apparent difference in sculpture. The coloration of the animal is, however, quite distinctive. The predominating colour of the head, proboscis and foot is black, or at times a lighter shade of greyish, lightening to pinkish-brown further back on the trunk and proximal parts of the foot. The tip of the proboscis is always deep black, and along the margin of the cephalic shield is a bright yellow or yellowish-cream line, passing backwards from the base of the short cephalic tentacle, which is itself yellow. A margin of the terminal pedal disc forms a broad band of yellow, and the black pedal disc itself is closely flecked or striated with yellowish. The pedal tentacles are mottled with black and yellow. The mantle is light brown in colour, its border edged widely with bright orange-red. The viscera are pure white, the digestive gland black. The opereulum is absent in the adult.

The radula (Text-Fig. 2) is distinguished from that of zelandicus by the narrowness of the central and lateral teeth, which are slightly greater in length than in breadth. The principal cusp of the central tooth is much smaller than in zelandicus, usually not more than a quarter the length of the tooth. The basal wings of the central tooth are shortly truncated, and the lateral teeth set closer to the mid-line. Their mesial cusps are also much smaller than in zelandicus.

The embryo shell of aotearoicus is illustrated in Figures 20–24. The nucleus is of one and a half whorls, with the apex a transparent spherical bulb, soon developing a finely lirate longitudinal sculpture, crossed by very delicate accremental striae. The lirate portion of the nucleus gives place abruptly to a second whorl which is without longitudinal sculpture, though marked with fine, distinct growth striae, and tinted with several longitudinal splashes of reddish-brown. The aperture at the one and a half whorled stage is approximately triangular, slightly notched or effused at the inner lip. The shell is at this stage imperforate, but the coiling early begins to loosen up, and the earliest part of the attached shell has the appearance of Fig. 25, somewhat resembling a worm tube of Spirorbis. The embryo at the free-moving two-whorled stage has a large chitinous operculum, concave or saucer-shaped, with the margin simple.

Type in the Auckland Museum. Slide of radula and colour record of living animal are deposited with the type. Paratypes in the Zoology Museum, Auckland University College.

Localities. Milford Reef, Rangitoto Channel (type)—under boulders at low spring tide, on outer fringe of reef, together with smaller numbers of zelandicus; Otata Island, Noises Group, on exposed rocks, covered with Melobesia at low spring tide.

Complete correlation was found in radula and animal coloration for 50 specimens of aotearoicus and 16 specimens of zelandicus examined from Milford. The two species occur intermixed in the same locality,

Genus Novastoa

Externals and Pallial Organs

When the animal is fully extended, the margin of the mantle forms a circular rim reflected over the edge of the shell, and the head and foot can be moved actively about, with the pair of long pedal tentacles directed upwards. The foot (Figs. 1, 2) is truncated and plug-like, its circular disc almost completely occupying the aperture. Lacking an operculum, the animal is able to withdraw far into the shell tube for protection by contraction of the slender columellar muscle. The edge of the foot is usually indented on either side by a small channelled lip, serving for the passage of the inhalant pallial current on the left and the exhalant on the right. This probably gave rise to the statement by Quoy and Gaimard, repeated by Suter, that the foot is cruciform. Dorsally to the pedal disc, the aperture is occupied by the flattened, broadly ovoid cephalic shield, terminating in a short, wide proboscis (Fig. 1) with the slit-shaped mouth at the tip. The cephalic tentacles (Fig. 2, TE.) are short and blunt, with minute eyes at their outer bases.

The organs of the pallial cavity (Fig. 4) fairly closely resemble those of “Vermetus” novae-hollandiae, described by Yonge (1932) and Yonge and Iles (1939); the animal is without doubt predominantly a ciliary feeder, employing the ctenidium for the collection of food particles—principally benthic diatoms and fine detritus of plant origin. The ctenidium (Figs. 1, 4, CT.) has about the same extent as in “Vermetus” novae-hollandiae and is thus much better developed than in Serpulorbis gigas, which feeds by the mechanism of mucus strings. The mantle margin forms an entire skirt in the male without trace of inhalent siphonal process; the female shares with other species of “Vermetus” and applied groups the longitudinal slit along the dorsal mid-line of the mantle along which the row of egg capsules is attached to the inner surface of the shell. The ctenidium lies to the left of the mantle cavity, and to the right of the slit in the female lies the hypobranchial gland (Fig. 4, HY.G.), which extends across the pallial (ventral) aspect of the rectum. The gill filaments (Figs. 7, 8) are typically triangular, with the frontal side supported by strong skeletal rods. These form a framework extending from the axis across the ventral aspect of the ctenidium, from which the wide respiratory lamellae pass dorsally to the mantle wall. According to Yonge's figure (1932) the filaments in “Vermetus” novae-hollandiae are much narrower and more elongate. As Yonge points out (1938) the filament width tends to be reduced by shrinkage of the respiratory area in the preserved material figured. This fact, however, does not appear to account sufficiently for the difference; in Serpulorbis zelandicus fixation did not cause appreciable shrinkage, and the filaments appear rather thick and non-membranous. No doubt, in “Vermetus” novae-hollandiae, the narrowing of the filaments provides a further adaptation to ciliary feeding, by which this species obtains the whole of its nutriment. In Serpulorbis zelandicus, on the other hand, it is clear that ciliary feeding may occur with retention of the primitive shape of the filaments, though in most ciliary feeding prosobranchs the filaments are either wholly linear (Crepidula, Orton, 1912; Stephopoma, present writer, in manuscript) or partially rod-like (Turritella, Graham, 1939; Struthiolaria, Morton, 1951). Conversely, in such genera as Strombus and Xenophora (Morton, 1949) the more general

The Foot and Pedal Gland (Figs. 3, 5, 6)

The food tract is continuous in front with the right side of the foot, and the lateral tracts of the foot on either side converge immediately below the proboscis. At this point the paired pedal tentacles arise close to the midline, and between them is situated the opening of the duct of the pedal mucus gland. In front of the gland opening is located a small triangular area of glandular epithelium, representing the original sole, or plantar surface of the foot (Fig. 3, G.FT.).

Text Fig. 3—Serpulorbis zelandicus. Diagrams showing the structure of the foot on emergence of the embryo from the capsule (A) and its modifications in the adult (B). F.TR, ciliated and glandular food tract; GL.FT, glandular sole or plantar surface of foot; OP, operculum; PD.G, opening of pedal gland; PD.T, pedal tentacle.

An understanding of the modifications that have taken place in the foot may be gained from Text-figure 3 and from Figs. 20–22 of developing embryos. In the capsule veliger (Fig. 21) the foot is triangular, with a transverse opercular rudiment upon its posterior surface and a bifid tubercle at the anterior edge, giving rise to the pedal tentacles. In the embryo which has just emerged from the capsule (Fig. 22) the velum is lost, and the foot has the same relations as in a typical adult free-moving prosobranch. The plantar surface is wide, enabling the animal to creep about, and the operculum is very prominent, forming a large saucer-like structure (Fig. 22 op.) somewhat overlapping the edges of the foot. Its concave shape is strongly reminiscent of the adult operculum retained throughout life in “Vermetus” novae-hollandiae. The anterior edge of the foot is bluntly rounded, with strong cilia, and in front of the foot is a pair of well-developed pedal tentacles, rugose and finely ciliated. The

The highly developed pedal gland is evidently not employed solely in connection with ciliary food collecting. In Serpulorbis gigas Boettger (1930) has shown that food is entrapped by means of long extruded mucus strings, formed by the pedal gland. Three or four strings are moved gently to and fro, and are periodically pulled in and ingested together with the planktonic organisms collected. In a further review of Serpulorbis gigas, Yonge and Iles (op. cit.) point out that the pallial cavity has entirely lost its ciliary feeding mechanism; the ctenidium is reduced, and weakly ciliated, the food tract little developed, and only a feeble water current is maintained through the pallial cavity. In Serpulorbis zelandicus, mucus threads are undoubtedly formed by the animal in its natural location, but were seldom able to be satisfactorily observed. They were sometimes seen at low tide on overturning a rock, when a thread two or three centimetres in length could be identified still attached to the opening of the pedal gland. The threads were always thin and delicate, though it would appear that when the animal is covered with water, and wave surge is reduced, mucus traps may be put out without interruption to form a supplementary means of food collection. The high degree of development of the pedal gland strongly supports this suggestion. “Vermetus” novae-hollandiae, which is typically a species of rough water, is stated never to form mucus strings: Serpulorbis gigas, which subsists entirely by mucus feeding, is a calm-water species. The pallial organs in Serpulorbis zelandicus are relatively much better developed than in gigas, corresponding with the retention of ciliary feeding. Zelandicus is clearly a transitional form between the V. novae-hollandiae and S. gigas types, and makes it easy to see how the two extreme groups of Yonge and Iles may be related. All attempts to observe the formation of mucus traps in zelandicus in the aquarium were unsuccessful. A large bolus of mucus was generally extruded indecisively over the rim of the operculigerous disc, but was never elaborated into strings. Yonge (personal conversation) states a similar difficulty was found in inducing gigas to feed under laboratory conditions.

II. The Alimentary Canal

The Vermetidae—as shown by Yonge (1932)—are among those prosobranchs that have developed a crystalline style, in correlation with the mode of continuous feeding on fine particles. The foregut is a simplified region producing an abundant mucus supply by which food is carried back to the stomach. In the specialized stomach region the food particles are stirred and subjected to preliminary digestion by the style, and sorted by ciliary action, after which assimilable material is passed into the digestive diverticula. As usual in current-feeding animals, the intestine is devoted to the formation of firm faeces.

The mouth in Serpulorbis zelandicus is a vertical slit, equipped with a pair of cuticular jaws, whose edges diverge below on either side of a protrusible odontophore. The apposable margins of the jaws are strengthened by a row of chitinous denticles, each secreted by a single underlying columnar cell; they are powerful enough to take a firm grasp of a needle placed in the buccal cavity. As the mucus-

The Stomach and Crystalline Style Caecum

The stomach in Serpulorbis zelandicus is a stout, obtusely angled sac with anterior and posterior limbs. The anterior portion is continuous in front with the style caecum and intestine, and the posterior receives the oesophagus on the left and gives exit behind to the posterior digestive diverticulum (Fig. 9, P.DIV.). A second, much smaller, diverticulum (A.DIV.) opens anteriorly, immediately below the mouth of the style caecum. The angle of the stomach is dilated into a spherical chamber, lined with cuticle, in which the head of the style rotates, bearing against a thin, curved plate of cuticle forming the gastric shield (G.SH.). The style caecum (S.CM.) is wide and cylindrical, slightly tapering anteriorly, and lined with tall transverse folds of ciliated epithelium.

