Some Coelozoic Myxosporidia from New Zealand Fishes I. —General, and Family Ceratomyxidae By Paul A Meglitsch [Received by the Editor May 27, 1959; communicated by J. T. Salmon and read by title before the Wellington Branch on September 9, 1959]. Abstract Surveys the Myxosporidia encountered in New Zealand fishes, and records the species observed belonging to the Family Ceratomyxidae. The classification of the Myxosporidia is discussed, and a new scheme of classification proposed. A system of angular measurements for use in measuring spores of the Ceratomyxidae is described, and the dependability of the measurements and amount of variability observed in various species is discussed. A total of 32 species, of which 29 are new, is described. Introduction and Acknowledgments This study has been conducted under the auspices of the U. S. Educational Foundation in New Zealand, and Victoria University of Wellington, the host institution. I am deeply appreciative of the opportunity they have afforded me. Especial thanks is due Professor L. R. Richardson, of the Zoology Department, whose assistance and co-operation have been a constant source of pleasure. I am also especially indebted to Dr. J. A. F. Garrick, who was kind enough to check the identification of the fishes examined, and also aided greatly in obtaining material. Dr. J. Moreland, of the Dominion Museum, also provided some of the material, as did Mr. Peter Castle, of the Zoology Department. Much of the work would have been impossible without the consistent co-operation of the crews of several commercial trawlers, who provided me with samples of their catches. In the northern hemisphere the Myxosporidia of fresh water and marine fishes have been studied extensively. A great deal less is known of the fauna of the southern hemisphere, and there have been but a few, sporadic studies of Myxosporidia from the southern Pacific region. As a result, the primary emphasis of the study was placed on a survey of the more available New Zealand fishes, with a view to determining what genera occur, and to providing a basic description of as many of the more common species as possible. The wide range and frequent occurrence of Myxosporidia in all kinds of fishes is impressive. Thus Davis (1917, p. 213) remarks, “The more the writer has studied the Myxosporidia, the more confident he has become that no group of fishes are free from these parasites… it is doubtful if there are very many fishes which are absolutely free of these parasites at all times.” In general, the results of infection experiments as well as the observations of natural populations have indicated a considerable degree of host specificity. Since there are about 20,000 species of fishes, it is evident that if the Myxosporidia are as ubiquitous as most students of the group have thought probable, and if there is even a moderate tendency toward host specificity, there must be many more unknown species than known ones. At the present time there are many more Families of fishes, according to Jordan (1923), than known species of Myxosporidia. We must expect eventually to encounter a large number of species, whose trophozoites and spores have a very simple construction and, therefore, but limited ways in which morphological traits may be expressed.

Regardless of our theories concerning the nature of the species group, the major clues for the recognition of species will undoubtedly remain morphological in the foreseeable future. Where the number of species is great and the morphological differentiation simple, the utmost care in the evaluation of morphological evidence is required. Unfortunately, we know but little of the factors which influence the form of Myxosporidia. Kudo (1921) made a careful study of the effects of fixation and dehydration upon spore dimensions. Otherwise, only scattered bits of information concerning the stability of the morphological trails are available. We know something of seasonal distribution, but very little of the effects of seasonal factors upon the appearance of either spores or trophic stages. Even less is known of the influence of biotic and geographical factors upon the form of parasites. Supposedly conspecific strains of Myxosporidia occurring in different hosts are known, in some instances, to exhibit slight differences in form, but we have very little knowledge of host-determined strains and almost no experimental evidence to support the interpretations suggested by morphological similarities. There is a dearth of information on such simple points as the extent of the variability that can be expected to occur in populations of Myxosporidia in a single host animal. Wherever time and material have favoured it, an attempt has been made to gather some information on variability and to make a preliminary analysis of the data. It is hoped that this secondary emphasis may help to define some of the problems, and to provide information that can help to establish a few principles that may aid the practising taxonomist in species recognition. In so far as possible, syntypes of the new species described below, will be deposited in the Dominion Museum at Wellington, New Zealand. The initial detailed drawing of the species is delineated in this paper: it may be considered as lectotypes. Materials and Methods Fishes have been obtained in a variety of ways. Some were caught on hook and line, some were taken in hand nets, and some were provided by other investigators. The great majority of them, however, were obtained from commercial trawlers. Trawler-caught specimens are not ideal for the study of parasitic protozoa, for it is unavoidable that trophic stages will often be missing or abnormal. On the other hand, it would be impossible to sample some host species in any other manner. It is surprising how often living, and apparently normal, trophozoites are found in fish which have died some hours before they are examined. To encourage trophic stages to remain as normal as possible, most of the trawler-caught fishes were chilled until they were landed. By far the most satisfactory method of studying spore characters in most coelozoic species is the examination of fresh material. Shrinkage is always associated with fixation, and in thin-walled spores it is difficult to accomplish dehydration without inducing deformities in the spore membrane. Where possible, therefore, the parasites were studied initially in the fresh state. For this purpose depression slides, ringed with ceder oil, were generally used. While the trophozoites soon deteriorated, spores remained quite normal for long periods of time. It was possible to notice changes in spores over a period of a month or more, and to compare similar material collected during that period directly. Fresh spores and trophozoites, in bile or in saline, were treated with several dyes to facilitate the study of nuclei and inclusions. For this purpose Bismarck brown, neutral red, and acidified methyl green were used. When attempts were made to stain mitochondria selectively with. Janus green B, it was found that concentrations too strong for selective mitochondrial staining were very useful for general staining Janus green apparently acts as a fungistatic and bacteriostatic agent in depression slides, and was particularly useful for this reason. Except for the acidified methyl.

