Introduction

The athecate hydroids of the family Solanderiidae Marshall, 1892 were until recently (Vervoort, 1966) regarded as unique in possessing a chitinous skeleton in the form of a network of longitudinal and transverse connecting fibres that originated in the mesoglea (Rees, 1957; Vervoort, 1962). Vervoort clearly showed in his 1966 paper, however, that the lattice work skeleton is formed as are all other hydroid skeletons by ectodermal cells. Solanderid hydroids range from cushion-like forms to tall, much branched colonies. The polyps have capitate tentactles. Six species are recognised, and a further nine have been given the title "doubtful species" by Vervoort (1962). These doubtful species are those which have not been fully described, and although they are probably members of the recognised species they cannot, at present, be placed in them. The various species have been recorded from localities including the West Indies, Australia, Japan, South Africa, and from other subtropical and temperate waters. Until the present paper one species only has been recorded from New Zealand. This is Chitina ericopsis described from the skeleton by Carter in 1873. The collection locality is stated simply as New Zealand. The skeletons of several colonies were dredged from soft mud in Wellington Harbour some ten years ago, but the first colonies from New Zealand waters with identifiable soft parts page 2are those described here from Auckland, collected in 1960. This year (1967) the largest complete New Zealand colony was dredged from bottom mud at 8 fathoms in Wellington Harbour.

Early systematic studies were based mainly on the growth form of the colonies and on the structure of the skeleton, in particular the structure of the hydrophore or "supporting cup" which is found at the base of polyps in certain species.

Materials

The material on which this study is based is a collection of five colonies preserved in formalin, from 4 to 12 cms high, taken from Auckland Harbour, N.Z., and two colonies, one 30 cms high and one 6 sms high, taken from Wellington Harbour, N.Z. Specimens of Solanderia fusca (Gray, 1868) from New South Wales and Queensland, Australia, and a specimen of Chitina ericopsis Carter, 1873 were studied to aid in identifying the Auckland and Wellington Harbour species which unfortunately were devoid of reproductive organs, so that descriptions of these organs were not made.

Classification of the New Zealand Material

The colonies from Auckland and Wellington Harbours are erect and have no hydrophore or supporting structure of any kind at the base of the polyps. This eliminates them from many of the species described by Vervoort (1962), but leaves them as possible members of the following genera and species.

(1)	Solanderia gracilis Duchassaing and Michelin, 1846 (distribution West Indies).
(2)	Solanderia leuckarti Marshall, 1892 (described as a "doubtful" species by Vervoort, distribution unknown, but probably the Pacific).
(3)	Chitina ericopsis Carter, 1873 (a "doubtful" genus according to Vervoort, distribution New Zealand).
(4)	Solanderia misakinensis (Inaba, 1892), found in Japanese waters.

The colony growth form of the Auckland specimens used in this study is identical with that of S. gracilis except that the polyps on the branch tips are borne in all planes, and not only in the plane of flattening of the colony. The skeleton of S. gracilis is purple coloured, but the skeletons of the Auckland and Wellington specimens preserved in formalin are a uniform light brown.

The present material lacks tubercles of fused fibres on fine branches such as are described for S. leuckarti.

A portion of a branch of Chitina ericopsis (dried material only) cannot be distinguished in structure from that seen in a branch of equivalent dimensions of the present material. Chitina may therefore be inseparable from Solanderia, as Vervoort suggests (1962, p. 537). However as no description of soft parts of Chitina exists, it is difficult to see how any living or well-preserved specimen can ever be assigned to this genus.

S. misakinensis, as described by Vervoort (1962) possesses a much closer, denser form of branching than the Auckland specimens, but is identical to the largest (30 cm high) Wellington Harbour colony. Growing from the same tangled root-like base of the Wellington material a little to the side of the large colony is a colony 6 cm high, and this smaller one has the same growth pattern as the Auckland Harbour specimens. There is little doubt that the Auckland and Wellington specimens are members of the species S. misakinensis, and that the Auckland material is at an earlier stage of development than that which is figured and described by Vervoort.

page 3

In the course of identifying the present material two specimens labelled as Solanderia fusca (Gray, 1868) were studied. One was collected in Sydney, Australia, (Australian Museum, No. Y.509) and the other from Queensland, collected by Dr. Pamela Pennycuik. The specimen from Sydney is undoubtedly S. fusca and agrees in every detail of skeletal structure with the specimen Ceratella fusca Gray, 1868 described by Spencer (1891). The specimen from Queensland, however, is certainly not S. fusca. The skeleton is very similar to that of Solanderia secunda (Inaba, 1892) with a flattened, thorn shaped projection either side of the polyp base and no hydrophore as in S. fusca. The colour of the Queensland specimen is dark brown as in S. secunda, and the method of branching is similar. As the distribution of S. secunda is tropical and subtropical regions of the Pacific Ocean (Vervoort, 1962) it seems very likely that the Queensland specimen is in fact S. secunda.