The cilia—as typically in this location—are extremely long and robust, equal in length (15μ) to half the height of the cells. Their rapid transverse beat brings about a clockwise rotation of the style. The caecum is in wide communication with the proximal part of the intestine, being bounded by two narrow typhlosoles which remain adpressed to the style in life. The dorsal typhlosole is clad with short eilia, beating backwards towards the stomach and imparting a down-

ward thrust to the style. Along its summit is a tract of columnar cells with darker staining contents, comparable with the style secreting zone in Struthiolaria (Morton, 1951). The ventral typhlosole also possesses short cilia, though its currents—while probably towards the intestine—are indistinct. To the left of the style caecum, a rapid forward current is maintained by the intestinal ciliated epithelium. The crystalline style (C.ST.) in Serpulorbis zelandicus is 4–5 mm. in length, relatively less stout than in “Vermetus” novae-hollandiae, and noteworthy for its delicate semi-fluid structure. It dissolves rapidly on cessation of feeding or when removed from the animal. The gastric end of the style generally continues directly into a mucus string containing ingested food material—chiefly diatoms—drawn out of the opening of the oesophagus. In contrast with Struthiolaria (Morton, 1951) there is no broad protective typhlosolar flange enwrapping the style, and its whole substance is often permeated with finely divided food particles. These appear to be swept into the caecum from the intestine and caught up in the viscid style substance, in which they are gradually carried back to the stomach as the style is thrust down, being no doubt partly digested meanwhile by the amylolytic style enzyme. Owing to the finely divided nature of the food entering the stomach, the delicate style performs no triturating function; its chief mechanical role is to promote the constant circulation of the stomach contents in the vicinity of the ciliary sorting area.

The sorting area (C.ST.A.) occupies all the left aspect of the stomach. It forms an extensive series of ciliated ridges and furrows, at first very narrow, and converging anteriorly to form about 12 wider folds which terminate abruptly at the opening of the intestine. The sorting surface is further increased by a long S-shaped fold (F.) of the wall of the anterior chamber, extending from the end of the typhlosole to the gastric shield. This flap incompletely separates the sorting area from the rotating style head. Its left aspect is thrown into a set of narrow folds, passing obliquely forwards to converge upon the intestinal opening. The grooves between the sorting ridges maintain strong ciliary currents towards the intestine, by which coarser particles are eliminated before passage of the stomach contents to the digestive diverticula. The ridges are also covered, especially the wider folds towards the intestine, by transversely beating cilia, which together with the rotating style, keep finer particles in circulation, for transfer to the diverticula.

The left digestive diverticulum opens by a narrow aperture from the posterior chamber of the stomach, which is lined behind the gastric shield with non-cuticulate ciliated epithelium. Particles are carried to the diverticulum by a posterior-directed ciliary current. Material returned from the digestive gland goes direct to the intestine, along an extension of the sorting area, which passes obliquely across the stomach, below the gastric shield. This ciliated region is separated from the cuticulate region by a narrow rim-like fold. Fine ciliary currents beat into the mouths of the diverticula, carrying the finest divided particles to the digestive epithelium, which forms a very extensive ingesting surface.

The Digestive Gland

The digestive lobules possess both digestive and excretory cells. The digestive cells (Fig. 13, DI.C.) are tall (100μ) and columnar, with basal nuclei; their cytoplasm is crowded with spherules, formed of greenish-yellow particles clumped together, in little boluses, 7μ across. From time to time these are egested from the cells and returned to the stomach. As in some other prosobranchs, the free cell surface is probably sparsely ciliated in life, but cilia are difficult to identify in sections, and the cell borders take on a convex pseudopodial appearance. Excretory cells (EXC.C.), presumably a means of rejecting absorbed chlorophyllous pigments, occur much less frequently, interspersed with the digestive cells. They are broad-based and pyramidal with brown-staining contents, and usually contain each a single brownish-black spherule, 25μ or more across. These spherules appear with the digestive cell contents in the egested material passed into the stomach. Faecal material is derived from two sources, first from the empty diatom frustules which are never found within the diverticula and apparently after preliminary digestion of their contents are carried directly to the proximal intestine, secondly from the egested particles from the two types of digestive gland cell. As in Struthiolaria, but contrary to the suggestion of Mansour (1946) in certain lamellibranchs, the greenish digestive cell particles are not enzymatic but represent waste products of digestion. In Vermicularia they were actually observed in the living stomach to pass at once by ciliary currents from the anterior diverticulum to the proximal intestine.

Intestine

III. Reproductive System

Serpulorbis, like the rest of the vermetid genera, is incubatory. The eggs are retained in capsules, attached to the inner surface of the shell, along the median pallial slit of the female. From 3 to 5 capsules are generally produced at one time, each containing about 10 eggs, 0.25 mm in diameter. The free-swimming stage is entirely lost, and the velum of the embryo much reduced in size. On emerging from the capsule membrane, the embryo creeps out of the aperture of the female shell, and is able for a limited time to crawl about, before attaching itself to the substratum and developing the uncoiled adult shell. The male is aphallic, and the sperms are shed from the vas deferens directly into the mantle cavity and carried out by the exhalant current. The female is thus fertilized by the entry of current-borne sperms with the inhalant pallial current. The pallial genital duct is for most of its length widely open for the entry of sperms from the pallial cavity. A similar case of current fertilization with an unclosed genital duct is described by Fretter (1946) in Turritella in which the capsule gland and the albumen gland are both widely open along the ventral surface. The pallial genital duct primitively originated as an unclosed groove, and examples of incomplete closure with current fertilization are to be found also in the Cerithiidae (Zeacumantus, unpublished observation) associated with well-protected pallial apertures in silty habitats, as well as in the sessile vermetids. These cases may have arisen by the re-acquisition of a primitive condition; in the Strombacea on the other hand (Morton, 1949b) copulation occurs, and the unclosed capsule gland is to be attributed to a primitive survival.

The female genital ducts of Serpulorbis zelandicus are illustrated in Figs. 10–12. The narrow ovarian duct (50μ) lined with ciliated epithelium is capable of great distention with the passage of eggs. It opens through the right wall of the albumen gland, which forms the first portion of the pallial oviduct. The lumen of the albumen gland (Fig. 11, ALB.G.) is curved and slit-shaped, closed along the lower edge, and lined by a single row of tall, columnar cells (30μ) filled with secretion spherules staining light-brown with Van Giesen's. These cells secrete the albumen coat surrounding the eggs after fertilization. The wall of the albumen gland is strongly ciliated, and sperms are carried backward from the capsule gland along the ventral side of the lumen to the receptaculum. This is represented by a small spherical pouch, 1.5mm.–2mm. in diameter, attached dorsally to the posterior end of the albumen gland. The lining is of short, cubical epithelium, and a dense row of haematoxylin-stained sperm heads is generally attached to the surface of the cells. It may be suspected that a chemotactic stimulus is responsible for the invariable aggregation of current-borne sperms in this part of the genital duct.

The capsule gland (Figs. 10, 12, CPS.) is much larger than the albumen gland. It forms a pouch-shaped organ, 9–10 mm. long, pale yellow in colour, open along its whole ventral surface into the pallial cavity. The epithelial cells are of two sorts, narrow ciliated cells, with a short ciliary coat (50μ) and dark rod-shaped nuclei, and columnar gland cells, 50μ tall, whose secretion stains lightly with haematoxylin. A single egg capsule occupies the whole lumen of the gland, and the

Discussion

The foregoing description of Serpulorbis zelandicus affords a comparison with “Vermetus” novae-hollandiae, an exclusively ciliary feeding species (Yonge, 1932) and Serpulorbis gigas which has been shown (Boettger, supr. cit.) to rely wholly upon mucus traps for food collecting. Yonge was inclined to separate fairly widely the two latter forms and suggested that the Vermetidae were probably divisible into separate groups on the basis of their widely differing feeding mechanisms, stating (1932) “the taxonomy of the Vermetidae clearly requires revision in the light of these results.” Serpulorbis zelandicus, is now found to show a condition intermediate in most respects between novae-hollandiae and gigas, suggesting that ciliary and mucus trap feeding represent merely extremes of a single evolutionary series. Without at present opening the question whether all of the forms at present included in the Vermetidae are closely related, the evidence is strong that the three forms above discussed should be regarded as belonging to a single natural group. In each case the adaptation of the foot is fundamentally the same, and the detailed structure of the pedal gland and tentacles is closely similar. In the alimentary canal likewise, novae-hollandiae as described by Yonge (1932) shows a well recognizable resemblance to sipho; the digestive system of gigas, though the description of Yonge and Iles (supr. cit.) is merely a short outline, appears to conform to the same plan. In each species the pharynx is relatively prominent and bulbous, and the salivary glands minute with short ducts. The anterior portion of the oesophagus is dilated, being crop-like in novae-hollandiae, cylindrical and rather less wide in gigas and zelandicus. The structure of the stomach and style caecum, and the disposition and relative size of the digestive diverticula also agree in novae-hollandiae and gigas. Between gigas and zelandicus there is especially close agreement in the radula, the figure of Troschel (Das Gebiss der Schnecken) for gigas being almost identical with Text-fig. 1 above, affording strong support to the return of the New Zealand shells to Serpulorbis. The chief feature in which zelandicus differs from gigas and resembles novae-hollandiae is in the retention of the ciliary feeding habit, and the consequent high development of the ctenidium and food tract. Bivonia, represented by triqueter, described by Lacaze-Duthiers, is also typical of the present group in structure of the foot, and in the alimentary and reproductive systems. The structure of the genital ducts is incompletely known, but there is evidence that zelandicus is here again typical of the whole of the present group.

“Vermetus” novae-hollandiae stands somewhat apart from the other forms above by its extremely large size, and the straight cylindrical shell. The ciliary feeding condition in novae-hollandiae is no doubt a more primitive stage of vermetid evolution, and it should

be noted that this species retains throughout life the large slightly concave operculum, with the same relative proportions as in the embryo of zelandicus. The gill filaments are strongly developed and powerfully ciliated, though from their triangular shape, as well as from the lack of an endostyle and of a specialized food groove, it would appear that members of this group never become so completely adapted for ciliary feeding as some of the other style-bearing prosobranchs. Moreover, although the pedal gland in novae-hollandiae is much less well-developed than in either zelandicus or gigas it is yet of relatively considerable size. It may be suggested that this structure was not originally developed in the Vermetidae solely for the production of mucus for ciliary feeding, for which requirement the food tract mucus supply would seem sufficient. Although novae-hollandiae—occurring as stressed by Yonge in a wave-beaten environment—does not resort at all to mucus trap feeding—it is likely that most other species of this group may—like zelandicus—possess to some extent the ability to form mucus traps. Bivonia triqueter is a mucus trap feeder according to Yonge (1932), while on the north American Pacific coast, Spiroglyphus lituellus and Aletes squamigerus—as figured by Johnson and Snook (1935) have the same structure of head, foot, pedal tentacles and food tract as described above. Ricketts and Calvin (1948) briefly state the last-named species feeds by the formation of mucus strings. “Vermetus” maximus and “Vermetus” giganteus are stated by Yonge and Iles (supr. cit.) to possess the more primitive ciliary feeding condition of novae-hollandiae, though their exact relationship to the above species has yet to be ascertained.