green, staining solutions were made up in a 0.25% stock solution. Small quantities were added to the preparations. With practice it was soon possible to add a quantity which gave good results. Permanent preparations were made to serve as permanent records and for the study of nuclear detail. These were fixed wet in sublimate-acetic and Carnoy's fixatives. They were stained in Giemsa, iron haematoxylin, or, in some cases, Feulgen's nucleal reaction. A diluted Giemsa was allowed to act overnight, overstaining the specimens. They were then differentiated during dehydration in isopropyl alcohol and mounted in cedar oil. Iron haematoxylin preparations were stained overnight in cold reagents, or from 10 to 15 minutes at 50° C. Differentiation was carried out in dilute iron alum, for the most part. For a time, Dobell's modification of the iron haematoxylin method, as outlined by Gray (1958), was tried. The standard procedure was found to be superior for routine purposes. In some cases good reactions were not obtained with Feulgen's technique. This appeared to result, at least in part, from slow penetration of the reagents into the spores with thicker walls. In these cases the time of hydrolysis was varied, at one or two minute intervals, to obtain the maximum reaction, and the time for washing extended. In all cases, except for Dobell's methods, the reagents were prepared according to the formulae given by Kudo (1957). Fresh material was used, preferentially, for the measurement of spores. Unless there is a specific mention of fixation or staining of material, all quoted measurements are fresh material. By far the greatest part of the measurements were taken from camera lucida tracings. For some of the structures, where the width of a pencil line represents a large percentage of error, preference was given to measurement with an ocular micrometer. All angular measurements were made with a circular protractor, read to the nearest degree. All calculations were carried out with the aid of a slide rule, and are subject to the usual rule error. A great deal of time can be spent in making measurements. Since a definite limitation on the length of the study existed, it was desirable that no time so spent should be wasted. Experiment has shown that if care is used in selecting examples for measurement, the mean of both linear and angular measurements tends to remain quite stable after the first 20 to 25 have been recorded. The addition of a second sample of 25. increases the range somewhat, but rarely has a significant effect upon the value of the mean. Accordingly, the practice has been to standardize on samples of 25 Where infections were very light, or where the same species had been measured before, smaller samples were sometimes used. Unless there is some specific mention of spore numbers, a small sample is a group of 10–20 spores, usually the latter number. If the nature of the spore population is to be reflected by the measurement of samples, it is important that the sample be a random one. Where angular measurements are taken, it is particularly important that care is exercised in choosing spores with the proper orientation. The rule followed in the selection of spores for measurement was to measure the first 25 spores which (a) were mature, (b) were not obviously deformed, (c) lay in a definite sutural or capsular view, (d) were not tilted, which is to say that in spores with a straight axis, both valvular tips were in focus at the same time, and (e) were not obscured by extraneous matter. Random samples of 25 spores selected in this way have proved to be surprisingly dependable. Unfortunately, three different microscopes were used during the course of the study, resulting in a variety of magnifications in sketches of the material. As it has not been either practicable or possible to repeat all observations with the best instrument, it has been necessary to include drawings made with all three. Where the original magnification is given as × 900 (scale a) or × 1775 (scale d), an older Watson microscope was used. Where the original magnification is given as × 1000

or × 2045 (scales b and c), an older Leitz microscope was used. Magnifications of 1290 (scale c) and × 2750 (scale f) indicate the use of a new Leitz instrument, equipped with dark field and dark contrast phase. All references to use of dark field or dark contrast phase relate to observations made with the new microscope Since its resolution was superior to that of the older microscopes, and might have an influence on the description of spore ornamentation, at least a brief check with the new instrument was undertaken on all species where spore ornamentation might be expected and might be of taxonomic value. Classification of Myxosporidia One of the aims of classification is to provide a convenient system of cataloguing species, including a key for the location of any species in the system. Another purpose is to provide a system of groupings which reflect our ideas of relationship. While the two aims are not incompatible, a system admirably adapted for the one purpose may fall short of achieving a satisfactory level for the other. An “unnatural” system may be admirably neat, dependable, and easy to use, and a “natural” system may be expressive of relationships but cumbersome and difficult for the new student. It is the writer's bias to prefer a system between the two extremes, sacrificing a little in “naturalness” to avoid extremely difficult placements, and yielding a little in neatness or compactness to emphasize what appear to be significant relationships. The system given below is based very greatly upon the efforts of the earlier students, especially Gutley, Doflein, Thélohan, Kudo and Davis, but differing from theirs somewhat in certain groupings. Because of the addition of new genera during the last few years, it seems wise to slightly emend the description of the Myxosporidia as a group, to de-emphasize the bivalve nature of the spore membrane, and to emphasize the nature of the spore formation rather than the number of valves or sutures. The Myxosporidia have been divided into Suborders in a variety of ways Doflein (1898) used the disporous and polysporous habit as a basis for the division of the group, but this proved untenable as species were found, belonging to the same genera but differing in the number of spores produced. Davis (1917) pointed out the difficulties in the use of Disporea and Polysporea as suborders, and instead divided the Myxosporidia into the suborders Myxosporea and Cystosporea, the former including the genera in which the spores were not laterally compressed, and the latter the genera with laterally compressed spores. Kudo (1920), like Davis, preferred to make the principal suborders depend upon the shape of the spore, and divided the group into the suborders Eurysporea, Sphaerosporea, and Platysporea to include the forms with laterally expanded spores, spherical spores, and laterally compressed spores, respectively. This system has generally been followed by students of the group. Tripathi (1948) proposed the suborders Umpolaria, and Bipolaria, basing his first division of the group on whether the capsules were clustered at one end of the spore or were located at opposite poles of the spore. I am inclined to believe that the difficulty in obtaining satisfactory major subdivisions stems primarily from the tendency to use a single criterion for the basis of the division. While the use of a single criterion makes for an orderly system, it has the deficiency of bringing together genera which are not always closely related, and separating others which are more closely related. The evolutionary process, whatever it may be, is not necessarily dichotomous, aflecting first one and then another of an organism's attributes. In the system proposed here several trails are used to characterize the suborders, largely those which have been used by one or another of the previous classifications, in addition to the relationship of valvular number to capsular number and the presence or absence of a definite anterior surface.

The basic spore shape, in relation to the position and nature of the suture, is a fundamental aspect of spore morphology, and has been used previously for the description of groups above the family level. The spore shape, whether laterally expanded, spherical, or compressed, arises as a result of the growth and differentiation of the valvular cells of the sporoblast, and, to an even greater extent, the operation of molecular forces such as surface tension, viscosity, and the like. The action of these forces is an elemental aspect of the biology of the Myxosporidia, and deserves high place in any classification scheme. The series of divisions which precede differentiation of the sporoblast, however, are no less basic. In the majority of cases the sporoblastic divisions are so ordered that there is a constant and equal number of polar capsules and spore valves. Where abnormalities occur, as in the production of triad Ceratomyxa, the abnormal forms have three capsules as well as three valves. Where abnormal spores with single polar capsules are formed in genera which usually have two capsules, an abortive capsulogenous cell can almost invariably be seen in the immature stages. This has been noticed particularly in Sphaerospora undulans and two species of Sinuolinea occurring in New Zealand. Accordingly, some weight has been given to the relative number of spore valves and polar capsules at the suborder level. This trait works to separate unicapsular forms, such as Coccomyxa, and tetracapsulate forms, such as Chloromyxum, from the more common types. It breaks down, however, in separating Kudoa and Hexacapsula from Chloromyxum, and also in separating Thelohanellus from Myxobolus, and so, while valuable in conjunction with other criteria, proves inadequate as a single criterion. By far the largest group of Myxosporidia have a definite anterior surface, through which the polar capsules open. Others, like Myxidium, lack an anterior surface and have the capsules opening in opposite directions. Still others, like Zschokkella and Myxoproteus have the capsules at some distance from each other, rotated so that they open in different directions, or otherwise so placed that there is no true anterior surface. Tripathi used these criteria for the description of Unipolaria and Bipolaria, his two suborders (1947), but made his line of separation between those with two definite poles and those with a single anterior pole, rather than basing the division on the absence of a definite anterior pole, as has been tried in the system outlined below. This trait, like others, seems inadequate alone to divide the Myxosporidia into appropriate suborders, but is a valuable aid. By far the majority of the family groupings are like those used by Kudo, and now outlined in his Handbook (1957). Several new families are proposed, as a result of the nature of some of the newer genera. Myxoproteus, which differs from Zschokkella primarily in the amount of torsion affecting the capsular placements, and the spore height relative to its breadth, and Zschokkella have been placed in the same family, and separated from the Myxidiidae. The genera Sinuolinea and Davisia, differing primarily in the presence or absence of discrete lateral appendages in the spore, have been placed in a family, distinct from the Sphaerosporidae. While the classification as proposed has the aim of visualizing some of the probable relationships within the group, there can be little question that it fails in some cases. The position of such little-known genera as Agarella, and Auerbachia is certainly not entirely satisfactory, but so little information is available that one cannot but place them arbitrarily. Relationships of the suborders are not clear, as we lack the information necessary to establish any but the most sketchy lines of relationship at that level. Order Myxosporidia Bütschli Cnidosporidia with a spore membrane formed by the valvular cells of the sporoblasts, the segments of which typically meet in one or more sutures. From one to six polar capsules occur, each developing from one capsulogenous cell of the sporoblast. Each spore contains one or two sporoplasms, with a maximum of two

