Diagnosis: Molpadids whose calcareous deposits include two-armed anchors associated with single perforated anchor-plates of varying shapes, usually having three marginal projections. No rosettes of racquet-shaped plates; no fusiform rods. Phosphatic bodies present, at least in adult specimens.

Type Species: Ankyroderma marenzelleri Theel.

Also included here: Ankyroderma tridens Sluiter.

Discussion: Heding (1931) in his subdivision of the genus Molpadia proposed a new genus Pseudomolpadia for those species which have the anchors either united with a single fenestrated plate, or supplied with more than two arms. In this genus Heding placed the following species:

1.	brevicaudata (Koehler and Vaney) 1905 type species.
2.	marenzelleri (Theel) 1886.
3.	tridens (Sluiter) 1901.
4.	inflata (Augustin) 1914.

Subsequently, Deichmann (1936) pointed out that it is only in tridens and marenzelleri that the anchor-plates are definitely known to be "not united in rosettes". Thus brevicaudata and inflata do not belong with the two other species unless it is proved that they have single anchor-plates. As brevicaudata is the type of Pseudomolpadia, this generic name cannot be used here, and it is necessary to propose a new genus, Heteromolpadia, with H. marenzelleri (Theel) as the type species.

Key to the Species of Heteromolpadia

1	(2)	Deposits in the body wall include tables, typically with three perforations	H. marenzelleri (Theel)
2	(1)	No such tables present	H. tridens (Sluiter)

Heteromolpadia marenzelleri (Theel) Plate II

• Ankyroderma marenzelleri Theel, 1886. p. 41. Pl. 3. fig. 1 a-g.
• Molpadia marenzelleri Clark, 1907. p. 171. Pl. 10. fig. 23; Benham, 1909, p. 70, Pl. 11, fig. 4, a-d; Deichmann, 1936, p. 464; Dawbin, 1950, p. 39, Pl. 2, fig. 17; Deichmann, 1960.
• Molpadia dendyi Benham, 1909, p. 71, Pl. 11, figs. 1–3.
• Pseudomolpadia marenzelleri Heding, 1932, p. 280.

Material Examined: VUZ 10, Palliser Bay, 200–250 fathoms, green mud, 1 specimen; VUZ 15, Palliser Bay, 100–150 fathoms, mud, 16 specimens; VUZ 21, Palliser Bay, 38 fathoms, mud, 1 specimen; VUZ 87, South of Cape Palisler, 400 fathoms, mud, rock and gravel, 1 specimen; VUZ 96, off Palliser Bay, 380 fathoms, mud, 13 specimens; Cook Strait, 40 fathoms, 2 specimens, collected by F. Abernethy, 14/11/1952; off Foxton, 50 fathoms, 1 specimen.

New Zealand Oceanographic Institute, Wellington: B 11, Hawke Bay, 35 fathoms, mud, 1 specimen; B 44, Hawke Bay, 14 fathoms, sandy mud, 1 specimen; page 12B 49, Hawke Bay, 44 fathoms, fine grey-green mud, 1 specimen; A 435, off Foxton, 64 fathoms, sandy mud, 1 specimen; C 185, off Wanganui, 25 fathoms, mud, 1 specimen; C 186, off Wanganui, 25 fathoms, mud, 1 specimen; C 189, entrance to Tasman Bay, 30 fathoms, soft mud, 1 specimen.

Diagnosis: Deposits in the body wall comprise spired tables with 3 large perforations, and anchors associated with single three-armed perforated anchor-plates up to 0.4mm in length. Tail deposits lozenge-shaped, 0.1–0.16mm in length.

Description: Short-tailed, fat-bodied holothurians which are greyish-white as juveniles, and gradually become red in colour with growth, until the largest specimens are almost uniformly dark red. The calcareous deposits comprise distinctive anchors, anchor-plates and tables. The tables in the tail persist unchanged throughout life, and are always readily usable as a diagnostic character. The deposits in the rest of the body become gradually transformed into phosphatic bodies with age, and thus they range from well-formed anchors, anchor-plates and tables in smaller specimens to simple, small, concentrically laminated ovoidal red phosphatic granules in larger specimens. Large specimens of the species often lack calcareous deposits altogether, except in the tail, and are dark red is colour, due to the presence of great numbers of phosphatic deposits.