It is probable—as suggested by Yonge and Iles (1939)—that the presence of operculum and a powerful ctenidial current tends to impede the employment of mucus strings. In zelandicus, however, mucus feeding appears to be carried out to a minor extent in the presence of a strong current passing into the pallial cavity. Pari passu with the loss of the operculum and the enlargement of the pedal gland, the ctenidium is reduced in size in vermetids, reaching its smallest proportions in gigas, which is probably mainly dependent on pallial respiration. A parallel evolutionary trend seems to be the reduction of the crystalline style. According to Yonge's figure the style is large and stout in novae-hollandiae; in zelandicus it is much more slender, and very delicate and impermanent. Its presence in gigas was first denied by Yonge (1932), but examination of living material (Yonge and Iles, 1939) later revealed a style and a small gastric shield “relatively much smaller than in ‘Vermetus’ novae-hollandiae.” It would be interesting to find whether the style is ever actually lost in mucus trap feeders; the ingestion of predominantly animal material such as small isopods captured by gigas (Boettger, 1930) would greatly diminish the need of a style. Yonge later found small gastropod shells in the stomach of gigas, and it seems well established that the mucus feeders are collectors of zooplankton. The presence or absence of a crystalline style is primarily a physiological character, which is probably not of major taxonomic importance—witness its separate appearance in gastropods in such groups as the prosobranchs and the thecostomatous pteropods.

Summary

The taxonomy of the New Zealand vermetids of the genus Serpulorbis is discussed. The shell, animal and dentition of S. zelandicus (Q. & G.) are described, and a new species, S. aotearoicus, is proposed. Novastoa zelandica becomes Novastoa lamellosa (Hutton). A full description is given of the external features, pallial organs and mode of ciliary feeding. Resemblances to “Vermetus” novae-hollandiae are indicated. In addition, S. zelandicus is shown to possess a large pedal mucous gland, and to feed to some extent by the supplementary mode of forming mucus strings. The species is regarded as illustrating the transition from exclusive ciliary feeding in “Vermetus” novae hollandiae to the complete development of mucus trap feeding in Serpulorbis gigas. The morphology and histology of the alimentary canal of Serpulorbis zelandicus is described, with an account of the ciliary currents and mode of function of the stomach and crystalline style caecum. The female reproductive system shows adaptations for current fertilization and for incubation of the eggs within the shell of the female. A comparative discussion follows, on the position of Serpulorbis zelandicus with relation to other members of the Vermetidae. Yonge's view that the ciliary feeders and the mucus trap feeders form two distinct groups is not upheld. A continuous evolutionary trend is shown from the more primitive ciliary feeding condition to mucus feeding, with reduction of the ctenidium and enlargement of the pedal gland. The genera Bivonia, Aletes, Petaloconchus, and Spiroglyphus are associated with the present series.

References to Literature

Boettger, C. R., 1930. Studien zur Physiologie der Nahrungsaufnahme festgewach sener Schnecken. Die Ernahrung der Wurmschnecke Vermetus. Biol. Zbl. 56, 581–597.

Cranwell, Lucy M., and Moore, Lucy B., 1938. Intertidal Communities of the Poor Knights Islands, New Zealand. Trans. Roy. Soc. N.Z., 67, 375–407, pls. 53, 54.

Finlay, H. J., 1927. A Further Commentary on New Zealand Molluscan Systematics. Trans. N.Z. Inst., 57, 320–485, pls. 18–23.

——, H. J., 1928. The Recent Mollusca of the Chatham Islands. Trans. N.Z. Inst., 59, 232–286, pls. 38–43.

Fretter, Vera, 1946. The Genital Ducts of Theodoxus, Lamellaria and Trivia, and a Discussion on their Evolution in the Prosobranchs. Journ. Mar. Biol. Assoc. U.K., 26, 312–349.

Graham, A., 1938a. On a Ciliary process of food collecting in the gastropod Turritella communis Risso. Proc. Zool. Soc. Lond. (A), 107, 453–463.

——, A., 1938b. On the Alimentary Canal in the Style-bearing Prosobranchs. Proc. Zool. Soc. Lond. (B), 107, 75–112.

Johnson, Myrtle E., and Snook, H. J., 1935. Seashore Animals of the Pacific Coast. New York: The Macmillan Coy.

Lacaze-Duthiers, H., 1860. Memoire sur l'Anatomie et l'Embryologie des Vermets. (Vermetus triqueter et V. semisurrectus Phil.) Ann. Sci. Nat. Zool. (4), xiii, 209–296.

Mansour, K., 1946. Feeding and Digestive Organs of Lamellibranchs. Nature, 158, 378.

Morton, J. E., 1949. The Adaptations of Xenophora, the Carrier Shell. N.Z. Sci. Rev., 7, (10), 188–189.

——, J. E., 1951. The Ecology and Digestive System of the Struthiolariidae (Gastropoda). Quart. Journ. Mior. Sci. (in the press).

The Structure and Adaptations of the New Zealand Vermetidae
Part II. The Genera Stephopoma and Pyxipoma

By J. E. Morton,
Department of Zoology, Auckland University College

neozelanic molluscan faunas. It is doubtful whether the genus Lilax, as proposed by Finlay (1927) is now really required for the Australian shell.

Stephopoma roseum is a typically low-tidal species, accessible only at low water spring tides. At Otata Island it is present as the dominant sessile member of a low-tidal animal association, on the vertical sides of rocky channels through which there is a strong wave surge. Here it forms a zone two or three inches in vertical height, immediately below a band of serpulid worms, and plentifully admixed with pink encrusting corallines. Serpulorbis aotearoicus is present in sparser patches immediately below. At Milford and Takapuna, Stephopoma roseum is a co-dominant sessile animal on the undersides of rocks, together with the barnacle Elminius modestus, sponges and hydroids, and the ascidians Corella eumyota and Didemnum candidum, and a rich collection of non-sessile associated animals. Unlike Serpulorbis, Stephopoma is very tolerant of sheltered backwaters, where the exposed rock surfaces are covered with silty Corallina officinalis, and there is a fine sandy deposit beneath the stones. Clusters of Stephopoma tubes are frequently aggregated at the sides of the boulders near the lower edge of the Corallina fringe, and are here densely covered with a green nulliporite or pink basal Corallina. Shells underneath the rocks in contact with the substratum are translucent white, or pinkish tinted. The species is evidently limited at its upper margin by the exposure factor. It tolerates wave surge as well as Serpulorbis, while in addition, having a protective operculum and more efficient cleansing mechanism, it thrives well in silty locations.

The Pallial Organs and Foot: Feeding and Cleansing Mechanisms.

When removed from the shell Stephopoma roseum is seen to be spirally coiled; the visceral mass is never irregularly vermiform as in Serpulorbis. The foot forms a stout plug-shaped column, surmounted by a very prominent operculum (Fig. 1 OP.). This is a thick chitinous disc, multiannular, and with each annulus fringed at its free edge with long multifid setae, figured in Text Fig. 1. The glandular regions of the foot are white in colour, forming a conspicuous area in front of the mouth. There are no pedal tentacles, but the front edge of the foot is produced into a median vertical process in front of the mouth, the pre-oral appendage discussed below. The rest of the foot and the exposed parts of the head are dark black or grey, and the trunk and pallial region whitish. The proboscis (Fig. 1 RO.) is short and bluntly rounded, jet black in colour, with the mouth a vertical slit at its tip. A pair of minute tubercles (TNT.) represents the cephalic tentacles, with small eyes at their outer bases. The pallial cavity forms a spacious chamber, and its contained organs may be conveniently considered in discussing the feeding process.

Stephopoma is a ciliary feeder on diatoms collected from the pallial current by the ctenidial filaments. The mechanism of feeding conforms to the fundamental type made familiar by the accounts of Yonge (1938) and other writers on ciliary feeding prosobranchs. There are, however, numerous differences in detail, and while most closely resembling Crepidula, Stephopoma is in some respects more

exploit to the full its available feeding area. Stephopoma is thus to be regarded as ecologically equivalent to the sessile cirripedes and the tubicolous serpulid polychaetes.

The Alimentary Canal.

The digestive system of Stephopoma roseum has the general plan typical of a style-bearing prosobranch, most resembling that of Turritella (Graham, 1939) in degree of specialisation. Attention is here devoted chiefly to the considerable differences in detail from Serpulorbis, previously described by the writer (1951a). In the buccal bulb, the jaw plates are reduced to a thin flexible cuticle lining the sides of the mouth. Chitinous mandibular rods are present in a small patch on either side, secreted by underlying columnar cells; they play little part, however, in seizing food material, which function is performed by the sharp, highly erectile teeth of the radula. When the odontopore is protruded the curved marginal teeth and the single-pointed laterals form a set of tiny grappling hooks by which a bolus is detached from the food groove mucus cord, and withdrawn into the mouth. The teeth are subjected to little wear and tear, and the radular caecum, from which the radula is replaced, is very short, terminating immediately behind the pharynx. The odontophoral musculature is slender and reduced, much as in Turritella (personal obs.), and in contrast with the ciliary feeding Crepidula, where the muscles that protrude and retract the odontophore remain unusually large.

The oesophagus (Figs. 5, 8, 9, Oe.) takes its origin in a glandular and ciliated dorsal channel (see Fig. 10) bounded by lateral folds, and forming the roof of the pharynx. A pair of tiny salivary glands open at this point by short ducts, and are composed histologically of mucus cells alone. Immediately behind the pharynx the oesophagus descends vertically, to pass through the nerve ring, after which it turns sharply backwards and passes to the stomach as a uniformly narrow tube (0.15 mm. in diameter). In marked contrast with the dilated crop-like structure in “Vermetus” novae-hollandiae, and the wide anterior region in Serpulorbis zelandicus, the oesophagus in Stephopoma has lost all trace of its primitive division into dorsal glandular and ventral food-conducting portions. The condition is as in the middle and posterior oesophagus of Turritella and Crepidula (Graham, op. cit.). Ciliary cells are uniformly present, with tall cilia (10μ–12μ) and the epithelial folds are about equally developed. Plump mucous goblet cells are prominent, their contents staining rather lightly with haematoxylin. Posteriorly there appear some well-defined longitudinal ridges carrying the oesophageal food string back to the stomach. The oesophagus is here frequently rather sinuous in course, becoming straightened out when the animal is protruded from the shell and the trunk region fully extended.

