tomyxoidea Tripathi in which the sporoplasm lacks an iodinophilous vacuole, and Myxoboloidea Tripathi, in which an iodinophilous vacuole is present in the sporoplasm. There is but one family in the sub-order Bipolaria. The superfamily Ceratomyxoidea is subdivided into four families, and Myxoboloidea into two, on the basis of the number of polar capsules in the spore.

Thus far, I favour Tripathi's classification over Kudo's. It not only does away with the anomalies involved in making the mere shape of the spore a subordinal criterion, but finally abolishes the use as family criteria of the microhabitat of the parasite and the ecology of the host—Kudo (1933) included two families in his sub-order Eurysporea, defining Ceratomyxidae Doflein, 1899, as “Typically coelozoic parasites of marine fish” and Wardiidae Kudo, 1933, as “Histozoic or coelozoic parasites of fresh-water fish.” Other considerations aside, it is obviously most undesirable to be unable to assign certain myxosporidians to such a major group as a family without full information on their microhabitat or the ecology of their hosts.

Furthermore, Kudo's (1933) classification left certain genera of obviously close affinities in different sub-orders. Thus Sinuolinea Davis (Sphaerosporea: Sphaerosporidae) has been used (in part) as a repository for forms having very close affinities with Ceratomyxa Thélohan (Eurysporea: Ceratomyxidae). By Tripathi's classification, Ceratomyxa and Sinuolinea are considered to belong to the same family. Considering that Tripathi's classification of the Myxosporidia is mechanically superior to Kudo's and affords a more satisfactory framework within which to demonstrate the relationships and affinities existing within the order, I have adopted it in this paper.

Sinuolinea Davis, 1917.

Davis defined this genus (the recognised species of which are listed in Table VI) as follows:—“Spores approximately spherical; with or without lateral processes. Capsules rounded, not convergent when seen from above; capsular pores some distance apart, sometimes on nearly opposite sides of the spore. Sutural line forming a prominent ridge, which takes a sinuous course around the spore. Sutural plane usually distinctly twisted on its axis. Disporous and polysporous.”

In all cases the organ infected is the urinary bladder, although S. murmanica has also been recorded from the kidney of the host (Basikalowa, 1932).

Seven of the species—S. dimorpha, S. capsularis, S. arborescens, S. gilsoni, S. murmanica, S. cyclopterina and S. rebae—have approximately spherical spores in which the suture line is markedly twisted on its axis. The other four—S. opacita, S. brachiophora, S. bidens and S. cella—are quite distinctive in that their spores have arm-like lateral extensions. The sutural line is only slightly curved in spores of S. brachiophora and S. cella, but is markedly twisted in those of S. opacita and S. bidens.

With regard to S. brachiophora, Davis (1917) remarked that “Possibly this species should be made the type of a new genus, but the spore undoubtedly more closely resembles that of Sinuolinea than any other genus. In many respects this species is very similar to S. opacita, which occurs in the same host.” Jameson, noting these remarks, held that there are insufficient grounds for considering Davis's three species having spherical spores as distinct from Sphaerospora Thélohan. He claimed that “the only character that the spherical members of the genus Sinuolinea possess that is not present in some species of Sphaerospora is the sinuous suture line and that is hardly a sufficient ground for creating a new genus. Such a character is found in the genus Myxidium and is regarded as being of only specific worth.” Jameson (who failed to notice S. gilsoni) thus considered that S. dimorpha should revert to the genus Sphaerospora in which Davis (1916) had originally placed it, and further elected to transfer S. capsularis and S. arborescens to that genus. He retained Sinuolinea for S. opacita and S. brachiophora, and described his S. bidens and S. cella in this same genus.

Sinuolinea, as emended by Jameson, contains those myxosporidians the spores of which have “first a clearly walled off, small spherical or oval central chamber in which the sporoplasm and pole capsules are located, and second a pair of prominent lateral appendages attached to this. The suture line is slightly curved but this is possibly not of great importance.”

Jameson considered that the acceptance of his emended definition of Sinuolinea would necessitate the transfer to it of Ceratomyxa spinosa Davis, 1917, the only species of that genus in which the spore has a clearly demarcated central chamber. Several other species of Ceratomyxa have laterally extended shell valves, but the spore cavity is continuous with those of the extensions.

Kudo (1933) reviewed the classification of the Myxosporidia, emending the definitions of certain of the genera and listing all the species known at that time. He neither accepted nor discussed Jameson's views on the status of the species of Sinuolinea, and retained Ceratomyxa spinosa in the genus to which it was allocated by Davis.