Methods

Several tips of branches about 1 cm long were stained in Delafield's haematoxylin and embedded in wax. Some of these blocks were cut to give transverse serial sections from the stem tip, and others were cut to give longitudinal serial sections. The sections were placed on slides in series, over-stained in Delafield's haematoxylin and differentiated in acid alcohol, countersigned in eosin (0.5% soln. in 90% alcohol) and mounted in DPX mountant. Most sections were cut at 10μ on a rocker microtome, but one transverse series was cut at 5μ. A thicker piece of stem (1cm × .5mm × .4mm) was embedded, cut in transverse section at 10μ and stained and mounted as above.

In addition, a 1cm length of stem from the tip of a branch was denuded of soft tissues by dipping into "Janola" (a commercial solution of sodium hypochlorite), and the remaining skeleton was washed and stained in a 0.5% soln. of lignin pink in cellosolve for 48 hours. After this time the skeleton was mounted in DPX for observation. Three lengths of stem from the tip to the thicker parts of the branch were also rid of soft tissues by the same method and were stained in the lignin pink soln. for ½ hour. After this time they were removed to absolute alcohol.

From one series of transverse sections of a branch tip, a model of the soft parts (endoderm only) was constructed to illustrate the method of branching of coenosarc tubes within a branch tip. (It was assumed that the growth pattern in a branch tip is indicative of the growth pattern of the colony as a whole).

Morphology

External features: The Colony (Plate I, Fig. 1).

The colony is erect, arising from a flattened, root-like base from which is formed a strong stem. This stem branches irregularly forming large and small branches which, near the base of a colony, have the appearance of being formed of intertwined branches, and which may anastomose if they come into contact. Branching in small colonies (4-12cms high) is in one plane only, so that the colony has a flattened, fan-shaped appearance. Larger colonies have a more bushy appearance. The colour of the colony skeleton is an overall brown, but is paler on the small branch tips. The polyps (white in preservative) are borne on all branches, but do not arise directly from the tip of any of the branches. They are elongate (up to 2½ mm in length in the preserved state) naked cylindrical structures with a conical hypostome. The base of the polyp is without hydrophore or supporting bract of any kind. Tentacles are given off throughout the length of the polyp and normally appear irregularly placed, but in a well extended polyp are seen to be arranged in a spiral at intervals along the polyp, except for an oral whorl of four tentacles.

page 4 page break page 5

Internal structure: The Polyp (Plate I, Figs. 2, 3; Plate 2, Figs. 1, 2).

In longitudinal section (Plate 1, Fig. 2) the polyp has three distinct regions, namely (a)—the hypostome, in which the main feature is the very strongly developed endothelial muscular cells; (b)—a secretory region of from ⅓ to ⅔ the length of the polyp consisting of many cells opening into the coelenteron and its branches [these cells are packed with acidophilic globules (Plate 2, Fig. 1)]; and (c)—at the base of the polyp there is a region of large vacuolated endodermal cells which extends for approximately a third of its length. The nuclei and cytoplasm of these cells are distally placed, and each cell usually subtends, at its base, two or three ectodermal cells.

The ectoderm of the polyp is somewhat cuboidal and is continuous with that over the surface of the whole colony. In extended polyps it is drawn out into a very thin layer (Plate 1).

The mesoglea is prominent in retracted zooids, but is thin when the zooids are extended. It is light staining, and appears structureless. In retracted polyps a very large number of cut ends of myonemes from the ectodermal cells are seen embedded in the mesoglea, and myonemes can be seen in the ectodermal cells.

The tentacles have a solid core of vacuolated endodermal cells and are capitate. Batteries of nematocysts occur in their swollen "cap-like" distal ends (Plate 2, Fig. 2). Nematocysts are sometimes seen in the ectoderm towards the base of a polyp, but apart from this are confined to the tentacles. Only stenoteles (penetrating nematocysts) have been seen, and they are usually about 5μ long, but often may be as large as 10μ long.