Three groups, based mainly on colour of specimens, may be recognised:

1. Small specimens: 15–30mm in total length. The smallest specimen on hand is 15mm in total length, with a tail length of 6mm. The ratio tail: body in this group is about 1: 3. The body is about twice as long as it is broad. These animals are uniformly greyish-white in colour, and the body wall is quite thin, but opaque. The body is often clothed in particles of mud and sand, which are caught on the arms of anchors and the spires of the tables, as they project above the level of the skin. When touched, the skin gives the sensation of carrying a number of short sharp spines. The inadequate development of the gonads indicates that the specimens in this group may not be sexually mature, and they may be regarded as juveniles.

2. Medium Specimens: 30–70mm in total length. The shape is approximately the same as that in the juvenile, but the tail: body ratio has now become 1: 6. These specimens are orange to dark red in colour, with many greyish spots. The tail and circum-oral regions still retain the grey colour of the juvenile. At this stage in growth, many of the calcareous deposits have been transformed into red phosphatic material, and thus there are but small numbers of anchor arms and table spires projecting from the skin.

3. Large Specimens: 70–101mm in total length. The largest specimen is 101mm in total length, with a tail length of 12mm. The tail: body ratio is here 1:8. In these specimens the body is uniformly dark red, while the tail is grey. In these large specimens, virtually all of the calcareous deposits have been transformed into phosphatic bodies. No anchor arms project above the skin, and the skin is quite smooth and leathery to touch.

The calcareous ring is made up of 10 sculptured pieces, 5 radials and 5 interradials, which are joined to form a solid ring (Plate II, fig. 8). Each radial piece has two rounded anterior projections and a slightly bifurcated posterior projection. There are no perforations for the passage of the radial nerves. The radial pieces each carry a groove for attachment of the radial muscle. Each interradial has a single anterior process, carries a sharp ridge, and has no posterior process. The ring has 15 grooves for tentacle ampullae. Sculpture of the ring varies considerably in this species. The calcareous ring in the juvenile has long and slender posterior projections.

A short thin-walled oesophagus leads into the intestine, which takes a very large loop and runs to the cloaca, which is undifferentiated, save for the numerous very fine muscle strands attaching it to the body wall. These strands also fill the cavity in the tail.

Two respiratory trees arise from the cloaca. Each consists of a single flattened tube which gives rise to a number of short side branches (Plate II, fig. 1). The left tree extends about one third of the way along the body cavity. The right is considerably longer, and runs to the anterior end of the body, attaching to the dorsal pieces of the calcareous ring (Plate II, fig. 1).

The gonads are represented in the mature specimens as extensively branched vesicular caeca, which are arranged in two bunches, lying one to each side of the dorsal mesentery. The caeca are loosely intertwined around and over the intestine and the right respiratory tree. The common genital duct runs anteriorly in the dorsal mesentery to open to the exterior as a well-defined genital aperture in the mid-dorsal interradius, immediately posterior to the ring of tentacles.

The longitudinal muscles are five broad strap-like double bands (Plate II, fig. 1). No "retractor muscles" were seen. Transverse muscles are visible as fine white fibres against the dark coloured background of the body wall (Plate II, fig. 1).

Four types of calcareous deposits are known in this species:

1. Tail Deposits (Plate II, fig. 2): The tail contains a large number of very closely aggregated tables which have elongate discs (0.1–0.16mm long), and carry short three-pillared spires. The discs each have 7–12 perforations. The three pillars of the spire are joined by one or sometimes two crossbars, and the pillars at their distal extremities give rise to a few short spines. Some tables were seen to lack spires and they merely took the form of flat perforated plates. The tail deposits in the juvenile differ little from those in adult specimens. The anus is surrounded by a small number of irregular and distorted tables.