The stomach (Fig. 8) is a small, triangular sac, opening widely in front to the crystalline style sac, a short, bluntly rounded caecum, 0·75 mm. long in open communication with the intestine along its left side (I.GR.). The anterior region of the stomach, which contains the rotating head of the crystalline style, is lined on the right side with transparent cuticle, continuous below with a small rigid gastric shield (SH.). This forms a triangular shelf of hard cuticle, secreted by a

II. Pyxipoma Weldii.

The genus Siliquaria (s. lat.) includes all those vermetids with the shell coiled in a cork-screw shaped spiral, and fissured by a longitudinal slit. There are two New Zealand species, of which the first is generally placed in the separate genus Pyxipoma. Powell (1940) predicted that the neozelanic Pyxipoma would be found to be distinct from the Australian weldii (type locality Tasmania) to which it has hitherto been referred. Though from our knowledge of other New Zealand vermetids, this might seem likely, New Zealand specimens in fact proved identical with Australian material from two separate localities—Seaspray, near Sale, Victoria, and Rhyll, Victoria. Adult shell features vary a good deal in compactness of spire, but the sculpture, and arrangement of growth lines is quite similar. The nucleus is golden brown and its

the vertical pre-oral appendage of Stephopoma. Pelseneer (1906) in his diagnosis of Siliquaria, states that the “pedal tentacles are rudimentary.” In Pyxipoma weldii, pedal tentacles do not exist; these structures in vermetids are associated with the presence of the specialised pedal gland, of which there is here no trace. The margins of the anterior flap are sometimes produced into short, blunt lateral processes.

The alimentary canal in Pyxipoma (see Fig. 2) is so like that of Stephopoma as not to need detailed description. The pharynx has the jaws reduced, and the principal means of grasping the food bolus is the small, sharply toothed radula. The oesophagus is a simple, uniformly narrow tube, producing abundant mucus, and lined with cilia which carry the food string to the stomach. There is no longer any trace of dorso-ventral division, or of the occurrence of torsion. The stomach is crescentic or triangular in shape, 0·5 mm. in length, and the crystalline style sac short and stout, with the same relations as in Stephopoma. The digestive gland consists of a single lobe, the diverticulum opening into the stomach as in Stephopoma. The first part of the intestine is cut off from the style sac by paired typhlosoles, and opens forward into the middle intestine, which loops back round the renal organ. This part of the intestine is somewhat sinuous in outline (Fig. 2, int.) as distinct from the simple loop described in Stephopoma. The rectum opens by the anal papilla, which projects slightly across the edge of the pallial slit.

III. Reproductive System.

In both Stephopoma and Pyxipoma the eggs are incubated by the female: the free-swimming veliger stage is eliminated, and the embryos emerge to wander about for a short period before becoming sessile. Unlike Serpulorbis and previously described vermetids, the eggs are not enclosed in common capsules attached to the inside of the shell. Each egg is contained in a separate capsule, and in Stephopoma roseum the ova—some 10–15 in number—are retained freely within the mantle cavity of the female, crawling out from the pallial aperture after hatching from the egg membrane. In the male Stephopoma, the testis is composed of two or three saccular lobules, drained by a narrow gonadial duct, convoluted, and lined with low-celled, non-glandular epithelium. The sperms in the gonadial duct are aggregated in dark-staining bundles 7μ across. There is no prostate, and the sperms leave the male aperture far back in the pallial cavity, being carried forward by cilia along the right margin of the food groove and discharged with the exhalant current. As in other current fertilised prosobranchs, cf. Turritella (Fretter, 1946); Serpulorbis (Morton, 1949b), the pallial genital duct is widely open ventrally for the reception of water-borne sperms. The small ovarian duct (Fig. 7, ov.) capable of great distention during passage of eggs, opens into a short, albumen gland (al.) dorso-ventrally compressed and situated between the oesophagus and the intestine. The dorsal and ventral walls are strongly ciliated, and possess columnar gland cells (75μ tall) filled with protein spherules secreted round the egg, and staining orange yellow in Van Giesen's. The anterior half of the albumen gland remains open to the pallial cavity by a narrow slit along the left side

of the lumen. Towards the capsule gland its cell contents begin to change to a mucoid secretion, staining deeply purple with haematoxylin. The capsule gland itself (cp.) is a long, straight tube, with dorsoventral lumen open to the pallial cavity, with white, glandular walls, deeply furrowed by vertical rugae. The gland cells are uniformly filled with mucoid secretion, lightly staining with haematoxylin. The eggs are surrounded by a tough mucoid membrane of the same staining reaction as the capsule gland cells. Fig. 20 is of an embryo about to break through the capsule membrane. Towards the hatching stage, the embryo is equipped with a large circular operculum, with the margin unornamented. By the sharp edge of the operculum the embryo breaks through the capsule membrane. At the crawling stage depicted in Fig. 21, the embryo shell is complete, with its wide trumpet-like mouth. The foot is long and narrow, rounded behind, with finely ciliated, glandular sole, and somewhat more expanded, strongly ciliated in front. The anterior margin of the foot is incised by a groove, into which open a cluster of mucus-secreting gland cells (Fig. 21, pg.). The cephalic tentacles are long and diverging with eyes well developed at the bases. The gill filaments are narrow and club-shaped, relatively few in number, and the food groove and exhalant siphonal rudiments are already present. The shell does not become attached till a slightly later stage, and as in Serpulorbis, the crawling embryo adds a portion of the first unsculptured whorl to the completed nuclear shell.

Pyxipoma differs little from Stephopoma in the arrangement of the genital ducts. The capsule gland opens at the bottom of a deep groove, overhung by a broad fold, by which the genital duct is shut off from the strong current passing through the pallial slit. The chief feature of interest in Pyxipoma is the possession of a spacious incubatory pouch, lying within the trunk cavity. This opens forward by a wide duct with its aperture immediately beneath the right margin of the foot (Text Fig. 3, bp.) just in front of the termination of the food groove. The eggs pass forward by a ciliated furrow along the right side of the food groove, and enter the incubatory pouch, from which they emerge as creeping embryos to wander about before attachment. The brood is larger than in Stephopoma—about 100–150—and the embryos smaller, 450μ in shell diameter. The course of development is otherwise similar. The embryonic operculum is simple, flat and translucent. The development of a brood pouch with a cephalic opening, and the retention of the primitive ciliated oviducal groove, has occurred also in the freshwater mesogastropods Tanganyicia and Melania (Moore, 1899). There seems to be no point in Moore's contention that this is a highly primitive character. Ovoviviparity is always an advanced feature in mesogastropods, and incubatory devices of various types have developed separately in a number of highly specialised forms.

Discussion.

similarity in histological structure.

Such details of the animal of Vermicularia as were provided by Mörch (1861) also suggest an affinity with Stephopoma: pedal tentacles are not mentioned, and the mantle is stated to be fringed at its margin with short filaments. The cephalic tentacles are short and conical. The condition of the gill is not described, but the presence of a well-developed food groove is indicated by reference to an elevated ridge that “runs along the back, becomes flattened into a membrane at the head, and passes round under the right tentacle forming a kind of canal, near which is the anus”. A relationship between the Turritellidae and the corkscrew-shaped vermetids is rather strikingly suggested by the developmental history of Vermicularia, as for example in V. spirata illustrated in Fig. 24, which begins life as a close-coiled turreted spiral almost indistinguishable from a Turritella. An examination of living material of species of Vermicularia, and in particular a comparison of animal structure, apex of shell and radula with the Turritellidae is much to be desired.

If, as has been suggested, it is considered more convenient to regard the genera Stephopoma, Pyxipoma and Siliquaria as a separate family from the Vermetidae, s. str., this group might be best placed alongside the Turritellidae, and would be known, from its oldest genus, as the Siliquariidae.

Synonymic List of the Recent New Zealand Vermetidae and Siliquariidae

Acknowledgment

The writer is deeply indebted to Mr. A. W. B. Powell for generous advice and encouragement, as regards both the Vermetidae and the mollusca at large, and for the frequent loan of specimens; to the Auckland Museum for making available specimens from the Finlay Collection; to the New Zealand Geological Survey for the loan of material from the Suter Collection; and to Mr. R. K. Dell, Conchologist, Dominion Museum, Miss J. Hope McPherson, of the National Museum of Victoria, and Mr. B. C. Cotton, of the South Australian Museum, for generous help in securing vermetid shells and animals. Mr. D. Whillans was kind enough to prepare the photomicrographs in Plate 2 of this part.

Summary

The sole neozelanic species of the genus Stephopoma is shown to be roseum (Q. and G.). The South Australian nucleogranosum is not represented, and shows slight but valid differences from roseum. Radular, nuclear and opercular characters are considered, and it is suggested that Finlay's genus Lilax is not required. The mode of life,

and feeding and cleansing mechanisms of Stephopoma roseum are described in detail, with the anatomy of the head, foot and pallial cavity. The species is a ciliary feeder, having but few detailed resemblances to the vermetids previously described in life. There is a ciliary method of food collection by the gill filaments most resembling that of Crepidula, and in addition, a supplementary mode of feeding by pro-extruding the anterior fourth of the gill from the pallial cavity to form a sweeping net of cirrus-like filaments. This method of feeding has not been previously described in gastropods. A pedal mucus gland for the extrusion of food-collecting mucous traps is not developed in Stephopoma. The alimentary canal differs to a considerable extent from that of previously described vermetids. An account is given of its structure, histology, and function, with suggestions as to the nature of the digestive mechanism.

Pyxipoma weldii, a member of the siliquariid group of Vermetidae, is validly represented in New Zealand. The operculum, radula, and embryonic shell are described and figured. The anatomy of the head, foot and pallial cavity is closely comparable with that of Stephopoma, and the significance of the pallial slit and shell fissure is considered. The female reproductive system in Stephopoma and Pyxipoma is described, with details of histology. In Pyxipoma there is a spacious brood pouch in the trunk of the female, opening anteriorly at the base of the right tentacle.

The resemblances between Stephopoma and Pyxipoma are emphasised and a tabular summary is provided, showing the dissimilarity between these genera and previously described Vermetidae. The family Vermetidae as at present constituted is concluded to be diphylectic and a separate family the Siliquariidae is proposed. The diagnostic characters of this family are listed, and its relationships with the Turritellidae are discussed. The genus Vermicularia is suggested to be most closely allied to the Siliquariidae.

References to Literature

Finlat, H. J., 1927. A Further Commentary on New Zealand Molluscan Systematics. Trans. N.Z. Inst., 57, 320–485, pl. 18–23.

Fretter, Vera, 1946. The Genital Ducts of Theodoxus, Lamellaria and Trivia, and a Discussion on their Evolution in the Prosobranchs. Jour. Mar. Biol. Assoc. U.K., 26, 312–349.

Graham, A., 1938a. On a Ciliary Process of Food Collecting in the Gastropod Turritella communis Risso. Proc. Zool. Soc. Lond. (A), 107, 453–463.

——, 1938b. On the Alimentary Canal in the Style-bearing Prosobranchs. Proc. Zool. Soc. Lond. (B), 107, 75–112.

Hulbert, G. C. E. B., and Yonoe, C. M., 1938. A Possible Function of the Osphradium in the Gastropoda. Nature, 139, 840.

Mansour, K., 1946. Feeding and Digestive Organs of Lamellibranchs. Nature, 158, 378.

Moore, J. E. S., 1899. The Molluses of the Great African Lakes. III. Tanganyikia rufofilosa and the genus Spekia. Quart. J. Micr. Sci., 42, 155–185.