Tripathi (1948a) followed Kudo in this regard, stating that “The spore of Ceratomyxa spinosa Davis, 1917 is not spherical, which is a necessary character of the members of the family Sphaerosporidae Davis, 1917…” Davis's definition of his Family Sphaerosporidae (1917, p. 219) reads as follows: “Spores pyramidal or approximately spherical; not distinctly longer than wide; with or without lateral processes.” As seen in front view, several members of the genus Sphaerospora itself are distinctly broader than long. For example, the genotype Sphaerospora divergens Thélohan may be more or less elongate (c.f. Auerbach, 1912; Pl. 5, fig. 4; reproduced by Kudo, 1920, Pl. IX, fig. 185); Davis (1917) himself stated that his Sphaerospora polymorpha is “sometimes slightly compressed parallel to longitudinal plane”; and Kudo (1920) described the form

of his Sphaerospora carassii as being “variable to some extent,” one of his figures (Pl. IX, fig. 204) illustrating a decidedly ovoid spore of this species. Similarly, Davis (1917) stated that the spore of his Sinuolinea arborescens is “rounded, slightly elongated along longitudinal axis” (his Pl. XXIII, fig. 110), giving the length of the spore as 15μ and the breadth as 12μ—this is hardly to be described as “not distinctly longer than wide.” Finally, the spores of both Sinuolinea bidens and Sinuolinea cella have ovoid central chambers, Jameson (1931) giving the dimensions of that of the former species as 8μ by 6.5μ to 11.5μ by 9μ and that of the latter species as 9μ by 8μ to 13μ by 10μ.

It is thus considered that the definition of the major group to which the genera Sphaerospora and Sinuolinea belong should be emended so as to include species having spores oval in shape or with oval central capsules.

In drawing a sharp distinction between those species of Sinuolinea having lateral extensions to the shell valves and those not having such extensions, Jameson (1931) contended that the species in the latter category only differ from the species of Sphaerospora in that their spores have a sinuous sutural line. He claimed that this is “hardly a sufficient ground for creating a new genus,” pointing out that the character of a sinuous suture line is regarded as being of only specific worth in the genus Myxidium. Jameson might also have instanced the genus Leptotheca in this connection for the spore of L. lobosa Davis (1917) has a sinuous suture line, whereas in most species of Leptotheca this line is more or less straight. As has already been indicated, Jameson proposed to retain only those species having a clearly walled-off spherical or oval central chamber, and a pair of prominent lateral appendages attached to this, in Sinuolinea. However, Jameson himself stated that in these species “The suture line is slightly curved but this is probably not of great importance.”

The author agrees with Jameson in considering that Sinuolinea as defined originally by Davis (1917) [and more recently by Kudo (1933) and Tripathi (1948a)] embraces species having spores of two such distinct types that they should be accorded separate generic status. He does not agree with Jameson in retaining those species having spores with lateral extensions in Sinuolinea and transferring the others to Sphaerospora. Although there are isolated cases in other genera (c.f. Myxidium, Leptotheca) where some species have spores with sinuous sutural lines, these cases are the exception rather than the rule in the genera concerned. Seven species of Sinuolinea having spores lacking lateral extensions but possessing a very sinuous sutural line have now been described. It is contended that these species should be retained in Sinuolinea—which in any case would hardly be an appropriate generic name for the species having lateral extensions, Jameson himself having emphasized that the spores of these species have the sutural line only slightly curved.

It is proposed, therefore, to establish a new genus, Davisia n. gen., to embrace those species of Sinuolinea Davis the spores of which have a clearly walled-off spherical or oval central chamber bearing a pair of lateral appendages. It is further proposed to follow Jameson in transferring Ceratomyxa spinosa to the company of these appendage-bearing myxosporidians previously considered to belong to Sinuolinea. Davisia n.gen. obviously has close affinities with Cera tomyxa Thélohan, 1892 (Kudo, 1933 emend.), the chief difference as regards spore characters being that the cavity of the central chamber is continuous with those of the appendages in the latter genus, but clearly walled-off from those of

the appendages in the former one. Elongate trophozoites having filopodia, so common in Ceratomyxa, have not been reported for any of the species here included in Davisia n.gen.

Davisia diplocrepis n.gen., n.sp., (Pl. 9, figs. 59—61).

Of 50 examples of Diplocrepis puniceus, ranging in length from 48 mm. to 90 mm. and collected at Point Jerningham, Lyall Bay and Island Bay, Wellington, in all months of the year (1950 and 1951), 44 had spores and trophozoites of D. diplocrepis n.sp. in the urinary bladder. There was no apparent pathological effect, apart from the fact that in extremely heavy infections the urinary bladder had a slightly milky appearance and its contents were very slightly viscous.

Trophozoites. Irregularly rounded in shape, the size ranging from 4.1μ by 4.0μ to 32.3μ by 30.6μ. The commonest forms are approximately 20μ in diameter. There is a broad and clearly demarcated peripheral zone of hyaline ectoplasm (Fig. 59), and the rather granular endoplasm contains many greenish refractive spherules averaging 0.7μ in diameter. Disporous and polysporous, up to eight spores of various stages of development being present in a single trophozoite. Extremely slow non–progressional movements of broad-ectoplasmic lobopodia take place, and the trophozoites soon become rounded and motionless in fresh preparations under microscopic examination.

Spores. Young spores (Fig. 60) have a more or less spherical central chamber. Their hollow lateral appendages are circular in cross-section. These appendages are shorter than those of mature spores, and at the point of attachment to the central chamber their bases are markedly inflated. The appendages, as seen in front view, may take up a variety of positions with regard to the central chamber. They may be straight and project at a slightly oblique downward angle, or be so recurved that they almost meet beneath the spore. The large shell valve nuclei, which are easily seen in fresh material, are located in the lateral appendages. Two rather smaller sporoplasm nuclei are present, lying close together one on either side of the rather curved sutural line. Small capsulogenous nuclei may be seen in contact with the round or ovoid polar capsules.