gamete nuclei, these apparently undergoing an autogamous union before or after the germination of the spore. Coelozoic or histozoic parasites of cold-blooded vertebrates, with one unconfirmed report of occurrence in an insect. Suborder I. Eurysporea Kudo Emend. Myxosporidia having spores with an equal number of capsules and valves. The axial length of each valve is greater than the diameter at the straight, or occasionally somewhat sinuous, suture. The sutural plane is at right angles to the greatest spore breadth. The capsules open through or near a definite anterior surface, and one is associated with each valve. Family Ceratomyxidae Doflein Emend Kudo With the characters of the suborder. Genus Ceratomyxa Thélohan Emend Kudo Ceratomyxidae with spores more than twice as broad as high, or arcuate forms with elongated valvular axes. Shell valves conical, often partially empty, but never septate Usually coelozoic in marine fishes. Genus Leptotheca Thélohan Ceratomyxidae with shorter, stubbier valves, the spore having a breadth less than twice the spore height. Usually coelozoic in marine fishes. Suborder II. Coccomyxea n. nom. Myxosporidia with spores having an anterior surface, through which the polar capsules opens. The spore has twice as many spore valves as polar capsules. The sporoplasm lacks an iodinophilous vacuole. Family Coccomyxidae Leger and Hesse Spore ellipsoidal, circular in cross section, with a single polar capsule. Coelozoic. Genus Coccomyxa Leger and Hesse With the characters of the family. Family Unicapsulidae Kudo Spore essentially spherical, with unequal valves meeting in a sinuous sutural line One polar capsule is present Histozoic. Genus Unicapsula Davis With the characters of the family. Suborder III. Sphaerosporea Kudo Emend. Myxosporidia with spores having a definite anterior surface, through which the capsules open. There are the same number of spore valves and polar capsules, and the sporoplasm lacks an iodinophilous vacuole. The spore may be spherical, or nearly so, may have one margin flattened, or may be pyramidal in shape. The spore valves are usually ornamented, and meet in a straight suture, occasionally slightly oblique in position. Family Sphaerosporidae Davis Emend. Spores spherical to subspherical, or somewhat ovoid in front view, and usually slightly flattened in profile, with the capsules lying approximately in the sutural plane. Coelozoic or histozoic. Genus Sphaerospora Thélohan With the characters of the family.

Family Wardiidae Kudo Spores with one side flattened, and suture set at right angles or oblique to the plane of the polar capsules. Often with filaments arising from the flattened spore surface. Genus Wardia Kudo Spore triangular in shape, with 2 convex sides. Suture straight, at right angles to the plane of the capsules. Genus Mitraspora Fujita Spores circular to ovoidal in front view and somewhat flattened in profile. Suture at right angles or oblique to the plane of the capsules. Shell striated, with or without posterior filaments. Suborder IV. Myxidea n. nom. Myxosporidia with spores lacking a definite anterior surface, the capsules being partially or wholly displaced toward the lateral extremities of the spore. Spore with an equal number of capsules and valves, the latter meeting in a suture which is usually sinuous. Sporoplasm without an iodinophilous vacuole. Family Sinuolinidae n. fam. Spore essentially spherical in shape, sometimes somewhat oval, with a sinuous sutural line, and with or without lateral appendages. Capsules only slightly rotated from the anterior surface of the spore. Genus Sinuolinea Davis Emend. Laird Spore spherical, without lateral appendages, and with a sinous sutural line. Coelozoc. Genus Davisia Laird Spore with a central capsule, spherical or somewhat oval in shape, with a somewhat sinuous suture. Lateral appendages set off from the central region of the spore by a definite septum. Family Myxoprotidae n. fam. Spore flattened at one side, with capsules opening near or from the flattened surface. Capsules rotated so that they do not have parallel axes. Suture obliquely set, and usually somewhat sinuous. Shell valves usually with offsets, in which the capsules lie. Genus Myxoproteus Doflein Spore pyramidal, with or without processes. Capsules lie near together, along flattened surface, in bulbous expansions of the valves. Suture oblique, and at least sometimes sinuous. Genus Zschokkella Auerbach Spore semicircular or crescentic in front view, fusiform in profile, and more or less circular in cross section. Suture usually sinuous, and set obliquely. Capsules large, set near the ends of the flattened surface of the spore. Coelozoic. Family Myxidiidae Thélohan Spore elongated, fusiform, arcuate or sigmoid in outline, with polar capsules at opposite ends of the spore and terminal capsular foramina. Suture straight or sinuous.