Mörch, O. A. L., 1861. Review of the Vermetidae. (Part I.) Proc. Zool. Soc. Lond., 30, 145–181, pl. 25.

Morton, J. E., 1951. The Ecology and Digestive System of the Struthiolariidae (Gastropoda). Quart. Journ. Micr. Sc. (in the press).

—— 1951a. The Structure and Adaptations of the New Zealand Vermetidae. Part I. The Genus Serpulorbis. Trans. Roy. Soc. N.Z., vol. 79, 1.

The Structure and Adaptations of the New Zealand Vermetidae
Part III. Novastoa Lamellosa and Its Affinities

By J. E. Morton,
Department of Zoology, Auckland University College.

the pedal gland of Novastoa lamellosa, a sort of triangular or fan-like sheet of secretion, which could become further spread out, loaded with particles and pulled in by the radula. Novastoa animals during feeding were carefully watched with water goggles. No employment of mucus traps could be observed. The water was rather disturbed by wave action near the surface—it is possible that when water disturbance below the surface is less, mucus feeding may be resorted to without disturbance. Undoubtedly, as claimed by Yonge, mucus feeding is chiefly a habit of still-water vermetids, but the rule has its exceptions, notably Serpulorbis zelandicus, which is able to put out mucus traps in rocky channels through which there is a constant wave surge.

Text Fig. 1—Single rows of radular teeth, the right marginals omitted in each case.
(a) Novastoa lamellosa (Hutton), Poor Knights Is., N.Z. (b) Spiroglyphus
megamastus (March), Cutalina Id., California. (c) Petaloconchus montereyensis
Dall., Carmel Point, nr. Monterey, California. (d) Aletes squamigerus Carpenter,
Newport Bay, California.

Alimentary Canal

The digestive system of Novastoa lamellosa conforms to the general vermetid plan, described by the writer for Serpulorbis. The chief features of comparative importance are the radula (Text Fig. 1, a), and the structure of the stomach and style sac, which form the most complex functional region of the gut.

As with the evidence of the apex and the operculum, the radula affords little doubt as to the close relationship of Novastoa and Spiroglyphus. In Novastoa lamellosa the central tooth is produced posterolaterally into slightly curved, back-directed wings, fitting into concavities in the mesial edges of the laterals. The posterior margin has a median obtuse point, being shallowly emarginate on either side. The anterior edge bears a long median cusp, flanked at each side by a row of three smaller denticles. The lateral tooth has a large cusp the size of the central cusp, with one smaller denticle to the mesial side and a row of three denticles laterally. The inner marginal is equipped with a spur-like forward-projecting cusp a short way behind the tip, with two shorter denticles further back. The outer marginal possesses the sharp, spur-like cusp alone. In Spiroglyphus (Text Fig. 1, b) the postero-lateral wings of the central are stout and peg-like, and the posterior margin is edged with a thin triangular flange. The long median cusp is flanked by four denticles, not three as in Novastoa: a larger denticle flanking the cusp on either side, with an outer row of three short denticles, which may occasionally increase to four. The laterals and marginals are scarcely to be distinguished from those of Novastoa.

The details of the stomach and visceral mass resemble most closely the description by Yonge of “Vermetus” novae-hollandiae. The stomach (Fig. 5) is a large cylindrical chamber, with a shorter style caecum (s.cm.) lying in front. The digestive gland is divided into two parts: a smaller anterior lobe (dg.r.) on the right side, close to the style caecum, and a larger posterior lobe opening from the stomach behind. The posterior lobe (dg.l.) is not greatly elongate and vermiform as in Serpulorbis but shorter, and bluntly tipped, as figured for V. novae-hollandiae, filling the shell tube back to the first septum. The stomach wall is thick, but contains little muscular tissue, being made up of an opaque parenchymatous connective tissue, apparently storing metabolic reserves. The style caecum is separated from the proximal part of the intestine by a single wide, low typhlosole (ty.) running along the ventral side. It opens into the spherical anterior portion of the stomach, the anterior digestive diverticulum being located just beneath the mouth of the caecum. The style rotates against the gastric shield (g.sh.), a transparent recurved plate of cuticle, wide and thin at its margin. The small and rounded gastric shield figured by Yonge in “Vermetus” novae-hollandiae, probably corresponds to the epithelial ridge secreting the actual substance of the shield, or to the strongly projecting fold of epithelium separating the oesophageal opening on the left from the gastric shield. The crystalline style (s.) is retained without resorption through the period between tides when the animals are exposed and feeding is interrupted. It is a narrow, very delicate, flexible rod, small in relation to the size of the animal. Its colour is golden-brown, and diatoms are frequently

contained in the matrix. Apparently a certain amount of food material carried into the intestine becomes caught up in the style substance and carried back to the stomach as the style is thrust downward. A similar “retrieving function” of the style is mentioned by Yonge (1926) in Ostraea, and by the present writer in Serpulorbis (Morton, 1950). In each case the typhlosole is short and the style sac remains widely open to the intestine. A mass of mucus-bound food material is always attached to the posterior end of the style, continuous with or augmented by the food string entering from the oesophagus. The posterior digestive diverticulum (dv.p.) is a wide-mouthed tube opening from the posterior end of the stomach well behind the gastric shield. The ciliary sorting area (s.a.) is on the left side of the stomach, below the intestinal aperture. It is rather small in extent, its size being no doubt correlable with the small size of the style and the relatively pure, well-sorted condition in which diatom material is brought to the stomach. The epithelium is raised into a series of small plicae, the grooves between them collecting rejected particles and empty diatom frustules for carriage to the intestine. At the opening of the intestine the grooves and plicae cease abruptly and the epithelium becomes smooth. The middle intestine (m.i.) passes to the right as a single loop, as in “Vermetus” novae-hollandiae and Petaloconchus montereyensis, separating the renal organ from the anterior lobe of the digestive gland and widening in front into the rectum.

Relationships

All the vermetid genera with mucus gland and pedal tentacles undoubtedly belong to a natural group, and the evolution of the family is a story of the varying degrees to which mucus trap feeding has been adopted in substitution for the earlier method of ctenidial ciliary feeding. The writer has examined critically the animals of representative species of five genera, namely, Serpulorbis zelandicus, Aletes squamigerus, Petaloconchus montereyensis, Spiroglyphus megamastus, and Novastoa lamellosa. Of these the last three may be placed together, Novastoa and Spiroglyphus especially closely, and Petaloconchus somewhat further removed. All appear to be predominantly ciliary feeders, showing approximately the same degree of development of the ctenidium, pedal gland and pedal tentacles as “Vermetus” novae-hollandiae. The alimentary canal seems to be much alike in all these ciliary feeders, although the material of Spiroglyphus was unfortunately mutilated, so that no accurate dissection of the stomach could be carried out. The strong similarities in the radulae and opercula of Novastoa and Spiroglyphus have already been noticed. An operculum of “Vermetus” novae-hollandiae is figured (Fig. 4, a) showing the same concave chitinous disc as in Novastoa and Spiroglyphus, widely overlapping the foot, but without development of the calcareous plug and mamilla. Radular material of “Vermetus” novae-hollandiae has so far proved unobtainable. The operculum, however, appears to represent a primitive type among vermetids. It resembles the simple concave plate recapitulated in the encapsulate embryo of such species as Serpulorbis zelandicus, which have lost the operculum in the adult. It conceivably gave rise to the Novastoa-Spiroglyphus type of disc, as also to the operculum of Petaloconchus, which is unlike that of any

An Index of the Rotifers in the C. B. Morris Collection
of Microscope Slides at the Cawthron Institute, Nelson

By C. R. Russell

The Hypocreales of New Zealand. I

By Joan M. Dingley, Plant Diseases Division,
Department of Scientific and Industrial Research

Throughout the order perithecia possess a well-defined wall, usually pseudoparenchymatous, but in some species the mycelial walls are densely thickened and pigmented and the nature of the tissue of the perithecial wall is difficult to discern. In other species pigmented inclusions within the cells gave colour to the perithecium. Where perithecia are immersed in the stromal tissues, the perithecial wall is poorly developed, but it is always distinct. An ostiolum is always present, usually papillate and lined with periphyses. The stroma varies from a loosely aggregated mat of mycelium to a well-developed pulvinate pseudoparenchymatous structure.

Asci develop from ascogenous hyphae at the base of the perithecium. They typically contain eight spores, although in some genera, the spores fragment within the ascus to form numerous part-spores. Within the order asci are divided into two types according to the method in which the spores are liberated.

Paraphyses are absent; it is possible that earlier records of paraphyses were based upon misidentification of pseudoparaphyses or young asci.

Spores may be non-septate, uniseptate, multiseptate or muriform. In some genera the spores fragment at the septa to form unicellular part spores. They are typically hyaline, although lightly pigmented spores are characteristic of some species.

In New Zealand few groups of the Ascomycetes have been studied. In Hooker's Handbook to the New Zealand Flora, published in 1867, two species of Nectria, one Hypocrea and two species of Cordyceps were listed and briefly described. Cooke (1879) listed and described a few further collections sent to Kew from New Zealand. Cunningham (1921) described species in the genus Cordyceps.

The present treatment is by no means complete, as collecting has not been sufficiently extensive throughout the Dominion.

Family I. Hypocreaceae

Perithecia light or brightly coloured, walls pseudoparenchymatous, fleshy or membraneous, rarely dark coloured, never hard or brittle; perithecia scattered or aggregated on or immersed in a stroma. Asci not operculate, spores liberated by rupturing of apical wall of ascus. Paraphyses absent.

Key to Genera

Hypomyces includes species with fleshy, often brightly coloured perithecia, superficial or semi-immersed in an effuse subiculum. Spores are one-septate, fusiform, apiculate and coarsely verrucose. Most species parasitize other fungi. Although originally defined to include only species with one-septate fusiform, apiculate spores, the genus for a time was treated as a biological one to include all Nectria-like fungi which parasitize other fungous fructifications. To-day the genus is limited to its original definition. Sydow (1920) separated species with unequally divided spores into the genus Apiocrea, but this character seems too variable to justify this treatment, consequently all species are included in Hypomyces. Species with smooth, elliptical or oval one-septate spores are placed under Nectria.

Hypomyces Tulasne. Annales des Science Naturelles, ser. iv, vol. 13, p. 11, 1860.

Clintoniella (Sace.) Rehm. Hedwigia, vol. 39, p. 223, 1900 (in part); Apiocrea Sydow. Annales Mycologici, vol. 18, p. 186, 1920.

Perithecia superficial or immersed in an effuse, byssoid stroma; perithecial wall pseudoparenchymatous, pigmented and hyaline. Asci containing eight spores, paraphyses absent. Spores one-septate, fusiform, apiculate, verrucose, hyaline or lightly pigmented.

Type Species. Hypomyces lactifluorum Schw.

Distribution. World wide.