Mature spores (Fig. 61) have an oval central chamber, the length of which (from the measurement of 50 fresh examples) is 9.0μ to 12.0μ (av. 10.7μ) and the breadth 12.1μ to 14 0μ (av. 13.0μ). The curved lateral appendages assume such an angle as seen in front view as to give the entire spore a crescent-shaped appearance (broken, of course, by the bulging contours of the central chamber). Both the shell valve nuclei and the capsulogenous nuclei persist to a late stage in development, and the sporoplasm, which almost or entirely fills the cavity of the central chamber, contains two nuclei positioned as in the young spore. The polar capsules, which are usually ovoidal, are somewhat flattened on the side adjoining the sutural line This line is clearly visible in living material, but there is no raised sutural ridge. The lateral appendages of the shell valves are somewhat narrowed proximally, expanding to an inflated base at the point of junction with the central chamber. They taper distally to a characteristic nipple-like projection. The appendages of an individual spore are usually of the same size, although they may be slightly unequal. They range in length from 10.0μ to 14.2μ (av 12.7μ) The greatest breadth of the appendages ranges from 2.8μ to 3.8μ (av. 3.2μ). The polar capsules of an individual spore are equal or subequal, ranging from 3.4μ to 3.8μ in length (av, 3.7μ) and from 3.2μ

to 3.5μ in breadth (av. 3.4μ). The tightly coiled polar filaments do not contact the periphery of the capsules, and have from 5 to 9 coils. Extrusion of the polar filaments is readily brought about with 5% phenol, these filaments being 40μ to 56μ in length (av. 47μ). The filaments of immature spores may be extruded by the same treatment, but these remain crinkled throughout their length (Fig. 60).

Discussion. Davisia spinosa, D. opacita and D. brachiophora all occur in the urinary bladder of the same host, the halibut Paralichthys albiguttus (from Beaufort, North Carolina); while D. bidens and D. cella are both found in the urinary bladder of a toadfish, Porichthys notatus, in California. Davis (1917) stated that his (Ceratomyxa) = Davisia spinosa was rare, and recorded (Sinuolinea) = D. opacita and (Sinuolinea) = D. brachiophora on but one occasion each. In view of the considerable variation in size and shape exhibited by spores of D. diplocrepis n.sp. during their development, the possibility that Davis was actually dealing with, at most, two species instead of the three which he described, is not to be lost sight of. Taking Davis's descriptions as they stand, however, three distinct species are quite clearly characterized. Similarly, further research may well indicate that D. cella (Jameson) is a synonym of D. bidens (Jameson). Jameson found no trophozoites of the former species, spores of which are only differentiated from those of D. bidens by the somewhat greater size of the central chamber (9μ by 8μ to 13μ by 10μ, as compared with 8μ by 6.5μ to 11.5μ by 9μ), the greater length of the appendages (25μ to 35μ, as compared with 6μ to 10μ) and the less strongly curved sutural line. All these differences might well be correlated with increasing maturity, as in D. diplocrepis.

All six species of Davisia are disporous, D. spinosa being monosporous as well, D. diplocrepis being disporous and polysporous, and the remaining four species being disporous only, in so far as they are known from very limited material. D. opacita and D. bidens have short lateral appendages, strongly recurved towards the postcapsular side of the spore. They share this character with certain of the immature spores of D. diplocrepis. Should a continuous series through such forms to arc-shaped spores with relatively longer appendages be found in some future investigation, D. brachiophora may prove to be the mature phase of D. opacita (and possibly spores of both these species may prove to be immature phases of D. spinosa), while D. cella may prove to be the mature phase of D. bidens. However, the former eventuality is not so likely as the latter, for spores of D. opacita differ from those of all other known members of the genus in having a prominent sutural ridge.

Accepting the five earlier species of Davisia as they stand at present, the affinities of D. diplocrepis lie with D. spinosa, D. brachiophora and D. cella. The spores of these species have an oval central chamber, which measures about 13μ by 7μ in D. spinosa, from 9μ to 11μ by about 9μ in D. brachiophora and 13μ by 10μ to 9μ by 8μ in D. cella, as compared with 12.1μ to 14.0μ by 9.0μ to 12.0μ in D. diplocrepis. Although the central chambers are of roughly comparable size in all four species, the lateral appendages are longer in the earlier described species—–approx. 33.5μ in D. spinosa, 18μ to 22μ in D. brachiophora and 25μ to 35μ in D. cella as compared with 10.0μ to 14.2μ in D. diplocrepis. On the basis of these measurements D. spinosa and D. cella bear close resemblances to one another, as do D. brachiophora and D. diplocrepis. However, although Davis (1917) and Jameson (1929) failed to state the breadth of the lateral

appendages in their material, a comparison of Davis's Pl. XXI, fig. 72 with Jameson's Pl. IV, fig. 13 indicates that these appendages are relatively much narrower in D. spinosa than in D. cella. The lateral appendages of the spores of D. diplocrepis are appreciably shorter than those of D. brachiophora, and the central chamber is more rounded in the latter species. Otherwise these two species have much in common. The polar capsules are of similar size (3.5μ diam. in the latter species and 3.7μ by 3.4μ in the former), and (from Davis's Pl. XXIII, fig. 113) the lateral appendages of D. brachiophora may end in a narrowed projection as in D. diplocrepis.