Genus Myxidium Bütschli Spores fusiform or sigmoid, with pointed or rounded ends, through which the capsules open. Polar filament long, fine, and coiled transversely. Coelozoic and histozoic. Genus Sphaeromyxa Thélohan Spores fusiform, with ends usually truncate. Polar filament short, thick, and coiled longitudinally. Coelozoic. Suborder V. Platysporea Kudo Emend. Myxosporidia with spores having an equal number of valves and capsules, except for Thelohanellus, which has one capsule and two valves. An iodinophilous vacuole may or may not be present. The polar capsules lie in the sutural plane, open on an anterior surface, and the spore is flattened in profile. Typically histozoic. Family Myxosomatidae Poche Spores with one or two polar capsules and one or two spore valves. No iodinophilous vacuole is present. Genus Myxosoma Thélohan Spores circular, ovoid or ellipsoid in front view, and lenticular in profile, with two polar capsules at the anterior end. Genus Auerbachia n. gen. Spore with a single polar capsule, containing a longitudinally coiled filament, and apparently covered with a membrane composed of a single piece. The posterior portion of the spore is extended as a hollow process, the cavity being continuous with the spore cavity. Coelozoic. Family Myxobolidae Thélohan Spores with one or two polar capsules, and two spore valves. An iodinophilus vacuole is present. Usually histozoic. Genus Myxobolus Bütschli Spores ovoid to ellipsoid, with two polar capsules at anterior end, and two spore valves. The spore is flattened in profile. Genus Thelohanellus Kudo Spores ovoid to ellipsoid in front view and flattened in profile. With a single, anterior polar capsule and two spore valves. Genus Henneguya Thélohan Spore nearly circular to ellipsoid in front view, and flattened in profile. With two polar capsules at the anterior end, and two spore valves, which are drawn out to form a bifurcated, median process. Genus Unicauda Davis Spore almost circular to ellipsoid in front view, and flattened in profile, with two anterior polar capsules and two spore valves. There is a single, median, posterior process, distinct from the spore valves. Genus Hofferellus Berg Spore ellipsoid, somewhat flattened posteriorly with two short, laterally placed, posterior processes. Shell valves are striated. Two shell valves and two polar capsules.

Genus Neahenneguya Tripathi Spore an elongated ellipsoid shape, flattened in profile. With two anterior polar capsules, set in tandem position, and two spore valves. The spore valves are drawn out into two anterior and two posterior processes. Genus Trigonosporous Hoshina Spore rather triangular in front view, with a flattened posterior and rounded anterior margin, and flattened in profile. There are two polar capsules and two spore valves. Two slender processes arise from each posterior corner of the spore, and are connected by filaments parallel to the base of the spore. Suborder VI. Chloromyxea n. nom. Myxosporidia with 4 polar capsules and 2 spore valves, or with 4 or 6 polar capsules and 4 or 6 spore valves. The capsules open together on an anterior surface. Sutures are straight or slightly sinuous. The spore is spherical, quadrate or stellate, with or without a posterior process. There is no iodinophilous vacuole. Coelozoic and histozoic. Family Chloromyxidae Thélohan Emend. Spores with twice as many capsules as valves. Coelozoic or histozoic. Genus Chloromyxum Mingazzini Spore with 4 polar capsules and 2 spore valves. No posterior process is present Coelozoic. Genus Agarella Dunkerly Spore with 4 polar capsules and 2 spore valves. With a posterior process. Histozoic. Family Kudoidae n. fam. Spores with an equal number of capsules and valves. Histozoic in muscle tissue. Genus Kudoa Meglitsch Spores with 4 polar capsules and 4 shell valves, with a rounded quadrate or tetra-radiate shape. Genus Hexacapsula Arai and Matsumoto Spores having six polar capsules and six shell valves, with a stellate shape. It will be noted that the genus Parvicapsula has not been placed. I have not been able to gain access to the description of this genus. Trilospora has also been omitted, tentatively. The triad form of the spore in this genus is not fixed, spores of a typical Ceratomyxa type also occurring. Triad spores are a common abnormality among the Ceratomyxa. Of course, as more than half of the spores of Trilospora californica exhibit the triad form, it is a moot question as to what is normal. It would seem preferable, however, to await the discovery of a species in which the tri-radiate symmetry has become fixed, as in Hexacapsula, before erecting a new genus for this type. Myxobilatus is also omitted, as being too poorly differentiated from Henneguya. The genus Auerbachia, to be described in a subsequent section of this report, has also been included. It seems probable that the Sphaerosporidae are a key family for the development of the histozoic habit. Both coelozoic and histozoic forms are included in the genus Sphaerospora. It may be found, eventually, that this genus should be divided into its two natural divisions, the polysporous, coelozoic type, typically with ornamented membranes and the sutural plane coinciding with the capsular plane, and the histozoic type, typically with unornamented membranes, and with the sutural

plane at right angles to the capsular plane. The histozoic Sphaerospora, with their tendency to form masses of pansporoblasts in the absence of a true cyst, appear to be the kind of forms which might serve as a transition between this suborder and the Myxsomatidae. Both Tripathi (1948) and Laird (1953) apparently feel very keenly the need for removing from all family or higher categories any remarks concerning what is known of their present distribution. Thus Laird (p. 98) complains of Kudo's (1933) classification because of such remarks as “Typically coelozoic parasites of marine fish”, with the remark that, “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 micro-habitat or the ecology of their hosts.” This reaction is, no doubt, based on differences in the interpretation of the nature of the characterizations we give to classificatory groups. According to one view, a classification may be an entirely arbitrary framework, with each category defined on an a priori basis. Another view is that a classification is merely a convenient summary of what is presently known of organisms, and that it is in actuality being modified constantly as new knowledge is acquired, whether or not the stated characterizations of the group have been restated to include the new information. If we adhere to the first point of view, the various categories are “defined” in the real sense of the word, as arbitrary abstractions. If we adhere to the second, each of the categories is “described” rather than defined, the descriptions being modified as new facts require. It is in the second sense that the present classification is constructed. The intent has been more to describe the common features of the organisms included in the categories, rather than to define them in the arbitrary sense. When the first histozoic Sinuolinea is found, the description of the genus has been altered, as has the description of the family. It is helpful, of course, to rephrase the written descriptions of the genus and family as well. It is not my intent, as I am sure that it was not Kudo's intent, to restrict the possibilities of assigning forms to appropriate groups, but rather to describe the groups as fully as possible, as a part of a continuously changing and expanding classificatory system. Family Ceratomyxidae Doflein The Family Ceratomyxidae includes two very closely related genera, Leptothecu and Ceratomyxa. It is a very widely distributed family in marine fishes, occuring as coelozoic parasites of the biliary and urinary systems. In the New Zealand material, it was the most common family, insofar as number of species is concerned. While some urinary parasites were found, the great majority of the species were biliary. Genus Ceratomyxa Thélohan Ceratomyxa, one of the oldest and largest of the genera of Myxosporidia, was established by Thélohan in 1892. At least 70 named species have been assigned to this genus, and a number of new ones have been found in the New Zealand material. In addition, 3 new species were partially described, but not named, by Awerinzew (1913, 1916). A number of records of Ceratomyxa, giving hosts, host organs, and localities, but without descriptions of the parasites, have been reported by Georgévitch (1916), Dunkerly (1921), Ray (1933), Setna (1942) and Meglitsch (1952). In 1895 Thélohan established a second genus, Leptotheca, to include Ceratomyxidae with oval or ellipsoid spores. As more species were described, however, a number were located which could reasonably be placed in either Ceratomyxa or Leptotheca, and the distinction between the two genera became increasingly difficult. Jameson (1929) expressed the opinion that the 2 genera should be combined.