Key to Species

1. Hypomyces aurantius (Persoon) Tulasne. Selecta Fungorum Carpologia, vol. 3, p. 48, 1865. (Plate 10, Fig. 3.)

Sphaeria aurantia Pers. Synopsis Fung., p. 18, 1801; S. aurantia Pers. ex Fr. Syst. Myc., vol. 2, p. 440, 1823; Nectria aurantia Fr. Summa Veg. Scand., p. 388, 1849.

Subiculum effuse, spreading over pileus of host, orange and umber, margins floccose, light yellow or white, pigment granules present on outer wall of mycelium; perithecia gregarious, crowded, superficial or semi-immersed in subiculum, globose or oval, 0·2–0·3 × 0·4 mm., orange or ferrugineous, translucent, ostiole papillated; perithecial wall pseudoparenchymatous, 50μ thick, cells 8–15 × 12–25μ, thin walled, lightly pigmented, pigmented granules present among the cells. Asci cylindrical, ends truncate, thin-walled, 80–140 × 4–6μ, 8 spored, spores obliquely uniseriate. Spores one-septate, equally divided into two cells, fusiform, slightly curved, apiculate, 17–26 × 3–6μ hyaline and verrucose.

Flora and Vegetation of the Caswell and George Sounds
District
Area Covered by the New Zealand-American Fiordland Expedition

By A. L. Poole, Director, Botany Division,
Department of Scientific and Industrial Research, Wellington

1847–51 in Milford Sound, Chalky Inlet, Dusky Bay and Preservation Inlet. After the activities of these earlier botanists spasmodic collecting was done by Petrie, Cockayne, Oliver, and others. During the expedition of the New Golden Hind in 1945, H. H. Allan collected extensively in the Sounds and adjacent country.

The Habitat

The basement rocks of most of that country stretching from Milford Sound to Preservation Inlet are metamorphic gneisses with veins of feldspar and quartz. They have been intensively glaciated. Benson, Bartrum and King (1934) have described the geology and geo-morphology of this type of country around Chalky and Preservation Inlets. While the country around Caswell and George Sounds and eastwards to Lake Te Anau is somewhat similar, it is higher and more dissected, and is probably the most rugged stretch of this inhospitable area.

Narrow mountain ranges, the remains of a deeply dissected upland, are separated by precipitous U-shaped valleys which extend as deep sounds to the coastline and to the lakes in the east. The mountains range about 4,000 ft. in height and seldom reach 5,000 ft. The valley floors are flattish, and in the lower reaches are covered with sands and sandy silts. Great steps are present in the valleys, and hanging valleys are frequent. Low hills of morainic material are sometimes present along the sides of the larger valleys, and lakes formed by the deposition of morainic material across valleys, or by rock bars or rock avalanches falling across valleys, are many.

Rivers and streams are numerous, as are waterfalls descending from the mountain tops directly down the valley sides. The vegetation is saturated at most times so that heavy rain finds its way quickly to the streams and rivers, which rise and fall with great rapidity. The Stillwater River, the largest in the area studied, rose twenty-four feet in six hours, and Lake Marchant, into which it flows, rose fourteen feet during the same rain storm.

Except on the sandy valley bottoms and on the moraines the forest virtually sits on unweathered rock in a thick mat of roots, peaty litter and abundant mosses and liverworts. The only hold this forest has is by the roots which penetrate the crcvices of shattered rock. It is astonishing the angle of slope forests are able to grow on. Nevertheless, the vegetation frequently becomes unstable and the scars of natural avalanches are plentiful. On new avalanches the unweathered white surface of the rock is left exposed, the vegetation peeling off it as though it were a skin. This form of gigantic slipping is such a constant feature of the valley sides that it must be a major factor in determining plant succession. Moreover, it has a marked indirect effect in that the process is sufficiently rapid to prevent the accumulation of all but the most minute amounts of coarse sandy soil. The forest, or rather scrub forest, on the valley sides is therefore present by virtue of the constantly moist conditions.

Area Studied

On the 26th February, in company with the first scientific party of the expedition, the writer travelled from Milford to the Stillwater Base Camp. Thereafter a period of six weeks was spent in the area.

This period included a day on Mary Peaks during which a complete traverse was made around the tops, five days at Leslie Clearing, ten days at the Upper Stillwater Camp, from which base Saddle Hill was climbed, two days at George Sound, two days on the Henry Saddle, and two days at Hankinson Hut. The remainder of the time was spent at the Stillwater Base Camp. Immediately after leaving the area, short visits were made to Mount Luxmore and several forest areas, some populated by red deer only, on the east and west shores of Lake Te Anau. These areas lie outside the main wapiti breeding country, but examination of them was useful for comparative purposes.

The area was later visited by three other botanists and a party of the National Forest Survey. Dr. W. R. B. Oliver spent the period from March 26 to April 5 collecting for the Dominion Museum in the vicinity of Caswell Sound, Stillwater base, and Leslie Clearing. Mr. V. D. Zotov worked over approximately the same ground during the same period, making an overland trip to George Sound, and collected both cryptogamic and phanerogamic plants. Miss R. Mason spent the period April 6–24 in the George Sound, Henry Saddle area collecting generally and paying particular attention to the water plants and to the feeding habits of the deer. Samples of the stomach contents of all animals shot were collected and were examined by Miss Mason. The Forest Survey party, under the leadership of Mr. J. T. Holloway, were in the area for about a month, from March 26.

Thanks are due to all these people who have kindly allowed the writer to use information and herbarium material collected by them.

Vegetation

Three major series of plant communities clothe the country: the forest communities ranging from sea-level to approximately 3,000 ft. altitude and dominated almost throughout by silver beech (Nothofagus menziesii); the alpine vegetation above 3,000 ft. altitude dominated for the most part by species of Danthonia (D. flavescens, D. crassiuscula and an unnamed species); and the bog communities differing greatly in nature according to altitude and the method of formation. Southern rata (Metrosideros umbellata) is present throughout the forest communities and on some steep faces equals or exceeds the silver beech in amount. On the Lake Te Anau side mountain beech (Nothofagus cliffortioides) is more abundant along the lake shore. South of the area on the eastern slopes of Mount Luxmore and extending southwards, the mountain beech is dominant up to 2,000 ft. elevation and pockets of N. fusca are present. Introduced plants are absent except for a few species, mostly grasses, around the old beach heads of George and Caswell Sounds, and around the much-frequented Hankinson Hut at the head of Lake Hankinson.

A feature of plant distribution is the number of species which extend from sea-level to timber line or even higher. Dacrydium biforme, silver beech, mountain beech, southern rata, Weinmannia racemosa, Phyllocladus alpinus, Hoheria glabrata, Aristotelia fruticosa, Nothopanax simplex, Pseudopanax lineare, Archeria traversii, Coprosma foetidissima, C. ciliata, are some of the trees and shrubs found throughout the altitudinal range of forest and scrub. Smaller

plants such as Nertera depressa and Oreobolus strictus extend throughout the complete altitudinal range and Rostkovia gracilis comes to low altitudes down waterfalls. The rapid transport of seed down waterways and the constant slipping of vegetation must account for much of this type of distribution.

Large numbers of mosses and liverworts throughout the forest is a characteristic feature. The trunks and limbs of trees and shrubs are almost invariably covered, and festoons of Weymouthia hang from trees and shrubs in the river bank forests. The forest floor is everywhere covered with such a thick mat that regeneration of trees is inhibited and much of it takes place epiphytically or perched on logs where the bryophyte layer is thinner. Epiphytism and perching of plants is widespread in numbers and is rich in species. It is much greater than is normally seen in silver beech forests.

The Forest and Scrub Communities

present are Muehlenbeckia australis, Rubus cissoides, Rhipogonum scandens (only within the immediate influence of the sea) and Metrosideros diffusa which scrambles over the floor as well. Climbing ferns and small epiphytic plants scattered throughout the forest are: Polystichum adianteforme, Polypodium diversifolium, P. billardieri, P. grammitidis, Asplenium flaccidum, Tmesipteris tannensis, Lycopodium billardieri, Earina autumnalis, E. mucronata (near the coast only), Dendrobium cunninghamii. Filmy ferns are abundant on trees, shrubs and forest floor. The most common species are Hymenophyllum mutifidum, H. dilatatum, H. demissum, H. scabrum, H. flabellatum, H. rufescens and Cardiomanes reniforme.

Regeneration of the forest takes place partly from perching plants or epiphytes; seedlings do not always seem to survive on the floor of the forest itself unless it is bared of mosses and liverworts.

(2) Lowland and Montane Forests of the Valley Sides

In this description is included the forest occurring on the moraines or avalanches stretching across the floor or along the sides of some valleys. Since these are composed of rocky materials they form a very different habitat from the sandy valley bottoms. The forest on them is more akin to that on the easier slopes of the valley sides themselves.

The forest is dominated throughout by silver beech except on very steep faces where southern rata becomes dominant or is the sole tree. On the easier slopes up to an altitude of 400 ft. to 500 ft. and within the influence of the sea, two to three trees of rimu (Dacrydium cupressinum) per acre are scattered throughout the forest. These trees, although above the beech canopy, are small for rimu, reaching heights of 70 ft. to 80 ft. and are seldom 2 ft. in diameter. Phyllocladus alpinus, infrequent in the valley bottoms, enters the communities; mountain totara, Weinmannia racemosa and Coprosma foetidissima become more plentiful, Nothopanax anomalum becomes less plentiful and in some extensive areas is absent altogether. This shrub disappears at an altitude of about 1,000 ft. Nothopanax simplex, N. colensoi, Suttonia divaricata and Myrtus pedunculata are present in about the same proportions as they are in the valley floor forests. Pseudopanax lineare increases in quantity with increasing altitude. Mountain beech makes its appearance on ridges and knolls, which, judging from the presence of Lycopodium volubile, might dry out more easily than the surrounding forest. Nertera depressa, and a species hitherto confused with N. dichondraefolia and all the valley bottom ferns are present, and Astelia nervosa is in places plentiful. Lianes are absent and the forest is not difficult to penetrate. Dracophyllum longifolium is plentiful along the coast line.

There is the usual luxuriant growth of mosses and liverwort on the forest floor and over the stems and branches of trees and shrubs. Mention was made earlier that the forest sits in a layer of these plants, mixed with litter and matted roots, on top of the unweathered rock. Its only footholds are the roots which penetrate shattered rock. Such a forest must be highly unstable both physically and ecologically. The sheer weight on steep slopes often causes it to peel off the rock and slip into the valleys. The exposed rock begins to heal again by the growth of hepatics, ferns and such small flowering plants as Nertera

spp., Hydrocotyle spp., Epilobium spp., etc. Such slipping must be an important factor in succession in these valley-side forests. That the forests are unstable ecologically can readily be argued. The nature of the forest floor, especially the absence of soil, is such that any change in the climate to dryer conditions would quickly be reflected in the vegetation. A major change to dryer conditions would undoubtedly see the degeneration of most of the forest to scrub or to still more xerophytic vegetation. The existence of the present forest on the valley sides depends primarily upon the high rainfall spread fairly evenly throughout the year.