The myxosporidian under discussion differs from the five previously described species now transferred to Davisia in that the lateral appendages of its spore are characterized by having an inflated base, which is particularly evident in the immature state. Although it is felt that further investigation of the earlier species may prove that other distinguishing features both among themselves and between themselves and the present species are not so significant as they seem at present while only brief and arbitrary descriptions based on limited material are available for comparison, it is considered best for purposes of reference to describe the New Zealand parasite as new. It is hence designated Davisia diplocrepis n.sp.

Leptotheca subelegans n.sp. (Pl. 8, figs. 51–58).

Trophozoites and mature spores were abundant in the bile of all three fishes found to be parasitized by this myxosporidian, 57 mm. and 72 mm. examples of Diplocrepis puniceus and a 45 mm. Callogobius atratus. Both the clingfishes were collected at Point Jerningham (3.9.50 and 26.2.51), while the goby was obtained at Island Bay, Wellington (14.7.51). The infected gall bladders were hypertrophied, the bile being distinctly sticky to the touch and of a light yellowish colour instead of the normal bright green.

Trophozoites. Both rounded and clubs-shaped forms occur, the rounded forms ranging from about 10μ to 28μ in diameter. Elongate forms (Fig. 51) have a terminal mass of cytoplasm containing many greenish refractive spherules, and a long, attenuated extension which is occasionally (Fig. 51) branched laterally and terminally. Typical elongate trophozoites appear rather like African knob-kerries in shape. They range in length from 25μ to 85μ. An ectoplasmic zone is rarely apparent, and none of the examples seen have been observed to display the slightest locomotor activity.

The species is disporous, sporulating trophozoites (Fig. 52) being of approximately ovoidal shape and averaging some 30μ by 25μ at their greatest diameters. The finely granular endoplasm contains numerous refractive spherules. These are particularly abundant peripherally, are of a greenish colour, and measure from 0.7μ to 1.0μ in diameter.

Spores. In the mature state, these are elliptical (Fig. 57) to egg-shaped (Fig. 58) in front view. Those from Diplocrepis range from 16.2μ to 20.3μ in breadth (av. 17.3μ) and from 8.3μ to 10.7μ in height (av. 9.0μ); while those from Callogobius range from 19.6μ to 26.2μ in breadth (av. 22.6μ) and from 8.6μ to 12.5μ in height (av. 11.6μ) (50 examples of each measured). From the polar aspect they are seen to be distinctly compressed and more or less arched, and to have lateral swellings at the extremities of the concave face of the arch. Because of the peculiar shape of the spore, almost all of those present in cover

slip preparations are so oriented as to lie with the front surface uppermost. The thickness of the spore is substantially less than the height, ranging from 5μ to 6μ in seven examples measured—–these seven spores were the only ones of hundreds seen which were lying still enough in the liquid medium to be measured from the polar aspect (although many other spores rolling over and over in fresh preparations, in consequence of the flowing of the medium beneath the cover slip, were momentarily glimpsed from this aspect).

The lateral limits of the shell valves are smoothly rounded, the valve walls being quite thin. One of the valves is often appreciably larger than the other (Fig. 54). The inequality may be so marked as to give the spore, as seen in front view, an egg–shaped appearance (Fig. 58). The sutural line is clearly visible by ordinary bright field illumination. This line traverses the spore transversely (Figs. 55, 56) or obliquely (Figs. 53, 54, 57), the valves uniting at a slight ridge which may be just sufficiently raised to break the even contour of the spore (Figs. 53, 55–58).

In developing spores (Fig. 52) the sporoplasm, although large, does not fill the entire space available to it, but in mature examples it almost or quite fills the spore cavity (Figs. 55–58). The sporoplasm is homogeneous and binucleate.

The polar capsules are spherical or subspherical, the side in contact with the sutural line often appearing somewhat flattened. Their diameter ranges from 2.1μ to 2.9μ in both hosts (av. 2.5μ). The polar filaments, which are easily extruded by means of 5% phenol, range from 64.5μ to 81.5μ in length.

Mature spores frequently remain for a time associated in pairs (Figs. 53, 54), joined to one another by means of a residuum of protoplasm derived from the trophozoite.

Discussion. The only member of this genus previously recorded as having spores displaying lateral swellings in polar view is Leptotheca elegans, described by Noble (1938) from a Californian kelpfish, Gibbonsia elegans elegans (Cooper). Noble later (1939, 1941) reported finding L. elegans in four additional Californian intertidal pool fishes belonging to the Families Cottidae, Clinidae and Gobiidae. Both rounded and club-shaped trophozoites, the former of equivalent size to those of the species under discussion, occur in L. elegans. Club-shaped trophozoites of the latter species differ from those of the New Zealand one in that they are not known to exceed 65μ in length. Noble (1938) observed sluggish movement in his trophozoites from Gibbonsia elegans elegans, but (1941) stated that those from Artedius lateralis (Girard) show no sign of movement (his measurements for the spore of L. elegans from the latter host, 22μ by 7μ, suggest that he may here have been dealing with a species of Ceratomyxa). The trophozoites of the New Zealand parasite differ from those of L. elegans from the type host in displaying no locomotor activity whatsoever. Both species have similarly shaped spores and polar capsules, and the size of L. elegans spores from Gibbonsia (17μ by 9μ) is equivalent to that of spores of the present species from Diplocrepis. As far as can be gathered from Noble's (1938) description, the only real point of difference between the two species lies in the character of the sutural line. This is indistinct in L. elegans spores, but in spores of the present species there is a distinct and slightly raised sutural ridge. Were it not for this latter point, I would have no hesitation in identifying the New Zealand Leptotheca as L. elegans. Disregarding the difference in zoogeographical dis-