Kudo (1933) attempted to solve the problem by redefining Leptotheca to include only those species with a breadth:height ratio of 2 or less, reassigning species as required by the new definition. This system has had general acceptance since. Even with the redefinition of Leptotheca, the distinction between the two genera can be difficult when dealing with a species that happens to fall near the borderline. The breadth:height ratio is a continuous variable, and examples can be found which approach 2 from both sides. As with any arbitrary distinction, it tends to result in assigning species which are very similar to different categories at times. Nevertheless, the difference between the extremely broad, crescentic spores of the typical Ceratomyxa contrast markedly with the stubby, oval spores of the typical Leptotheca. The two genera, as defined, tend to separate the ceratomyxid species into groups which have apparently differentiated in different directions. Some attention has been devoted to the breadth: height ratio of Leptotheca and Ceratomyxa spores obtained during this study. In spores of the same species there is a tendency for the spore breadth to be proportionately greater in larger spores, with the result that the breadth:height ratio is a function of spore breadth. In the smaller species, the spores lying near the lower end of the breadth range tend to have a bredth height ratio about 0.50 below those near the upper part of the breadth range. In general, the more slender the spore valves the greater is the difference between the breadth:height ratio of the smallest and largest spores. In order to describe this aspect of spore form, the breadth range has been broken into quartiles, and the mean breadth:height ratio of the spores in each quartile has been comptued. For all species in which a reasonable sample of spores in all quartiles of the breadth range have been measured, the statement of the breadth:height ratio for the various quartiles of the breadth range is given in conjunction with the species description. Since there were relatively few spores in some quartiles, the results obtained are not as dependable, statistically, as one might wish. However, it will be noted that the results obtained in samples of different species are surprisingly consistent, tending to support each other in establishing the tendency for the ratio to change with increasing breadth. Only in spores with a very great spore curvature do the quartiles fail to exhibit this trend, with but one or two exceptions. The fact that species with more stubby valves tend, in general, to exhibit less difference between the lowest and highest quartiles in breadth:height ratio would be understandable if the rate of increase of spore breadth, during the later stages of spore formation, were different in different species and highest, of course, in spores with the most slender valves. In this case, small increments in the growth time which preceded the firming of the spore membrane as it assumes its final contours, would have a greater effect on spore breadth, and on breadth:height ratios. The shorter, stubbier spores would be those which retained a more juvenile condition. Since the sporoblasts tend, in general, to be more oval in shape than the mature spores, there is a strong likelihood that this is actually the case. For this reason it is particularly important that mature spores be used for the calculation of spore dimensions. It is not impossible that the slope of the breadth:height regression on spore breadth is a characteristic feature of a species. The variability of breadth:height ratios within a species has considerable significance to the taxonomist, in comparing his samples with those described by others, and in considering the placement of species in Leptotheca or Ceratomyxa. Even in the stubby Leptotheca-like species of Ceratomyxa, a range of 0.50 in breadth:height ratio is not uncommon. This means that if the smallest spores, in the lowest quartile of the breadth range, have a breadth-height ratio of 1.75, the largest ones, in the upper quartile, will have a breadth: height ratio of 2.25. It further emphasizes the arbitrariness of the present definitions of Leptotheca and Ceratomyxa.

A careful study of the attributes of the spores in these two genera, however, reveals no better method of dividing them. A series of spore types can be constructed for all of the principal spore traits, showing a gradual transition from one extreme to another. The curvature of the spore, the amount of valvular taper, the narrowness of the valvular extremities, as well as other traits, vary in the same continuous fashion. In my opinion, therefore, any attempt to divide the species into two or more groups must be on the basis of an arbitrary character. The breadth:height ratio is no less satisfactory than any other, and is far easier to determine than most of the other traits that might be selected. The wisdom of retaining the two genera on the basis of a purely arbitrary distinction may be argued pro and con. The system is retained in this paper. Grouping of Ceratomyxa Species The large number of species of Ceratomyxa makes the comparison of new and known material laborious and difficult. To facilitate this task, the species have been assembled into groups on the basis of the following criteria: (a) the curvature of the posterior margin of the spore and the general spore shape, (b) the amount of valvular taper and acuteness of the valvular tips, and (c) the presence or absence of some form of lateral appendage. Of the various traits which might have been used, these appear to be the most universally available, and seem best to reflect the evolutionary trends in the genus. Because of the nature of the species differences in this genus, some species fall at or near the border-line between two groups, or are known to vary regularly so that they may fall into several groups. In using the groups it is necessary to understand that they are not infallible in these cases. Where it has seemed advisable, a species has been listed in two or more groups. The basis for the grouping system is outlined below. Spore Shape I. Spore elliptical or flattened. Posterior margin convex, or flattened; never more than very gently concave. II. Spore crescentic or arcuate in sutural or capsular view, with the posterior margin or one lateral margin concave. Where the spore is curved in capsular view, the species name is preceded by an asterisk in the lists below. In some instances, of course, the spore is curved in both sutural and capsular views. Valve Shape A. Valves taper little or not at all, with the anterior and posterior margins approximately parallel. Alternatively, spore oval, with a similar convexity of anterior and posterior margins. B. Anterior and posterior margins not parallel, and valves tapering appreciably, but terminating in bluntly rounded or truncated tips. C. Valves tapering to narrowly rounded or narrowly truncate tips. D. Valves tapering to sharply pointed tips, exclusive of appendages. Spore Appendages 1. Spore without a lateral appendage of any kind. 2. Spore with a lateral appendage of some kind. Several species have been omitted from the following list because of the lack of information. These are C. pallida, C. gobiodesi, C. hilsae, C. awerinzciwi, C. reinhardti, C. orientalis, C. shasta and the three C. sp. mentioned by Awerinzew Following Laird (1953), C. spinosa is omitted as belonging to the genus Davisia.