2. Mid-body Tables (Plate II, fig. 6): These tables are small (about 0.15mm across), and they typically have three large perforations. A three-pillared spire with a spinous distal extremity arises from the centre of each table and the pillars are joined to each other by 3–7 crossbars. The tables lie so that the spires project through the skin. Developing tables are common in small specimens (Plate II, fig. 7). In juveniles, the tables are present in large numbers, closely aggregated together, but their numbers decrease with growth of the animals as they become transformed into phosphatic deposits. Medium-sized specimens have scattered tables, while large specimens have very few tables or none.

3. Tables from the extreme anterior end of the body (Plate II, fig. 4): These are elongate tables which often tend towards a fusiform shape, and have 3–7 perforations. The disc is surmounted by a three-pillared spire with crossbars.

4. Anchors and Anchor-plates (Plate II, figs. 3, 9): Anchors with anchor-plates are present in numbers in juveniles, while in large specimens they are absent, due to their dissolution into phosphatic bodies. Medium-sized specimens may still possess anchors and anchor-plates, but when present, they are commonly in the process of dissolution.

The anchors are of varying sizes, lengths ranging between 0.2mm and 0.3mm. The attaching portion of the anchor is saucer-shaped and has 3 perforations. The shaft is straight and cylindrical. The arms are short and curved, and each has two to five small serrations (Plate II, fig. 9).

The very characteristic anchor-plates are not all the same shape, but the basic form is an irregular plate having three (sometimes two) elongate marginal projections. The plate is freely perforated, while the projections each have one or two perforations or none at all (Plate II, fig. 3). Each anchor-plate supports one anchor in such a manner that part of its shaft and arms lie outside the skin. The method of support is simple (Plate II, fig. 3a), so that the anchors are easily detached from their plates. Some anchors were found on their own, holding to the surface of the body by the ends of their arms or by their serrations.

Anchor-plates from medium-sized specimens show various stages in their transformation into phosphatic material (Plate II, figs. 3a, 3b), while juvenile anchor-plates show no trace of phosphatic material (Plate II, fig. 3c). One anchor-plate was found to possess a spire composed of three rods joined by crossbars (Plate II, fig. 3b). This plate had two large perforations and 12 smaller ones. A mid-body table from a juvenile specimen carried a marginal process which showed some resemblance to part of the shaft and arms of an anchor (Plate II, fig. 6a).

Phosphatic Deposits: Apart from the tail deposits which remain unaffected, the anchors, anchor-plates and tables in turn become transformed into phosphatic spherules with advancing age. It is therefore possible to encounter specimens which lack anchors, anchors and anchor-plates, or anchors anchor-plates and tables. Transformation into phosphatic material is a gradual process and one often sees deposits which are in the process of dissolution (Plate II. figs. 3a, 3b, 10).

page 15

The resulting phosphatic bodies are amber or red, ovoid to spherical, and they superficially resemble starch grains (Plate II, fig. 5).

Knowledge of the changes in deposits with growth serves to explain clearly the differences in colour between small, medium and large individuals of this species. The greyish-white juveniles have very few phosphatic bodies and many calcareous deposits. The medium-sized specimens, dark red with greyish spots, have clusters of phosphatic bodies, and the greyish patches represent areas where calcareous deposits still remain. The uniformly dark red large individuals have great numbers of phosphatic bodies and very few calcareous deposits.

Distribution: Theel (1886) described the type specimen from east of East Cape in 700 fathoms. Benham (1909) recorded specimens of H. marenzelleri from 38 fathoms in Hawke Bay, and Molpadia dendyi from deeper water off the coast of the North Island. The new localities recorded here indicate that H. marenzelleri is a common species about the deeper waters of the southern half of the North Island. As the species is eurybathic, it probably has a wider distribution.

Ecology: This species lives on a muddy or sandy bottom.