tribution of these parasites, the intertidal pool habitat of the Californian host is very similar to that of the New Zealand ones. One of the hosts for L. elegans, Typhlogobius californiensis Steindachner, belongs to the same family (Gobiidae) as does the New Zealand Callogobius atratus. Furthermore, the habitat of T. californiensis is precisely the same as that of the New Zealand clingfish Diplocrepis puniceus, for according to Jordan and Evermann (1898) the former species has a ventral disc and lives attached to the underside of rocks in shallow water or surf.

L. obovalis Fantham, 1930, parasitizes various South African fishes of the pelagic and intertidal zones, the hosts of this myxosporidian including a blenny and three species of kelpfishes (Fantham, 1919, 1930). This species, like the New Zealand one and L. elegans, is disporous. Its club-shaped trophozoites range from 30μ to 60μ in length, and its spores as observed in front view are ovoidal or somewhat arched. Spores from Blennius cornutus (L.) measure 12μ to 18μ by 6μ to 9μ when fixed and stained, and thus enter the size range of those of L. elegans and of the species under discussion. Fantham failed to describe the appearance of L. obovalis in polar view. It would be of considerable interest to ascertain from the study of fresh material of this species whether or not its spores have lateral extensions as do the Californian and New Zealand parasites compared herein. If it should prove that the shape of L. obovalis spores entails their being seen most frequently in front view, it might eventuate that all three species are very closely related to one another if not conspecific.

It is considered inadvisable to follow Kudo (1933) in transferring L. obovalis to Ceratomyxa. The breadth of the spore of Fantham's species from the type host, Blennius cornutus (L.) is only equal to twice the height; and Kudo's emended generic diagnosis of Ceratomyxa requires the breadth of the spore to be more than twice the height. However, it is possible that some of the spores referred by Fantham to L. obovalis did not in fact belong to this species; for those from kelpfishes (1919), (Dentex) = Argyrozona argyrozona (Val.) (1919, 1930), and Lepidopus caudatus (Euphrasen) (1930) were all three times as broad as they were high. Fantham may thus have been dealing with a species of Ceratomyxa which he failed to recognise as such and which remains undescribed, as well as his undoubted Leptotheca. The same remarks also apply to the 22μ by 7μ spores which Noble (1941) found in the gall bladder of the North American sculpin Artedius lateralis (Girard) and identified as his L. elegans.

Noble failed to state whether or not L. elegans induces any changes in the gall bladder or bile of its hosts. Fantham (1930) noted that the bile of Blennius cornutus parasitized by L. obovalis remained light green and clear, although that of his earlier hosts (1919) was slightly turbid and yellowish green or yellow in colour. As already stated, the New Zealand species produces marked alterations in the gall bladder and bile of its hosts.

On the available data there is no alternative to considering the Leptotheca of D. puniceus and C. atratus as a distinct species, while recognizing its very close affinities with L. elegans and possibly with L. obovalis as well. To emphasize its relationships with the Californian parasite, it is hence described as Leptotheca subelegans n.sp.

Myxidium incurvatum Thélohan, 1892 (Pl. 9, figs. 62–71).

Trophozoites and spores of this species were abundant in the gall bladder of three examples of Diplocrepis puniceus (50–69 mm.) from Point Jerningham (19.9.50) and of a 46 mm. example of Notoclinus fenestratus collected at Lyall Bay, Wellington, on 15th October, 1950. A 36 mm. Oliverichtus melobesia collected at Long Beach, Russell, on 6th February, 1951, was also heavily infected with M. incurvatum; as were 14 of 28 examples of the same clingfish (28 mm.–39 mm.) obtained at Princess Bay and Island Bay, Wellington, from July to September, 1951. Scanty spores of this species were observed in the bile of three kelpfishes (Acanthoclinus quadridactylus) from Point Jeningham (14.8.50) and Island Bay (9.9.51).

Trophozoites. As seen in a drop of fresh bile these are irregularly rounded, and make slow, progressional movements by means of one or two broad and blunt lobopodia. The endoplasm is hyaline and finely granular, and contains numerous refractive spherules of a yellowish green colour. Larger and darker bodies, representing the initial stages of sporoblasts, can often be distinguished. The trophozoite is not surrounded by a clearly demarcated zone of ectoplasm, although the lobopodia present a less granular appearance than does the body proper. Trophozoites very soon become rounded and motionless in cover slip preparations, exhibiting from one to several large, clear vacuoles (Fig. 62). The majority of the examples seen ranged from 15μ to 20μ in diameter, although a few larger forms of up to 28.5μ were noticed. Monosporous (Fig. 63) and disporous trophozoites were observed. Young spores have round polar capsules and are straight in both front and side views, as Georgévitch (1916) indicated.