The Branch Tip (Plate 3, Figs. 1, 2, 3, 4).

Thicker branches of the colony give the appearance of being formed of intertwined smaller branches. The very tips of branches, although of small diameter show a similar structure when seen in transverse section.

In Plate 3, Figs. 1 and 3 are transverse sections through a typical branch tip and are 10μ apart. The stem or branch tip is seen to be composed of several coenosarc tubes. There is a central tube (that is endoderm, mesoglea, and ectoderm) and radially-placed coenosarc tubes.

Between these tubes are the skeletal elements. The skeleton is not formed around the whole circumference of the stem, but only between the inner ectoderm layer of adjacent coenosarc tubes (Plate 2). The whole stem is covered with a columnar epithelium.

The "inward facing" layer of the ectoderm of the outer coenosarc tubes, and the ectoderm of the central coenosarc tube are composed of vacuolate cells with the cytoplasm and nucleus distally placed. These ectodermal cells are clearly delimited from the endoderm of the tubes by a thin layer of structureless, lightly staining mesoglea. Small eosinophil particles are frequently present in the ectodermal cells, especially those cells which abut against the skeleton. The coelenteron of the coenosarc tubes appears as narrow branching channels (Plate 2). Gland cells packed with large eosinophil globules are seen in the endoderm of the tubes, especially in thicker stems, even though the coelenteron cannot be seen. In some transverse sections of a branch tip the coelenteron can be followed through the polyp to the endoderm of a coenosarc tube, but in the latter the coelenteron is not prominent. The "closing-over" of the coelenteron in the coenosarc tubes is probably due to contraction during preservation.

page 6 page 7

The ectodermal and endodermal cells are markedly different. Endodermal cells take stain more readily than do ectodermal cells; they also differ in the characteristics of their nuclei. Both the outer columnar and the inner secretory ectoderm cells have round nuclei of about 3μ. diameter, while endodermal nuclei may be up to 6μ diameter and are often irregular in outline. The nucleus of the ectodermal cells (Text Fig. 1b ) has a purplish staining nucleolus which is surrounded by about six small, darkly staining, chromatin bodies. From each of these small chromatin bodies there is a fine darkly staining radial process which connects with a small peripheral body. The nucleoplasm is irregularly granular. The nucleus of the endodermal cells (Text Fig 1a ) has a nucleolus which stains a brighter red, and fewer darkly staining bodies. But these bodies are larger than in the ectodermal cell nucleus, and are found irregularly scattered throughout the nucleoplasm. The nucleoplasm of the endodermal cell nucleus is much less granular than that of the ectodermal cell nucleus.

Bipolar and multipolar nerve cells occur between endoderm and ectoderm of the coenosarc tubes. The nucleus of these cells occupies the greater part of the cell body and stains very darkly. The cytoplasm is also deeply stained, and is agranular.

Although the outer ectoderm appears to be free of nematocysts, except at the base of a polyp, nematocysts occur in the internal etcoderm. In the latter cell layer, they total about 40% of the ectodermal cell number in thicker branches (Plate 2, Figs. 3 and 4). Both large (10μ) and small (5μ) stenoteles are present, and many stages of development can be seen.

page 8

The Skeleton (Plate 1, Fig. 4; Plate 2, Fig. 3).

The skeleton of a branch tip is a network of "chitinous" fibres (Plate 1, Fig. 4). There are major fibres running longitudinally which curve to meet at the tip of the branch. These main fibres are joined by struts which are flattened in the direction of the branch and which occur about 0.09 mm apart. Combining the information given in the accompanying Plates, the skeleton of a branch tip can be described as a network cylinder surrounding the central coenosarc tube, with radially arranged, longitudinal strips of chitinous material which are flattened laterally, and joined to the central cylinder at intervals by transverse struts. These separate the peripheral coenosarc tubes from each other. In a well preserved specimen the whole skeleton is enclosed by soft tissue.

The skeletal fibres are of lamellar construction (Plate 2, Fig. 3). A branch tip skeleton had to be left in a 0.5% solution of lignin pink for 48 hours before it stained evenly throughout its length. In three pieces of skeleton taken in series from a branch tip, the colour deepened greatly away from the tip of the branch when they were all stained for ½ hour. The tip stained very faintly, while the furthermost piece stained deeply.