Discussion: The status of this species has been in doubt for some time, owing to insufficient knowledge of the juvenile and its deposits. The present findings indicate that H. marenzelleri is a valid species, characterised by the peculiar anchor-plates.

Molpadia violacea (Studer) Plate III, figs. 4–8

• Trochostoma violaceum Studer, 1876; Theel, 1886, p. 42, Pl. II, fig. 4; Pl. XI. fig. 1.
• Molpadia musculus H. L. Clark, 1907, p. 165, Pl. XI.
• Haplodactyla violacea Heding, 1931, p. 280.
• Molpadia violacea Deichmann, 1960.

Material Examined: VUZ 87, South of Cape Palliser, 400 fathoms, mud, 1 specimen; VUZ 101, off Palliser Bay, 550 fathoms, mud, 6 specimens.

Diagnosis: Deposits in the form of large fusiform rods with two to three arms, up to 1.1mm in length. No anchors or anchor-plates. No rosettes of racquet-shaped plates. Tail deposits two-armed fusiform rods up to 0.8mm in length. One anterior process on each radial piece of the calcareous ring perforated for the passage of the radial nerve.

Description: The smallest specimen has a total length of 47mm; the largest is 78mm in length. The specimens are approximately cylindrical in shape, elongate, with the posterior end attenuated to form a distinct caudal appendage which occupies up to 20% of the total body length.

Colour in alcohol ranges from a light-brownish red to a dark brick-red. The anterior extremity of the body and the tail are greyish-white. The skin is quite thin and coarse to touch.

The calcareous ring is composed of 10 pieces, five radials and five interradials, joined to form a solid ring. Anteriorly, the radials each have two short and blunt processes, one of which carries a small perforation for the passage of the radial nerve. The interradials each have one anterior process and no perforation. The radials have a forked posterior process, while the interradials have none. The ring is sculptured on its outer surface, and the sculpture varies within the species.

The internal anatomy is similar to that in Heteromolpadia marenzelleri,

page 16

Two types of calcareous deposits are present in the skin:

1. Fusiform Rods: These are found everywhere in the body wall, especially in the tail, where they occur in great numbers, closely aggregated together, lying transverse to the longitudinal axis of the body. The rods vary in length up to a maximum of 1.1mm, and they have an expanded central portion which carries a small number of perforations.

(a) Rods from the tail (Plate III, fig. 6): The tail rods are in general smaller (average length 0.7mm) than those from other areas, and have fewer perforations. The ends of the rods tend to project above the level of the skin, and they can be seen with the naked eye.

(b) Rods from the posterior third of the body, near the tail (Plate III, fig. 8): These are massive deposits, with an average length of 1.0mm. Many of the rods have three arms, while others have two, and there are four to eight central perforations. The rods are grouped into small clusters.

(c) Rods from the middle of the body (Plate III, fig. 7): Mid-body deposits closely resemble those from the posterior third of the body in general features, but even more variability in shape is displayed.

(d) Rods from the extreme anterior end of the body (Plate III, fig. 4): These are similar to those from the tail, and are of a comparable size.

2. Perforated tables (Plate III, fig. 7a): Tables (length 0.3–0.6mm) with short central spires and 3–6 perforations are scattered sparsely in the skin. The spire is composed of a single column, and in many cases it is absent. Developmental stages are occasionally seen (Plate III, fig. 5).

Red phosphatic deposits are present, grouped together in small clumps. They are similar to those in Heteromolpadia marenzelleri (Theel).

Distribution: Theel (1886) described specimens taken from the vicinity of Kerguelen Island at depths between 20 and 120 fathoms, and from about 50 miles east of East Cape, New Zealand in 700 fathoms. The two new localities recorded here, Palliser Bay, 550 fathoms, and south of Cape Palliser, 400 fathoms, indicate that this species may be relatively common in deeper waters about New Zealand, and probably elsewhere, achieving its distribution by spreading across the seafloor in deep water.