Spores. These were studied in both fresh and Schaudinn/iron baematoxylin preparations, all measurements being made from fresh material. In front view they are spindle-shaped (Figs. 66, 68), in side view twisted into the shape of a thick figure S (Figs. 69–71). The shell valves are not striated, and their extremities are rounded. It is unusually difficult to make out the sutural line, which is sometimes distinguishable in side view running along the proximal sides of the polar capsules (Figs. 69–71).

The measurements of 50 spores from N. fenestratus, and of the same number from D. puniceus and O. melobesia, are as follows:—

The polar capsules of the examples from N. fenestratus ranged from 3.7μ to 4.5μ (av. 4.0μ by 1.4μ to 2.2μ (av. 1.8μ), those of the spores from O. melobesia, 3.2μ to 4.4μ (av. 3.8μ) by 1.4μ to 2.2μ (av. 1.7μ), and those of the spores from D. puniceus, 34μ to 3.7μ (av. 3.5μ) by 1.2μ to 1.7μ (av. 1.5μ). Polar filaments, extruded by means of 5% phenol, ranged in length from 42.4μ to 73.4μ (av. 64.9μ). HC1 gave much less satisfactory extrusion, the filaments remaining partly coiled and measuring from 15μ to 30μ in length.

As seen in side view, the axes of the polar capsules may be parallel to one another (Fig. 71), but more frequently are not (Fig. 70). Due to the curvature

of the spore, the openings of these capsules are on opposite sides of the shell. The lightly granular sporoplasm, which does not completely fill the spore cavity, always contains two nuclei. These may be distinguished in fresh preparations under dark field illumination, as may the capsulogenous nuclei (Figs. 68–71) and one or both of the shell valve nuclei. In both fresh and Schaudinn/iron haematoxylin preparations, the undischarged polar filament is clearly visible, from five to eight closely wound coils being apparent.

Discussion. M. incurvatum has been reported from about twenty different species of fish, from both marine and freshwater habitats. It is of cosmopolitan occurrence, having previously been recorded from various European waters and from the Atlantic and Pacific coasts of North America.

Thélohan (1892, 1895) described this myxosporidian from six different North Sea hosts, including the blenny Blennius pholis L. The trophozoites from New Zealand fishes agree with his description in all essentials. Most authors have remarked on the presence of refractive granules in the endoplasm, although Georgévitch (1916) failed to observe these. The disporous condition is the typical one, although the monosporous one was recorded by Parisi (1912), Davis (1917) and Jameson (1929). Georgévitch and Dunkerly (1920, 1925) observed polysporous individuals.

Spores of M. incurvatum are somewhat variable in size Their dimensions, according to various authors, are as follows:—

Slight differences have also been noted in the shape and degree of curvature of the spore. These, together with the size differences, possibly indicate the occurrence of host-determined varieties or races. Thus, regarding the New Zealand material, the spores from N. fenestratus are somewhat longer and relatively rather narrower than those from O. melobesia, while those from D. puniceus are smaller still. Similarly, the measurements given by Noble (1941) suggest that his M. incurvatum spores from Dialarchus snyderi Greeley were longer and relatively narrower than those described by Jameson (1929) from another Californian marine fish, Sebastodes caurinus Richardson. However, Noble failed to state whether his measurements were made from fresh or fixed spores, although his measurements of other myxosporidians in earlier papers (1938, 1939) were based on living material. If he followed the same practice in regard to this species, the disparity could be attributed to shrinkage, for Jameson's figures were derived from the measurement of fixed and stained spores. There is a further difference between the parasites from these two Californian hosts which is not attributable to technique, Jameson's trophozoites being monosporous and Noble's being disporous.

Although there are thus grounds for supposing that host-determined varieties or races of M. incurvatum do exist, there is little evidence to indicate the occurrence of geographical variation. The overall size range of the spores from all

known hosts in Europe, North America and New Zealand is much the same in each of these localities, although the upper limit of that range appears, from the available information, to be rather higher in the latter country than in the other areas.

There is considerable variation in the figures published by earlier authors for the size of the pyriform polar capsules and the length of the polar filament in M. incurvatum. Noble's (1941) average of 3.8μ by 1.8μ for the dimensions of the capsule comes closest to the present ones (4.0μ by 1.8μ and 3.8μ by 1.7μ). The length of the extruded filament was given as 12μ by Thélohan (1892), who stated that extrusion was difficult to produce. The same author (1895) gave this length as 10μ to 15μ, while Parisi (1912) gave it as 28μ. Davis (1917) stated that when extrusion was brought about by means of HC1, the filament remained tightly coiled. This observation was confirmed in the New Zealand material, partially extruded and still coiled filaments measuring from 15μ to 30μ in length. Employment of Bond's (1938) technique utilizing 5% phenol gave much more satisfactory extrusion, the average length of the filament (64.9μ) being more than twice as great as the maximum derived from HC1-treated material. Earlier authors neglected to furnish details of the number of coils in the undischarged polar filament. The following figures are derived from their illustrations:—

These figures compare favourably with those from my material (5–8).