Discussion

The fact that the skeleton of the Solanderiidae is internal and is completely covered by a layer of epithelial cells in the living animal has undoubtedly led to the earlier belief that the skeleton is mesogleal in origin. However, it is evident from this study that each branch of a colony is composed of several coenosarc tubes. The appearance given in cross sections of a branch tip (Pl. 3, Figs. 1 and 3) is very similar to that of a cable containing many wires, cut in cross section. Each of the coenosarc tubes consists of endoderm, mesoglea, and ectoderm, and therefore could be expected to be capable of producing a perisarc, just as are other skelton-bearing hydroids.

The outer epithelium of the colony probably is formed by fusion of the superficial layer of the ectoderm of the peripheral coenosarc tubes in each branch. This idea is given weight by the fact that the outer epithelium can be seen to be continuous with "inner" ectoderm of the coenosarc tubes, especially at polyp bases (Pl. 2, Figs. 3 and 4).

The "inward facing" ectoderm of peripheral coenosarc tubes and the ectoderm of the other coenosarc tubes contained in a stem apparently become modified into a secretory tissue which forms the skeletal fibres. Skeletal material is secreted only where the ectoderm of two or more coenosarc tubes make contact. The skeleton is thus a "perisarc" contributed to by two or more coenosarc tubes. The cross sections shown in Plate 3 support this interpretation, and there is no doubt that the skeleton is ectodermal in origin. The skeleton may differ in composition in young and older parts of the stem. Lignin pink, which is considered a specific stain for chitin, does not stain the skeleton of branch tips as readily as it does the skeleton from thicker parts of the stem.

The method of growth of the colony can be seen from serial sections of a branch tip. From the tip to the thicker parts of a branch there is a central coenosarc tube, with peripheral tubes which anastomise with the central tube and with each other (but the central tube anastomoses far more frequently with the peripheral tubes than do the latter with each other). The polyps communicate with more than one coenosarc tube at their base (Plate 3, Fig. 1), and this base may be very large and extend over two or more skeletal "holes". (Plate 2, Figs. 3 and 4).

A serial reconstruction of the soft parts (endoderm only) (Text Fig. 2) shows these and other features of a stem tip. The central coenosarc tube has no terminal polyp. Some peripheral tubes may have a terminal polyp, but others do not. This type of anastomosing of coenosarc tubes is intermediate between that described for Clathrozoon wilsoni and Plumularia procumbens by Spencer in 1890, although there are fewer tubes in any one branch tip.

page 9

The skeleton shown in Text Fig. 2 was not reconstructed from serial sections. Skeletal rods were drawn between the peripheral endoderm tubes and then joined through the holes between the anastomosing tubes. This showed a skeleton of a branch tip similar to that described from serial cross sections as illustrated in Plate 3.

The stem is clearly a fasicled or bundled structure. But the status of each of the elements composing it is more difficult to define. The alternative possibilities are firstly, that the central coenosarc tube is equivalent to a hydrocaulus and is surrounded by its branches and stolons, which anastomose with it and with each other, and secondly, that the whole stem is a rhizocaulome, being composed of upright, branching and joining stolons.

Two features indicate that the first alternative is the more likely because (a) the stem tip clearly has a central coenosarc tube, and (b) the fact that this central tube anastomoses much more frequently with the peripheral tubes than the peripheral ones do with themselves. This suggests that the central tube is a primary structure, and the others are secondary to it.

The method of growth seems to be "secondary monopodial" (Hyman 1940, p. 406) in which the hydrocaulus and branches end in non-polyp bearing growing points. The growing points elongate and bud off laterally both polyps and new branches (this method is more typical of thecate hydroid colonies) — (Hyman 1940, p. 405). The picture is obscured somewhat, however, by the hydrorhizal elements of the stem.

page 10

Summary

(1)	The status of the New Zealand species of the family Solanderiidae is discussed, and material from Auckland and Wellington Harbour is recognized as Solanderia misakinensis (Inaba, 1892).
(2)	It is concluded from a study of the internal morphology of the Auckland Harbour specimens that, (a) the skeleton is of ectodermal origin (b) the branches of the colony are fasicled structures consisting of a central coenosarc tube (corresponding to a hydrocaulus) which branches, and which anastomoses with its own branches and with hydrorhizal elements of the stem. These latter elements also branch and anastomose with each other.
(3)	The method of growth of the colony is "secondary monopodial". That is, the hydrocaulus and its branches end in permanent growing points, and elongate by budding off polyps and other branches laterally.