Discussion: The specimens described here are similar in most respects to those described and figured by Theel (1886). There appears to be a complete lack of anchors and rosettes, even in smaller specimens. Thus the species is sharply distinguished from M. musculus (Risso). Deichmann (1960) believes that M. violacea is an extreme form of M. musculus with narrow rods and lacking the anchor and racquet stage completely.

Diagnosis: Tentacles with two pairs of digits. Caudal appendage usually long and slender. Deposits not tables but cups (buttons), perforated plates or irregular rods (Heding, 1931, in part).

Type Species: Paracaudina chilensis (Muller).

Heding (1931) diagnosed this genus and listed seven species, one of which (pigmentosa Perrier) was included with some doubt. Clark (1935) agreed with Heding's new genus, but did not agree with the part of Heding's diagnosis which stated "retractor muscles more or less well developed". After examination of many specimens of Paracaudina, Clark found no true retractor muscles.

Paracaudina chilensis (Muller) Plate IV

• Molpadia chilensis Muller, 1850, p. 139; 1854, Pl. VI, fig. 14, Pl. IX, fig. 1.
• Molpadia coriacea Hutton, 1872, p. 17; Hutton, 1878, p. 307.
• Caudina meridionalis Bell, 1883, p. 58. Pl. XV, fig. 1.
• Caudina coriacea Theel, 1886, p. 47, Pl. III, fig. 4; Dendy, 1896, p. 28, Pl. 3, figs. 9–18; Dendy, 1897, p. 456, Pl. 29; Farquhar, 1898, p. 324; Ludwig, 1898, p. 63; Dendy and Hindle, 1907, p. 108; Mortensen, 1925, p. 363, figs. 46–47.
• Caudina pulchella Perrier, 1905, p. 117, Pl. V, figs. 14–17.
• Caudina coriacea var. brevicauda Perrier, 1905, p. 121.
• Caudina chilensis H. L. Clark, 1907, p. 175; Benham, 1909. p. 28; Hozawa, 1928, p. 363; Ohshima, 1929, p. 39.
• Pseudocaudina coriacea Heding, 1931. p. 283.
• Paracaudina coriacea Heding, 1932, p. 455; Heding, 1933, p. 127, Pl. IV, figs. 8–13, Pl. VII, figs. 6–7, Pl. VIII, fig. 4; Dawbin, 1950, p. 39, Pl. 1, fig. 5, Pl. 2, fig. 16.
• Paracaudina chilensis var. coriacea H. L. Clark, 1935, p. 267.
• Paracaudina chilensis forms coriacea Deichmann, 1960.

Material Examined: VUZ 15, Palliser Bay, 100–150 fathoms, mud, 2 specimens; VUZ 87, South of Cape Palliser, 400 fathoms, mud, 2 specimens; VUZ 101, off Palliser Bay, 550 fathoms, mud, 135 specimens.

New Zealand Oceanographic Institute, Wellington: Stn. B 8, Hawkes Bay, 39° 06′ S., 177° 23′ E., 15.5 fathoms, 26/8/56, fine grey sand, 326 specimens.

Diagnosis: Body cylindrical, attenuated posteriorly into a long tail. Colour in alcohol white; old specimens frequently light brown. Calcareous deposits in the form of thick, solid crossed cups with small perforations, the marginal projections when present being low and rounded. The cups, especially in young specimens, approximately octagonal in shape, while the points of the octagon may be obscured in old specimens. Diameter of cups 0.06–0.1mm.

Description: These are caudinids whose tail length is about 40% of the total body length. Total length of largest specimen 115.0mm, tail 44.0mm, diameter of body at widest point 37.0mm. Total length of a juvenile specimen 15.0mm, tail, 6.0mm, diameter at widest point 6.0mm.

In all specimens the body is cylindrical, tapering abruptly to form a long tail. The body wall is thin and firm, and is marked by numerous transverse striations. In juveniles the body wall is semi-transparent, and through it the gut can be seen as a dark coloured mass. Colour in life and in alcohol varies from white to light yellow or brown. Anal papillae are present. The mouth is circular, lying in the middle of a circular oral disc. Tentacles 15, usually retracted.