The spores of a number of other species of Myxidium are of broadly similar morphology to those of M. incurvatum. These species either differ from the latter one in their vegetative stages, or else have very much larger spores. The Myxidium of N. fenestratus and O. melobesia so closely resembles M. incurvatum as described by earlier investigators that it is considered, bearing in mind the wide geographical and host range of this species, that its description as new would be unwarranted.

Sphaeromyxa tripterygii n.sp. (Pl. 10, figs. 72–74).

This myxosporidian was first recorded from the gall bladder of a 72 mm. example of Tripterygion varium collected at St. Heliers, Auckland, on 12th January, 1951. Spores only were present. Spores and a trophozoite were found in the gall bladder of a 64 mm. example of the same host collected at Russell. North Auckland, on 7th February, 1951. Trophozoites alone were recorded from three examples of T. medium (67 mm.–74 mm.) inhabiting the same rock pool as this last blenny. Only one trophozoite was present in each of the infected gall bladders, while very few spores were found, six being present in one instance and five in the other.

Trophozoites. Far exceeding the gall bladder itself in length and breadth, each trophozoite is folded upon itself into a hollow ball closely applied to the inner wall of the bladder, through which it is clearly visible in situ. Unfolded and flattened trophozoites measure up to 7 mm. by 5 mm. in the fresh condition. The example illustrated (Fig. 72) was fixed in Schaudinn's fluid and stained with Delafield's haematoxylin. It measures 5.47 mm. by 4.03 mm., and is thin, opaque and bluish white in colour.

Spores. The few available for study are in Schaudinn-fixed smears stained with Heidenhain's iron haematoxylin. They are arcuate, resembling boomerangs with truncated ends (Figs. 73, 74). The measurements of the 11 examples discovered are as follows:—Length along the median line, 17.2μ to 21.1μ (av. 18.9μ); length along the inner side of the arch, 16.4μ to 20.0μ (av. 17.9μ); length along the more strongly curved outer side of the arch, 18.6μ to 22.3μ (av. 20.2μ); greatest breadth 3.4μ to 3.8μ (av. 3.5μ).

The pyriform polar capsules, located one at each end of the spore, measure from 46μ by 1.4μ to 5.1μ by 1.4μ (av. 48.μ by 1.4μ). A thick polar filament (not yet seen in the discharged state) is coiled lengthwise within the capsule. In the younger spores (Fig. 73) the rather granular sporoplasm contains two nuclei, and prominent polar capsule nucler are present at the inner ends of the capsules. These latter nuclei are not apparent in the older spores (Fig. 74), which also have only one sporoplasm nucleus, a synkaryon formed at autogamy.

Discussion Since 1933, when Kudo listed 10 species of Sphaeromyxa, three further species have been described. The inclusion of the present species brings the total up to 14. All of these inhabit the gall bladder of marine fishes. Their distinguishing characters are summarized in Table VII.

It is apparent from Table VII that the spores of Sphaeromyxa fall into two broad groups:—

The known vegetative stages of the species of this genus, with the exception of S. reinhardti, are of very large size. Within the incurvata group, the trophozoites of S. exneri, S. arcuata and S. curvula have not yet been described, and measurements are not available for those of S. hellandi. The thin and leaf-like trophozoite of the species under discussion is comparable in size and form with that of S. incurvata, and is appreciably larger than that of S. sabrazesi, the diameter of which usually approximates 2 mm. (although Schröder, quoted by Kudo, 1920, recorded examples up to 5 mm. in diameter).

The spores of S. incurvata and of S. exneri are very much larger than are those of the present species. They also differ from these in having actually and relatively larger polar capsules. The length of a single polar capsule expressed as a percentage of that of the spore itself approximates 40% in S. incurvata and S. exneri, but only 25% in S. tripterygii n. sp.

S. sabrazesi, S. hellandi, S. arcuata and S. curvula have spores which, allowing for differences in methods of examination and fixation, are comparable in length with those of the species under discussion. S. arcuata may, however, be eliminated from further comparison. From Fantham's (1930) figures, the sporeb of this species are actually and relatively much broader, and have longer and relatively more slender polar capsules than those of S. tripteryii n.sp. The breadth of a polar capsule expressed as a percentage of its length is approximately 20% in the former species, but 30% in the latter.

In S. sabrazesi, S. hellandi and S. curvula, the polar capsules, while being of much greater bulk than those of the New Zealand parasite, are of equivalent length-breadth ratio. The spores themselves are relatively broader in S. hellandi and S. curvula than in S. sabrazesi and S. tripterygii n.sp. The breadth of those of the first two species, expressed as a percentage of the length, is approximately 25%, while the equivalent figures for the last two species are 15% and 20%. Spores of S. tripterygii n.sp. are still further differentiated from those of these other three species by the fact that the length of a single polar capsule expressed as a percentage of that of the spore is only 25%, as compared with 35% for S. sabrazesi, 45% for S. hellandi and 40% for S. curvula.

Longitudinal striations have been seen on the shell valves of spores of three species of the incurvata group:—S. sabrazesi, S. arcuata and S. curvula. Such striations are often evident only in living material under the most favourable lighting conditions, and the fact that they have not been reported from other members of this group is not necessarily indicative of their absence.