I-A-1 minuta n.sp. (9.6–14.2μ by 5.0–7.9μ) inconstans Jameson, 1929 (11.2–13.33μ by 5.45–7.66μ) faba n.sp. (10.7–14.1μ by 5.6–6.7μ) gracilis Jameson, 1929 (11–14μ by 4–5.6μ) castigata n.sp. (9.2–15.3μ by 5.1–6.9μ) recta n.sp. (14.7–16.7μ by 6.8–8.0μ) dubia Dunkerly, 1921 (17.5μ by 8μ) polymorpha n.sp. (23.0–44.5μ by 11.1–16.4μ) I-A-2 coris Georgévitch, 1916 (size ?). I-B-1 herouardi Georgévitch, 1916 (size ?). minuta n.sp. (9.6–14.2μ by 5.0–7.9μ). inconstans Jameson, 1929 (11.2–13.33μ by 5.45–7.66μ). faba n.sp. (10.7–14.1μ by 5.6–6.7μ). castigata n.sp. (9.2–15.3μ by 5.1–6.9μ). castigatoides n.sp. (9.8–15.9μ by 5.5–7.3μ, fixed). obovalis Fantham, 1930 (12–18μ by 6–9μ). gemmaphora n.sp. (14.2–23.0μ by 5.9–8.3μ, fixed). maenae Georgévitch, 1929 (20μ by 8μ). yoichiensis Fujita, 1923 (21μ by 9μ). starksi Jameson, 1929 (20–28μ by 6.25–8μ). agglomerata Davis, 1917 (24–28μ by 5μ). streptospora Davis, 1917 (34–39μ by 4μ). inaequalis Doflein, 1898 (31μ by 6μ). japonica Fujita, 1923 (37–49μ by 11–13μ). angusta n.sp. (34.2–58.8μ by 5.4–6.2μ). venusta Jameson, 1931 (63–78μ by 4–6μ). I-C-1 navicularia Davis, 1917 (14–22μ by 5–7.5μ). japonica Fujita, 1923 (37–49μ by 11–13μ). mesospora Davis, 1917 (50–65μ by 8μ). tylosuri Awerinzew, 1913 (124–140μ by 40–45μ). I-D-2 linospora Doflein, 1898 (50μ by 5μ). II-A-1 inconstans Jameson, 1929 (11.2–13.33μ by 5.45–7.66μ). obesa Jameson, 1929 (13.2–14.8μ by 4.5–5.9μ). intexua n.sp. (9.3–20.1μ by 3.4–5.4μ). hama n.sp. (18.5–29.8μ by 6.8–8.3μ). arcuata Thélohan, 1892 (20–30μ by 5–8μ). constricta n.sp. (23.6–29.3μ by 5.6–9.0μ). crassa Jameson, 1929 (27.5–32.0μ by 5.6–9.0μ). hopkinsi Jameson, 1929 (28.75–39.0μ by 5.9–7.5μ). polymorpha n.sp. (23.0–44.5μ by 11.1–16.4μ). spari Awerinzew, 1913 (50–60μ by 11–13μ). fukuiensis Fujita, 1923 (65–75μ by 11–13μ).

II-A-2 arcuata Thélohan, 1892 (20–30μ by 5–8μ). jamesoni Kudo, 1933 (95–117μ by 7.5–9.5μ)= C. tacnia Jameson, 1931, preoc. taenia Davis, 1917 (140–150μ by 6μ). II-B-1 hippocampi Cunha and Fonseca, 1918 (size ?). inconstans Jameson, 1929 (11.2–13.33μ by 5.45–7.66μ). faba n.sp. (10.7–14.1μ by 5.6–6.7μ). castigatoides n.sp. (9.8–17.8μ by 5.1–7.3μ, fixed). declivis n.sp. (13.5–15.2μ by 5.1–6.8μ). obovalis Fantham, 1930 (12–18μ by 6–9μ). intexua n.sp. (9.3–20.1μ by 3.4–5.4μ). gibba n.sp. (14.2–18.9μ by 5.6–8.0μ). lata Dunkerly (19μ by 7μ). gemmaphora n.sp. (14.2–23.0μ by 5.9–8.3μ). vepallida n.sp. (16.0–21.4μ by 7.8–9.6μ). scatophagi Chakravarty, 1943 (16–26μ by 4.2–7.2μ) *blennius Noble, 1938 (22 by 7.3μ). monospora Davis, 1917 (18–25μ by 5–6μ). undulata Davis, 1917 (22–24μ by 6μ). hama n.sp. (18.5–29.8μ by 6.8–8.3μ). * laxa n.sp. (20.8–30.3μ by 7.8–9.8μ). elegans Jameson, 1929 (23.6–29.3μ by 6–7.5μ). constricta n.sp. (23.6–29.3μ by 5.6–9.0μ). * renalis n.sp. (22.5–33.4μ by 7.3–8.5μ). uncinata n.sp. (27.8–36.0μ by 10.9–14.1μ). amorpha Davis, 1917 (27μ by 11μ). *toitae Fujita, 1923 (30–40μ by 13μ). dispar Kudo, 1933 (35–37μ by 15–16μ) = C. inaequalis Fujita, 1923, preoc. *limandae Fujita, 1923 (43–45μ by 11–13μ). nitida n.sp. (52.9–63.6μ by 10.8–13.7μ). *protopsettae Fujita, 1923 (50–65μ by 10–12μ). drepanopsettae Awerinzew, 1908 (50–80μ by ?). inversa n.sp. (51.1–73.7μ by 6.9–9.3μ). tenuis Fujita, 1923 (100–112μ by 10–15μ). robusta Fujita, 1923 (115–120μ by 18–20μ). II-B-2 *acadiensis Mavor, 1915 (40–50μ by 5–7μ). *appendiculata Thélohan, 1892 (50μ by 5–7μ). II-C-1 parva (Thélohan, 1895) (8–10μ by 3–4μ). flexa n.sp. (13.6–16.9μ by 5.6–7.0μ, fixed). insolita n.sp. (13–23μ by 10–13μ). gibba n.sp. (14.2–15.9μ by 5.6–8.0μ). subtilis n.sp. (15.7–26.0μ by 2.4–4.5μ). lunata Davis, 1917 (15–30μ by 7–9μ). *renalis n.sp. (22.5–33.4μ by 7.3–8.5μ). urophycis Fantham. Porter and Richardson 1910 (25.0–39.1μ by 1–7.5μ). moenci n.sp. (25.3–35.5μ by 4.5–7.1μ). *hokarari n.sp. (24.2–48.4μ by 11.0–13.2μ).