(3) Bog Forests

Bog forests occur frequently in gently sloping valley bottoms not subject to flooding, on valley spurs where rock has been polished by glaciers, even when these spurs are very steep, and on the faces of moraines where seepage must occur. Such forests are to be distinguished from scrub and stunted forest surrounding sedge and moss bogs formed behind river levees. True bog forests are evidently the result of continuous water seepage or water flow over the smooth polished rock, which encourages an extra strong growth of mosses including Sphagnum. Sometimes Sphagnum spp. are dominant over sizable areas.

The plants in these communities differ in composition according to altitude, and possibly a number of other undetermined factors, such as the acidity of the seepage, etc. One such subalpine forest on a valley step face at Leslie Clearing contains widely spaced silver beech, Dacrydium biforme and some Nothofagus cliffortioides growing to a height of about 25 ft. Many seedlings, saplings, and suckers of the Dacrydium are present. Intermediate shrubs reaching 10–15 ft. height are Nothopanax simplex, Coprosma pseudocuneata, Suttonia divaricata, Archeria traversii, and Dracophyllum longifolium (mountain form). The open floor is soft and semi-boggy with a thick covering of mosses, including much sphagnum, and liverworts. Other plants growing on this floor include Lycopodium volubile, L. scariosum, Carpha alpina, Oreobolus strictus, Astelia cockaynei, A. linearis, Gahnia procera, Pratia angulata, Nertera spp., Drosera stenopetala and Colobanthus acicularis. The unnamed species of Danthonia descends to this bog and is frequent, while Gentiana montana and Forstera sedifolia which are usually alpine plants are also to be seen.

Near the Upper Stillwater Camp, at about 1,000 ft. elevation, a bog forest is present on the comparatively steep slopes of a valley spur. It is dominated by Dacrydium intermedium and Nothofagus cliffortioides and has a much thinner and dryer layer of mosses than the Leslie Clearing bog forest. Alpine species of flowering plants are absent from the floor. Archeria traversii and Gaultheria spp. are plentiful.

(4) Subalpine Forest and Scrub

As with the lowland and montane forest of the valley sides these communities are present on steep to precipitous slopes. They lie between approximately 2,000 ft. and 3,000 ft. elevation, though the timberline is very irregular, depending on the topography and other factors, and may drop in places several hundred feet. J. T. Holloway

of the Forest Survey Party, attributes this extreme irregularity to a disintegration of the timberline following on a change in the climate of the south-west to cooler and wetter conditions, and in the south-east to cooler and dryer conditions.

Silver beech is the dominant tree right to the timberline, though in places where slips have occurred and are healing over, tongues of Hoheria glabrata are present. Over limited areas Olearia colensoi and Senecio bennettii become dominant and in many places southern rata is co-dominant with the beech. Weinmannia racemosa and mountain beech are the only other tree species present. The canopy, usually about 35 ft.–40 ft. high, is fairly complete, though the trees become greatly reduced in height and somewhat scattered at or near the timberline. The shrub layer is composed of Senecio bennettii, Olearia colensoi, Nothopanax colensoi var. montana, N. simplex, Pseudopanax lineare (many seedlings and juvenile plants of this species are present), Pseudowintera colorata, Suttonia divaricata, Coprosma colensoi, C. foetidissima, C. pseudo-cuneata, C. ciliata, Aristotelia fruticosa, Archeria traversii, Dracophyllum longifolium (mountain form) and occasional Pittosporum crassicaule and Dracophyllum menziesii.

The forest floor has a thick covering of mosses and liverworts in which grow Blechnum procerum, Leptopteris superba, Polystichum vestitum, Blechnum penna-marina, Astelia nervosa, A. cockaynei, Phormium colensoi, Luzuriaga parviflora and Libertia pulchella.

Some of the alpine plants come down into this forest, such as Danthonia flavescens, Ourisia macrocarpa, and Forstera sedifolia.

The irregular timberline at about 3,000 ft. consists of patches of stunted silver beech and southern rata accompanied by shrub species interspersed with patches of alpine vegetation. Shrub groves unaccompanied by the stunted tree species extend to at least 4,000 ft. Characteristic members of these are Dracophyllum fiordense and Olearia crosby-smithiana; other species are Dracophyllum menziesii, D. longifolium (mountain form), Olearia colensoi and Nothopanax colensoi var. montana.

(5) Coastal Scrub of Southern Fiords (Preservation Inlet to Doubtful Sound).

The expedition had no opportunity to examine the coastal scrub, but H. H. Allan has kindly supplied the following account of an area further south which would be comparable with that in the vicinity of George and Caswell Sounds.

“Along the rocky shores of the Sounds where exposure to wind is considerable the forest is margined by a narrow belt of scrub, the prevailing brownish tone colour being due to the dominance of epacrids, especially Dracophyllum longifolium, D. menziesii, Archeria traversii var. australis and Cyathodes acerosa. Over 50 species occur in the association as a whole, of which Pimelea gnidia, Griselinia littoralis, Gaultheria antipoda, Coprosma foetidissima and Pseudopanax lineare are often conspicuous. In more sheltered embayments Olearia operina, O. angustifolia, Carmichaelia arborea, C. grandiflora,

Alpine Vegetation

Bog Communities

Herbaceous Swards

Herbaceous swards of quite limited extent occupy sandy or silty deltas and bars at the mouths of larger streams and rivers where they flow into lakes. These bars are subject to continual alteration as sand and silt is deposited on them. Should they become stable they are quickly invaded by shrubs, in particular Pseudowintera colorata. The plants in these swards are Pratia angulata, Cotula squalida, Gunnera monoica var. albocarpa, Scirpus spp. and Eleocharis gracilis.

The Deer Population and the Vegetation

above the bottoms; and (b) the subalpine forest and scrub and mountain tops above the timberline. Certain easy spurs leading from the tops to the valley bottoms are used as lanes for frequent passing up and down, but by and large the belt of forest between the subalpine and that of the lower sides of the valleys is almost uninhabited. It is the opinion of experienced stalkers that many animals are either permanent valley or permanent top dwellers, but snow during the winter is almost certain to drive the animals into the valleys. Another factor which must be considered when trying to arrive at a comprehensive statement of the effect on the vegetation of the deer population, is that much of the feeding on the tops is done by herds which seem to browse an area heavily and then leave it for a length of time. Extensive areas of tops are thus free of animals and appear to have been so for some time. In the lowland and montane forests there are also large areas almost free of animals. They are the rough steps between the relatively level parts or the valleys. The terrain of these is difficult alike for man and beast. One of the roughest seen has shut off Lake Thomson and is composed of shattered rocks, many of which are each the size of a large house.

The wapiti population seems to be confined towards the east by the presence of red deer and the competition for food. This may not be the full explanation, for the natural rate of colonisation under such adverse conditions is very slow; for instance, the population has not yet spread in numbers farther north to Bligh Sound or south to Charles Sound where there is no competition from red deer and there is ample food.

Valley Bottom Feeding Grounds

Where deer are present in numbers their effect on the vegetation is obvious, but apart from severe depredations on a few species this effect could not so far be described as damaging. Along well-worn tracks on the sandy levees and in the valley floor forest in general browsing is severe on all species of Nothopanax, Griselinia littoralis, Coprosma foetidissima and Schefflera digitata, and on the ferns Polystichum vestitum, Leptopteris superba and in places Dicksonia squarrosa. A number of other shrubs and the grasses Microlaena avenaceae and Danthonia cunninghamii are occasionally eaten. A list of the most important plants eaten is given in Appendix A.

The question arises to what extent shrubs such as Shefflera digitata and species of Carmichaelia, relished by deer and present in very small numbers, have been eaten out. The former shrubs if ever present in any quantity would have been plentiful as an epiphyte as well as a ground plant, but epiphytes are absent. The plant is also absent in some areas which have never been frequented by animals. Carmichaelia grows in inaccessible places on the river bank and it is doubtful if much of it ever occurred in accessible places.

Seedlings and saplings of the silver beech are browsed along tracks and spasmodically elsewhere. It would seem therefore that the major succession of the forest has, as yet, been unaffected by the presence of the animals. This statement does not mean that the same population might not in time bring about a change as preferred foods disappeared

Subalpine Forest and Scrub and Alpine Vegetation

In their use for feeding, these two types of vegetation must be considered together, for animals which frequent the tops browse extensively in the subalpine scrub and forest, and may spend some of the winter months in this belt.

The effect of animals feeding on the alpine vegetation is difficult to assess because it is not easy to determine, apart from a few obvious species, what or how much of a species has been or is being eaten. It would seem that Danthonia flavescens probably forms the main food, though quantities of woody and herbaceous plants are also eaten.

It might be as well to discuss briefly the evidence seen on Mary Peaks and to use this as a standard with which to compare other areas. These peaks are isolated and extend around a large cirque. Stalkers say that animals have been present consistently on them for some ten years. During the course of a day spent in traversing completely around them some twenty-seven wapiti were seen and the effect of one herd of seventeen feeding over a confined area was examined. This area consisted mainly of a good Danthonia-covered slope lying to the south-west. Danthonia flavescens had been browsed heavily in places and the vegetation opened up particularly on spurs. Shrubs browsed extensively were Hebe laingii, H. buxifolia, Coprosma serrulata and the tips of Dracophyllum uniflorum. Herbaceous plants noticeably browsed were Astelia cockaynei, Celmisia holosericea, Aciphylla spp., and Ranunculus spp. A number of other plants listed in Appendix A were eaten to some extent.

Where animals had not been feeding for some time past on these tops the vegetation was closed, though there were many tracks through it. The effects of the browsing on shrubs was noticeable and very few intact or healthy plants of Hebes were to be seen. It seemed possible that the wapiti had reduced considerably the numbers of these as well as of such herbaceous plants as Ranunculus, Astelia and some Celmisia species. A plant such as Anisotome haastii, which was untouched by deer, was plentiful, so it appeared that other herbs eaten by the wapiti might also at one time have been plentiful. They may have suffered more readily from grazing than the grasses. These species were much more plentiful on the steep places inaccessible to animals, though it was possible that this was just a habitat difference. Such appeared to be partly the explanation, for comparison with other tops,

where deer were sparse or according to stalkers had never been present, showed that these plants had probably never been plentiful on the easier slopes dominated by Danthonias.

The subalpine scrub and forest around Mary Peaks showed the effect of animal browsing more obviously than did the alpine vegetation. Damage done to shrubs must take a greater time to recover than does the damage to herbaceous plants and grasses. In the subalpine belt Coprosma foetidissima and N othopanax spp. were eaten or barked where available and were frequently killed. Danthonia flavescens and Phormium colensoi growing in this belt were usually eaten to the ground. On the easier slopes tracks were numerous.