The calcareous ring comprises ten pieces, five radials and five interradials. The radials each have a bifurcated posterior projection and three anterior projections. The interradials each have no posterior projection and one anterior projection. Many workers (Hozawa, 1928; Heding, 1933; Clark, 1935) have described the calcareous ring in detail.

page 19 page 20

A short thinwalled oesophagus meets the intestine which takes a large loop (Plate IV, fig. 1), and joins the cloaca, which is attached to the body wall by fine muscle strands which also fill the cavity in the tail. In the juvenile the intestine fills most of the body cavity, and the cloaca has a small number of muscle strands.

The gonad consists of two bunches of irregularly branching vesicular caeca (Plate IV, fig. 1). The bunches join to form the common genital duct which runs anteriorly in the dorsal mesentery, and opens to the exterior as a minute genital pore. The juveniles show no trace of gonads.

A single bulbous and elongate Polian vesicle, up to 10mm in length arises from the right or left ventral side of the water vascular ring. The vesicle has a small patch of dark brown pigment at its distal extremity. A short coiling stone canal lies in the dorsal mesentery, terminating in a madreporite which takes the form of a depressed sphere (Plate IV, fig. 5). The madreporite has a sculptured surface, due to the presence of an investing layer of irregular intertwining deposits (Plate IV, fig. 7). In the juvenile the madreporite has a diameter of about 0.2mm, and the deposits are essentially the same as those in the adult, but they do not constitute a full enveloping network.

Respiratory trees are composed of two main trunks which carry numerous small side branches (Plate IV, fig. 1). The right tree is considerably longer than the left, and extends to the level of the calcareous ring, to which it is attached, while the left tree extends about half way along the body cavity. In most specimens the left tree gives rise to lesser trunks which tangle about the intestine, and lie in association with the rete mirabile.

The longitudinal muscles take the form of five double bands (Plate IV, fig. 1). There are no retractor muscles. In juvenile specimens, the muscle bands are very thin and straplike, and between each member of a pair of double bands the radial longitudinal nerve can be clearly seen as a thin white line. Transverse muscles are visible as fine lines (Plate IV, fig. 1).

Calcareous deposits: In adult specimens the deposits are "crossed cups" of varying shapes. These are crowded together in vast numbers in the skin, from the extreme anterior end of the body to the end of the tail. These cups are 0.06–0. lmm across. They each consist of a "cross" which overlies a "square" (Plate IV, fig. 2). The "square" has a single large perforation which is usually rectangular, with a tendency to become circular. The margin of the "crossed cup" is approximately octagonal in shape, but in adult specimens this shape is often obscured, as the points of the octagon become rounded off. Marginal projections are usually present as low rounded knobs. In the tail the cups are slightly more irregular in shape than those in the rest of the body wall.

The cups in the juvenile are more angular in outline (Plate IV, fig. 2) than those in the adult (Plate IV, fig. 3), and the typical "crossed cup" structure is more readily observed. Half-grown specimens show a mixture of angular and rounded deposits.

Stages in the development of "crossed cups" are readily observed in juveniles. The "cross" is the first to form (Plate IV, fig. 6a). It is a simple four-armed cross, each arm measuring about 0.008mm in length. A perforated square then develops on the cross (Plate IV, fig. 6b). The cross is invariably the starting point in the development of crossed cups. The extremities of the cross and the corners of the square then begin to expand laterally, and these lateral expansions eventually meet to form the "young" deposit which has smooth rounded edges (Plate IV, fig. 6e). The "young" deposits then assume the "classical" form (Plate IV, fig. 6f), with eight sharp projections regularly spaced around the margin, and a few small spines (3–8) on the cross side. In most cases the cross faces the outside of the body, and the short sharp spines project slightly above the level of the skin.

The anal papilla deposits of the juvenile are similar to those in the adult. They arc small, irregular spicules, and take the form of branched rods or perforated plates (Plate IV, fig. 4).