reticularis Thélohan, 1895 (45–50μ by 12–15μ). majimae Fujita, 1923 (46–50μ by 16–17μ). aggregata Davis, 1917 (50μ by 6–7μ). californica Jameson, 1929 (48–59μ by 7.5–9.0μ). ramosa Awerinzew, 1907 (50–80μ by 10–20μ). inversa n.sp. (51.1–73.7μ by 6.9–9.3μ). microstomi Fujita, 1923 (51–84μ by 11–13μ). osmeri Kudo, 1933 (? by 11–13μ) = C. furcata Fujita, 1923, preoc. furcata Fujita, 1923 (80μ by 15μ). elongata n.sp. (72.6–99.0μ by 7.7–11.0μ) flagellifera Davis, 1917 (118μ by 12μ). II-D-1 curvata Cunha and Fonseca, 1918 (size ?). recurvata Davis, 1917 (16μ by 6–9μ). globulifera Thélohan, 1895 (50μ by 10μ). *microcapsularis Fujita, 1923 (53–72μ by 11–13μ). sphaerulosa Thélohan, 1895 (90–100μ by 10–12μ). *attenuata Davis, 1917 (115μ by 9μ). sphairophora Davis, 1917 (115–140μ by 12μ). In addition to these named and partially described species, there are a number of records of Ceratomyxa which include information only as to the host organ and host species. While these records are of but limited value, they tend to direct attention to the hosts and thus aid in their eventual rediscovery and complete characterization. The following list, given in Table I, is no doubt incomplete, but is offered as a reasonably comprehensive survey of such records. Table I.—Undescribed Ceratomyxa. Host Species Host Organ Reference Bairdiella chrysura Gall bladder Meglitsch, 1952 Carcharinus bleekeri (2 spp.) Gall bladder Setna, 1942 C. limbatus (2 spp.) Gall bladder Setna, 1942 C. melanopterus Gall bladder Setna, 1942 C. menisorrah (2 spp.) Gall bladder Setna, 1942 C. pleurotaenia Gall bladder Setna, 1942 Cesbracion blochii Gall bladder Setna, 1942 C. zygaena (2 spp.) Gall bladder Setna, 1942 Hemigaleus balfouri (2 spp.) Gall bladder Setna, 1942 Hypoprion macloti Gall bladder Setna, 1942 Lagodon rhomboideus Gall bladder Meglitsch, 1952 Macrones gulio Gall bladder Ray, 1933 Molva molva Gall bladder Dunkerly, 1921 Muraena sp. Gall bladder Georgévitch, 1916 Pristis cuspidatus Gall bladder Setna, 1942 Rhinchobatis djeddensis Gall bladder Setna, 1942 Scoliodon sp. Gall bladder Setna, 1942 S. palasorrah Gall bladder Setna, 1942 S. sorrakowah Gall bladder Setna, 1942 S. wahlbeehmi (2 spp.) Gall bladder Setna, 1942 Sphaeroides maculatus Gall bladder Meglitsch, 1952 Angular Measurements It is difficult to describe the distinctive features of Ceratomyxa and Leptotheca spores with any degree of accuracy. To describe precisely the differences in spore curvature, valvular taper, or convexity of a spore margin is difficult at the best, and almost impossible in view of the different interpretations given to words by different investigators with different native tongues. Drawings can be used to state

form, as seen in a few spores, but cannot be used to describe the range of variability encountered in a typical population sample. Only a mathematical vocabulary can be sufficiently precise for descriptive purposes, but the difficulties involved in describing curvatures of a variable population in mathematical terms are enormous. In order to surmount these difficulties partially, three angles have been measured routinely. These have been termed the anterior angle, posterior angle, and tangential angle All angles were measured toward the posterior margin of the spore. The anterior angle is formed by lines originating at the anterior tip of the suture, ignoring any slight constriction in which it may be set, and passing through the valvular Text-Fig. 1.—Diagram of Ceratomyxa spore, showing angular measurements. Anterior angle, bac; posterior angle, bdc; tangential angle, edf. Curvature index, 360° minus (bac plus bdc); taper index, bdc minus edf; valvular index, bdc minus bac. The distances bg and cg are the valvular axes. tips (Text-fig. 1, bac). The posterior angle is formed by lines orginating at the posterior tip of the suture, again ignoring any slight constriction in which it may lie, and passing through the valvular tips (Text-fig. 1, bdc). The tangential angle is formed by lines tangent to the posterior margin of the spore, and meeting at the posterior tip of the suture (Text-fig. 1, edf). Points for the valvular tips are easily located in most species. The extremities of the valves are limited by a terminal curvature which is generally distinct from the curvatures of the anterior and posterior margins. The midpoint of this terminal curvature is the valvular tip. There is necessarily some slight inaccuracy in the placement of these points, as well as in the construction of tangent lines, but in practice the means and ranges of the angles measured have proved to be quite reproducible. These angles were chosen because they seem most closely related to aspects of spore form which are most difficult to express objectively in any other way. None of the angles are a simple statement of any one aspect of spore form, each varying with relation to two distinct architectural features. The anterior angle is determined by the position of the valvular tips in relation to the anterior tip of the suture. This position is the resultant of spore breadth and spore curvature. In oval spores, with the valvular tips opposite the midpoint of the sutural line, the anterior angle will vary directly with the breadth:height ratio. In spores of different sizes, but with identical shapes, the anterior angle, as all of the others, will be identical. In two spores with the same breadth:height ratio, the anterior angle will be smaller in the spore with the greatest curvature. In practice the anterior angle serves as a rough statement of the approximate convexity of the anterior margin of the spore, as seen in sutural view. The posterior angle is affected by the same factors that affect the anterior angle. All other factors remaining constant, the posterior angle will be more obtuse in

spores with a greater height in relation to breadth, or with a lesser curvature toward the posterior margin, or, in short, a spore with a more convex or less concave posterior margin. Taken in conjunction with the anterior angle, the posterior angle can be used to reflect the direction and amount of spore curvature, and the stubbiness of the valves. In a perfectly straight spore, with the valvular tips directly opposite the midpoint of the sutural plane, the sum of the anterior and posterior angles will be 360°. Since both angles are measured on the posterior side, a sum of less than 360° indicates a displacement of the valvular tips toward the posterior margin of the spore, while a sum of more than 360° indicates a displacement of the valvular tips toward the anterior margin of the spore. The curvature index has been defined as the sum of the anterior and posterior angles, subtracted from 360°. When positive, it indicates a tendency for the spore axis to be deflected toward the posterior margin, either through a flattening of the posterior margin or a definite curvature of the spore. The curvature index should not be thought of as an angle, but rather as an index of spore curvature. The valvular index is the difference between the anterior and posterior angles. In spores with short, stubby valves, the valvular index is high, while in slender spores it is low. When the spore is curved, the valvular index tends to be smaller, so that it reflects valvular shape but imperfectly. It is essentially sensitive to breadth:height ratio and spore curvature. The tangential angle varies with the position of the most posterior point of the valvular margin in relation to the posterior end of the suture. It will be more acute in spores with a stronger curvature. It has a necessary limit of 180°, and so is useless in describing the posterior margin of spores varying from flat to convex. As a result its descriptive value is much greater in the Ceratomyxa than the Leptotheca. The tangential angle will be more acute in a spore with bulkier valves than in one with slender valves, assuming the curvature to be the same. By using the tangential angle in conjunction with the posterior angle, it is possible to obtain a rough index of the amount of valvular taper. In spores with a very stumpy valvular shape, the tips are some distance anterior to the posterior margin of the spore. As the posterior margin projects well beyond the line connecting the posterior end of the suture with the valvular tip, the difference between the posterior and tangential angles is relatively large. In spores with slender valves, the posterior and tangential angles tend to be very similar. The taper index is the difference between the posterior and tangential angles. The taper index is far from precise, because of the limitation of the tangential angle to a maximum of 180°, and the fact that as spores become arcuate, the terminal curvature of the valves forms an arc which is not parallel to the sutural plane. Thus the taper index is reduced in spores with a greater curvature. Despite its limitations, however, the taper index has some descriptive value, particularly when it is interpreted with due regard to the amount of spore curvature. During the course of the study, a number of species were encountered in which the spore was curved horizontally instead of vertically. In order to distinguish between horizontal and vertical curvatures, spores which are curved vertically have been termed “curved” or “arcuate”, while spores which are curved horizontally have been termed “bent”. In a bent spore, seen in capsular view, one of the lateral margins is convex, while the other is flattened or concave. These two margins bear the same relationship to spore curvature and valvular shape as the anterior and posterior margins of curved spores. For some of the species with bent spores, therefore, the same three angles were measured, in the same manner. To distinguish them from the angles measured in arcuate spores, they have been termed the convex angle, concave angle, and horizontal tangential angle. While the concave margin may be somewhat less than concave in spores which are only moderately