As far as can be gathered from the figures published by earlier investigators, the outer side of the arch formed by spores of members of the incurvata group is usually more strongly curved than the inner one. Three species besides the present one—S. hellandi, S. exneri and S. arcuata—have the outer curvature of the arch sufficiently marked for the spore to appear humped or boomerang–like.

It is concluded that the Sphaeromyxa of New Zealand blennies, having affinities with other members of the incurvata group but differing from them in detail as described above, is sufficiently distinctive to warrant description as a new species. The type slide of Sphaeromyxa tripterygii n.sp. has been deposited in the collection of the Dominion Museum, Wellington (catalogue number Z104).

Zschokkella sp. (Pl. 10, fig. 75).

A single sporulating trophozoite was found in the urinary bladder of a 50 mm. example of Tripterygion varium collected at Moa Point, Wellington, on 15th March, 1951. No other life history stages were present, and no further infections have been recorded.

The following notes were made from a fresh cover slip preparation. The trophozoite (Fig. 75) which measures 18.4μ by 14.9μ, had rather granular cytoplasm containing 13 greenish refractive spherules averaging 0.7μ in diameter. An ectoplasmic zone could not be distinguished. Both of the young spores were reniform in shape, one of them being broader and more strongly curved than the other. Their greatest dimensions (length being measured along the median line) were 16.6μ by 5.4μ and 17.2μ by 7.4μ. The polar capsules, seen in anterior view, averaged 3.6μ in breadth. In each case the polar filament was wound in three coils Neither the sporoplasm nor the various nuclei of the spores could be made out.

Discussion. The genus Zschokkella, in so far as at present known, comprises 14 species. Twelve of these have been described from the gall bladder, bile ducts, kidney or urinary bladder of marine and fresh water fishes, one from the gall bladder of a toad and one from that of a tortoise. In seven of the species the shell valves are striated, in the other seven they are smooth (Tripathi, 1948a). Striations could not be made out on the shell valves of the immature spores studied in the present instance. Other authors (e.g, Davis, 1917; Tripathi, 1948a) have drawn attention to the presence of large refractive granules in the trophozoites

of species of Zschokkella, Tripathi stating that these disappear in fixed and stained preparations.

The affinities of the myxosporidian under discussion cannot be accurately assessed until more material is available. Two of the previously described members of the genus, Z. acheilognathi Kudo, 1916, and Z. russelli Tripathi, 1948, need not be considered further in this connection, for both of these have very large polysporous trophozoites. The spores of three of the other species from fish, Z. hildae Auerbach (data from Kudo, 1920), Z. rovignensis Nemeczek, 1922 and Z. salvelini Fantham et al., 1939, are very much larger than those of the present species. Some of the remaining species are of comparable size. Before useful comparisons can be drawn between these and the New Zealand parasite, mature spores of the latter will have to be obtained, for the shape of young myxosporidian spores often differs markedly from that of mature examples of the same species. It is possible that the present species will prove to be systematically close to Z. ovata (Dunkerly, 1920) of English rocklings. Its immature spores closely resemble the mature ones of Dunkerly's species in both size and shape.

Myxosoma tripterygii n.sp (Pl. 10, figs. 76 and 77).

This species is described from the subdermal connective tissue of the caudal peduncle of a 102 mm. example of Tripterygion varium, collected at Lyall Bay, Wellington, on 1st September, 1950. Cysts were not evident, only a few disporous trophozoites and mature spores being seen.

The following measurements and observations were all made from fresh material.

Young Spores (Fig. 76). Ovoidal or pyriform, conjoined in pairs by a small residuum of protoplasm. Two sporoplasm nucler and the capsulogenous nuclei are clearly visible. The spores themselves measure from 11.1μ to 12.9μ (av. 11.6μ) by 7.9μ to 9.0μ (av. 8.3μ) (14 measured). Their polar capsules are also ovoidal or pyriform, their dimensions ranging from 4.8μ to 5.8μ (av. 5.4μ) by 3.7μ to 4.0μ (av. 3.8μ).

Mature Spores (Fig. 77). These are circular in front view and apparently markedly flattened in side view, the position adopted by all of them in the liquid medium allowing of their being studied in front view only. Their diameter range from 11.7μ to 12.4μ (av. 12.1μ) (10 examples measured). The sutural edge is 0.6μ thick, and has from eight to ten triangular folds. The polar capsules are pyriform, and have a prominent projection anteriorly. They are relatively very large, measuring from 6.8μ to 7.8μ (av. 7.3μ) by 4.0μ to 4.6μ (av. 4.5μ). The polar filament, which is coiled five or six times, occupies the central portion of the capsule only, the sides remaining empty. The binucleate sporoplasm is rather small, and tapers anteriorly to a slender tongue which is pushed between the polar capsules and reaches a point about half way along them.

Systematic Position.

Mature spores of many of the species of Myxosoma are ovoid or pyriform. The spores of the following species, however, are circular or almost so in front view.

M. cerebralis Hofer, 1903. Data from Kudo, 1920.

M. encephalinum (Mulsow, 1911) Kudo, 1933. Data from Kudo, 1920.

M. dermatobium (Ishii, 1916) Kudo, 1933. Data from Kudo, 1920.