Other tops visited, Saddle Hill, Leslie Clearing tops, and the tops around Henry Saddle, had deer on them or had had them not long previously. The type of browsing and its effect on the vegetation were the same as described for the Mary Peaks. The most noticeable effect was the presence of tracks, the damage to the Hebes and certain other plants such as Phormium colensoi, and the paucity of Ranunculus. spp. Otherwise the vegetation seemed little affected and where the terrain permitted communities were closed. One of the members of the expedition visited Mount Pluvius, an area to which deer have evidently not spread. He reported that the vegetation was very little different from that seen on the deer-frequented tops.

To sum up, it would seem that the present population of deer has modified the vegetation to the extent of affecting a certain limited range of species in the valley bottoms and in the subalpine and alpine communities. Some of these species are affected seriously in that they are being reduced; others, including Pseudowintera colorata and some grasses must be increasing. There is, however, no permanent effect on vegetation cover or succession except in very restricted areas: regeneration of the major forest trees is virtually unaffected.

Since deer are in the area their presence must be tolerated, for total destruction of animals in such country would be impossible. It is therefore necessary that the problem of vegetation conservation—a problem which must now be considered for all remaining native vegetation in New Zealand and not just for particular isolated areas—should be faced rationally. Here is an area, prized by hunters, who are almost the only people visiting its interior, with a population of deer so far not highly destructive to the vegetation. Almost the only management policy practicable would be an endeavour by encouraging shooting, to keep the animal population somewhat below its present level. Reasonable accounts of the vegetation and deer herds are now available, so that inspections from time to time should show whether such a policy can be put into effect and how it works. A danger, possibly a very real one, will come from the infiltration by immigration and by crossing of numbers of red deer which may prove more adaptable to the country, and by the immigration of thar into the area. Thar are already in the Clinton Valley to the north. Large numbers of browsing animals of any kind would have a disastrous effect on such an unstable vegetation cover.

Acknowledgements

Acknowledgment is made to V. J. Cook for the identification of Scirpus, to N. Potts for assistance with the identification of Coprosma, to several members of the staff of the Botany Division for other identifications, to H. H. Allan for the identification of Celmisia, for supplying the section on Coastal Scrub and for checking the manuscript, and to W. R. B. Oliver for numerous suggestions. Acknowledgments to the botanist members of the expedition have already been made in the text.

References

Benson, W. N., Bartrum, J. A., and King, L. C., 1934. The Geology of the Region about Preservation and Chalky Inlets, Southern Fiordland, N.Z. Part II. Trans Roy Soc. N.Z., vol 64, pp. 51–85.

Cheeseman, T. F., 1925. Manual of the New Zealand Flora. Wellington, N.Z. Government Printer.

Cockayne, L., 1928. Vegetation of New Zealand (2nd ed.). Veg. der Erde, Vol. 14. Leipzig, W. Engelman.

Poole, A. L., 1949. Brief Account of the New Zealand-American Fiordland Expedition. N.Z. Science Rev., vol. 7, no. 8.

Wodzicki, K., 1947. Interim Report on Wild Life Problems in New Zealand. D.S.I.R. unpublished report.

Appendix A
Plants Commonly Eaten by Deer
(Compiled from field observations only. Plants are listed as far as
possible in order of preference. The list is not a complete one.

Appendix B
List of Ferns, Lycopods and Flowering Plants

Notes On New Zealand Algae

By V. J. Chapman, Auckland University College

Rhodophyceae

1.	Erythrocladia irregularis Rosenv., on Sertularia sp. New record for New Zealand.
2.	Schmitziella cladophorae sp. nov.

Thallo formante patinam expansa complanatam, crassitudine unicellulari, ex ordinibus cellularum radiantibus intra parietem hospitis positis constantem; non calcario; ordinibus cellularum dichotome ramosis; corporibus genitalibus nemathecia intra protrusiones tumidas formantibus.

Thallus forming an expanded flattened plate one cell thick, composed of radiating rows of cells, within the wall of the host; non-calcareous; branching of cell rows dichotomous; reproductive bodies forming nemathecia within swollen protuberances (fig. 1).

Fig. 1—Schmitziella cladophorae. a, b, margin of thallus; c, section of thallus in wall of host; d, margin of young thallus; e, section of a nemathecium.

Occurrence: Parasitic within the cell walls of Cladophora feredayi wherever the host plant occurs.

This species appears to be specifically restricted to the one species of Cladophora in just the same manner that S. endophloea is restricted in Great Britain to C. pellucida.

Chlorophyceae

Plantis tenuibus, stipitatis, filamentosis, ad 1 mm. latis, 7 mm. longis, colore gramino-viridi; cellulis in binas, trinas, quaternas catervatis, 3–7μ dia., 7–8μ altis in T.S.; membrana gracili, 11·2μ crassa, parietibus vix crassatis.

Plant fine, stipitate, delicate, up to 1 mm. wide and 7 mm. long; cells in 2's, 3's and 4's, 3–7μ diameter, 7–8μ high in T.S.; membrane fine, 11·2μ thick, walls scarcely thickened; grass green; adheres well to paper (fig. 2). Ditch, Russell. Distributed by V. W. Lindauer as No. 227.

Fig. 2—Prasiola delicatula, cells in surface view (a), and t.s. of the thallus (b).

Myxophyceae

Thornthwaite's New System of Climate Classification In Its Application to New Zealand

By B. J. Garnier, Department of Geography, University of Otago

conditions. Thornthwaite's 1931 system has been shown as effectively recognising these differences (Garnier, 1946) and the present study has been undertaken to see if the 1948 proposals provide a better or equally good indication of the climatic variety within the country.

Source Material for the Study¹

In any study of climate in New Zealand one is handicapped by the absence of statistics for many places. Not only is the number of stations having both temperature and precipitation normals derived from periods of twenty years or more small, but the location of them, usually in centres of population, is not ideal from the viewpoint of portraying climatic variety. The number of rainfall recording stations is, however, considerable and totals over five hundred. These are widely distributed over the country, but temperature estimates for them must be arrived at by interpolation from those places for which statistics are available. In the present instance temperature figures were obtained by the application of a lapse rate of 2·74°F per 1,000 feet of elevation (Kidson, 1931a), by the use of temperature anomaly maps specially prepared for the work, and by graphs of the variation of average sea-level temperatures at various latitudes in New Zealand, also specially prepared in connexion with the present study.

The classification of climatic types under Thornthwaite's new system is a formidable task. Even with the use of a maximum number of calculating devices such as nomograms, slide rules, and the tables Thornthwaite has prepared it was found that a rate of not more than three stations an hour was the maximum achieved. Where temperature normals were absent, therefore, full monthly calculations were not made, since the amount of work involved was not considered justified in view of the approximate nature of the source statistics. To obtain moisture categories for rainfall stations where no temperature records are kept, therefore, reference was made to a graph (see Figure 1) which was devised from the results of 53 stations for which full computations were made from monthly temperature and precipitation normals.² Thermal categories for the rainfall stations were obtained on the basis shown in Table I, which was also prepared from the 53 results mentioned above.

Fig. 1

Moisture Categories¹

The essential variety of moisture areas in New Zealand seems, therefore, to be brought out equally well by either Thornthwaite's 1931 or 1948 system of classification. Such differences as do exist are minor in character or, if anything, favour the earlier proposals. A case in point is in the vicinity of Karioi. This is classed as B₄ (1931 P–E Index 114) under the old classification and as A (1948 P–E Index 130) under the new system. The latter arrangement places Karioi within the perhumid area of the central North Island, yet it is at Karioi that an abrupt change takes place in landscape aspects. Northwards is the heavily forested country near Mount Ruapehu with, near Ohakune, gaunt skeletons of trees where burning has taken place in recent times to enable dairying and vegetable growing to take place. South of Karioi one finds open country with the spaciousness of a grassland environment and, near Waiouru and the Desert Road, a tussock setting of treeless, rolling country, which is reminiscent of South Island subhumid conditions. While one cannot discount the importance of historical development, clearing policy, and volcanic ash showers, climatic differences are also there and the vicinity of Karioi seems more allied to the region to the south than to the north. Similarly, the disappearance of B₁ in the Wairarapa valley (Masterton, formerly B₁ now becomes B₂) is an alteration which tends to mask the undoubtedly drier regime of this interior lowland area as compared with the higher land towards the east.

Seasonal Variation of Effective Moisture

The oceanic situation of New Zealand is such that, over the country as a whole, the seasonal variety of climatic conditions is not very marked. In certain localities, however, noteworthy seasonal contrasts appear to exist and some attention has been paid to them. Kidson (Kidson, 1931b), for instance, has divided the country into three rainfall regions: (a) areas with a winter maximum, (b) areas of summer maximum, and (c) areas with February and August minima. Such distributions, however, considered without relation to temperature, are of little use as a guide to effective moisture. Over much of Kidson's summer maximum area, for example, the summer is actually the season of least effective moisture conditions.

Thornthwaite's 1948 system recognises a seasonal variation at one station only in New Zealand. This is Blenheim, where a summer seasonal deficiency is revealed. Other stations are all either r or d for their third letter. No seasonal variety is indicated by the 1931 system either, all stations in this case being also either r or d in their third letter.

There is, however, a tendency in some parts of New Zealand, especially eastern South Island, for summer moisture deficiencies to become apparent. Occasionally these result in very marked changes in farm production levels and it is undoubted farming experience that summer and, to a less extent, autumn are seasons when moisture conservation practices or some application of irrigation water, if available, are necessary if harvests are to be assured. Such contrasts between moisture conditions in summer and winter are mainly confined to the subhumid parts of the country and some indication of them seems desirable in a climate classification for use in New Zealand if it is to bring out one of the significant features of these climates.

For these reasons, therefore, it might be a useful modification of Thornthwaite's original proposals, as far as their application in New Zealand is concerned, to neglect water storage in computing moisture indices. By so doing, the major classification for much of the country is on the whole unaffected. Such alterations as do occur are confined to the drier half of humid climates and to the subhumid group. The general effect of these alterations is to increase the major moisture index values (see Table VI) which in some cases causes

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

a change of class but also brings out the seasonal contrasts already discussed.¹

Thermal Efficiency

Of the stations for which temperature normals are available, two (the Chateau Tongariro and Manorburn Dam) are classed as microthermal (C′₂) and the remainder are mesothermal. Of the latter, four stations (Te Paki, Auckland, Gisborne, and Napier) are second category mesothermal (B′₂) and the remainder are first category mesothermal (B′₁). The warmer group of mesothermal climate occupies, for the most part, favourable pockets of the North Island and only in the far north of the country is the delimitation of a continuous area of any size possible. B′₁ is the typical thermal category for New Zealand as a whole and stations fall into this group from the far south almost to the far north. An increase in altitude creates cooler conditions, but it is not until altitudes of above 2,500 feet above sea-level are reached that the evidence warrants the recognition of cooler climates. These are mapped in Figure 5 on the basis of the figures shown in Table I.

As regards the seasonal concentration of thermal efficiency, little need be said. All recording stations in the country indicate a result typical of oceanic situations and are, accordingly, classified as a′ in the fourth letter. This is probably not true of the colder, mountain areas, but no attempt at mapping was undertaken since, in the absence