Ecology: Paracaudina chilensis has been taken in the New Zealand region from muddy and sandy localities to depths of at least 550 fathoms. Some specimens have been recorded from fish stomachs, but it is not known how extensively the species is used as food by bottom-feeding fish.

Discussion: Dendy (1897) gave a very thorough account of the structure and disposition of the anal papillae in this species.

Clark (1907) placed eight species of Caudina Stimpson into the single species Caudina chilensis (Muller), as many of the original species descriptions had page 21been inadequate, based as they were on such characters as size of the specimens, colour, texture of the body wall, all of which are known to be subject to much individual variation. In this synonymy Clark included Hutton's (1872) species coriacea from New Zealand and australis (Semper) from Australia.

Mortensen (1925) criticised Clark's synonymy and declared that C. coriacea from New Zealand, C. australis from Australia and C. chilensis from Chile were different species, and he used apparent differences in spiculation and calcareous rings as his evidence. At the present time, C. australis is still regarded as a distinct species. But the history of C. coriacea is rather more complex.

Hozawa (1928) regarded C. chilensis and C. coriacea as the same species using Clark (1907) as his authority. He may not have seen Mortensen's (1925) paper. Ohshima (1929) agreed with Clark (1907) and Hozawa (1928), and criticised the work of Mortensen, stating that his figures were inadequate. Heding (1932) accepted Mortensen's view and included C. coriacea as a separate species in his new genus Paracaudina, together with C. chilensis. Later Heding (1933) vigorously opposed Ohshima's (1929) opinion, and used the same characters as Mortensen (1925) for distinguishing the species coriacea and chilensis, but on a much more elaborate scale. He used characters such as body form, "retractor muscles", genital papillae, and presence or absence of "Cuvierian organs" as additional evidence. Thus Paracaudina chilensis was re-established as a separate species, but Clark (1935) "re-entered the lists" in his own words, after examining a great number of specimens of Paracaudina. His paper shows that he disagreed with Mortensen (1925) and Heding (1933). He discarded body form, "retractor muscles", "Cuvierian organs", genital papillae and the calcareous ring as bases for classification, and stated that the spicules were the only safe criterion for separation at the species level. As a result of his thorough studies Clark compiled a key to the species in genus Paracaudina, and named the New Zealand form Paracaudina chilensis var. coriacea, adding that Deichmann was in agreement with him. Deichmann (1960) suggested that the New Zealand form be named P. chilensis forma coriacea as, in the words of W. K. Fisher, "it does not protest too much".

I have examined only the New Zealand specimens of the genus Paracaudina and they display some considerable variation in their calcareous deposits. Comparison of these deposits with those figured by Hozawa (1928) and Heding (1933) has served to convince me that they resemble each other in so many features, and show such diversity of form, that the subdivision of the species chilensis into subspecies or even "forms" is unwarranted. Clark (1935) himself stated that if he had a specimen from Chile mixed with specimens from another area he would not be able to identify the Chile specimen with certainty.

Thus the suggestion lies at hand that P. chilensis is a circum-Pacific species, having possibly the Indo-West-Pacific region as its centre of distribution. Near the centre of distribution, the genus Paracaudina gave rise to tetrapora and australis, now in Australia, and to chilensis, which spread north to Japan, and to California and Florida via the Aleutian Islands, and south to New Zealand, leaving a remnant in North-west Australia. The Chilean representatives may have reached South America via New Zealand. Fell (1953) states that it is quite likely that New Zealand supplied contributions to the fauna of southern South America. He does not propose an Antarctic shoreline as does Deichmann (Clark, 1935), but indicates that the west to east circum-polar current may be responsible for this New Zealand affinity in certain elements of the South American fauna. The gap in the distribution of P. chilensis lies between California in the north and Chile in the south. The gap may possibly be due to unfavourable environmental conditions, or the species may still be undiscovered there. As P. chilensis is eurybathic to a certain degree, there should be few depth barriers to dispersal.