bent, the simplicity of these terms was preferred to the more clumsy descriptive terms which might have been used. It is evident that a bending index may be obtained in exactly the same manner as a curvature index, and that horizontal valvular index and taper index may also be determined. In many species of Ceratomyxa there is some inequality in the length of the axis of the two spore valves. The lengths of each valvular axis were measured routinely during this study (Text-fig. 1, bg and cg), and were interpreted as the distance separating the valvular tips and the midpoint of the sutural plane. It is evident that even a slight tilting of the spore, or a moderate curvature of the valve, will cause the valvular axes to appear different because of foreshortening, even though they are actually the same. This was held to a minimum by measuring only spores with both valvular tips in focus at the same level. This was, of course, not possible in spores with the valves both curved and bent, in which case the axes were measured as accurately as the conditions permitted. Despite the attempt to prevent foreshortening from being a major problem, there can be no doubt that slight differences in the length of the axes is not significant. It is probable that only as differences approach about 10% of the valvular length they may be thought of as meaningful. The routine use of angular measurements has made it possible to form some estimate of their value. They have proved to have some excellent qualities. Angles, like linear dimensions, are continuous variables, and may be treated statistically with the same statistical tools. Despite the wide range of variability which is seen in most angular measurements, the extremes of the ranges, as well as the mean values, have tended to be very reproducible. Apparently the small element of subjectivity which enters into placing the valvular tips and locating the tangent lines is not a material difficulty. The ranges of the angular measurements tend to make an objective comment on the amount of variation in spore shape which would not be expressed as compactly in any other way. A discussion of the nature of the variability found and on the reproducibility of the angular measurements will be found in the section following the description of species. It appears that one may expect a reproducibility in the range of 15–20° in the species studied, and that on the whole the valvular index and taper index are more stable than the angles themselves. One of the most desirable features of the angular measurements is that they depend on spore form rather than directly on spore size. In two spores with identical shapes, but different sizes, the angular measurements are identical. They are, therefore, independent of modifications in magnification, and any errors in calibration. It may also mean that in comparison of material which is fixed with material that is fresh, the angular measurements may be considerably more comparable than linear dimensions. Fixed spores, studied in water suspension, tend to have angular measurements which are a little more variable, but with almost identical means, than similar spores studied in the fresh condition. In the thin-walled Ceratomyxa, however, the forces involved in dehydration tend to deform the spores so that angular measurements are not so dependable in fixed and stained, permanent preparations. On the whole, they have been found to be a valuable adjunct to the linear dimensions that have been used in the past. Description of Species Ceratomyxa minuta n.sp. (Text-fig. 2, Figs. 1–8) Habitat. This species occurs in the gall bladder of Thyrsites atun (Euphrasen) and Jordanidia solandri (Cuv. and Val.). Two infected Thyrsites and one infected Jordanidia have been seen, the latter very lightly infected In two of the three hosts, Leptotheca annulata was also present. In all of the hosts, the bile was a

Text-fig. 2.—For this and all subsequent plates, the following symbols are used for conditions of material and magnification:—F (fresh); S. A. (sublimate-acetic); C. (Carnoy's fixative); U (unstained), N. R. (neutral red); J. G. (Janus green B), M. G. (acidified methyl green); G. (giemsa); H. (haematoxylin); Fe. (Feulgen's nucleal reaction). The scales given refer to original magnifications of: a, × 900; b, × 1,000; c, × 1,290; d. × 1,775; e, × 2,045. f, × 2,750; The appropriate scale is designated in parentheses following the description of the figure. Figs. 1–8.—Ceratomyxa minuta n. sp. from Thyrsites atun. Fig. 1—Mature trophozoite with two well-developed spores. F., U. (e). Fig. 2—Young trophozoite with four nuclei. S. A., H. (f). Fig. 3—Mature unisporous trophozoites. S. A., H. (f). Figs. 4–5—Spores in sutural and capsular view. F., U. (e). Fig. 6— Spore from sutural aspect. S. A., H. (f). Fig. 7—Outline of spore with convex posterior margin. F. U. (e). Fig. 8—Outline of spore with triad form. F. U. (e). Figs. 9–20.—Ceratomyxa inconstans. Fig. 9—Mature trophozoite, from Usacaranx S. A., H. (e). Figs. 10–11—Spores from Scomber japonica in sutural and capsular views. S. A., U. (e). Fig. 12—Spore from Usacaranx lutescens, in sutural view. F. U. (d). Figs. 13–14—Spores from Trachurus. novae-zelandiae, in capsular and sutural views. F. U. (d). Figs. 15–16—Spores from Helicolenus percoides, in sutural and capsular views. F. U. (f). Fig. 17—Spore from Scomber S. A., H. (e). Fig. 18— Spore from Helicolenus. S. A., H. (f). Fig. 19—Spore from Trachurus S. A., H. (f). Fig. 20—Immature spore, from Usacaranx. S. A., H. (e).

reddish colour and quite viscous. The gall bladder and bile duct appeared to be inflamed. The infected fishes were taken by trawlers working out of Wellington in August and October. Trophic Stages. The trophozortes, immobile when studied, are rounded to irregular in shape, and vary from about 2μ to about 15μ in diameter. While there is very little visible distinction between endoplasm and ectoplasm, the occurrence of the inclusions in an internal position suggests that some differentiation exists. The ground-plasm is extremely transparent, and presents no visible structure. In living trophozoites a few tiny, irregular, granular inclusions may be seen, which are selectively stained by dilute neutral red. These inclusions are found in all but the smallest of the trophozoites. They are not seen in permanent slides. Trophozoites from about 5μ up usually have a few to many larger spherical inclusions. They are somewhat refractive, and are lightly stained by dilute neutral red. They average about 1 to 1.5μ in diameter. They are persistent in stained preparations. The trophozoites have a tendency to aggregate which increses with incresing maturity. The clumped tropozoites fuse together to such an extent that the individual boundaries are completely lost to view. Masses of sporulating trophozoites occur, containing large numbers of developing spores, and densely packed with the spherical inclusions. Isolated trophozoites containing spores are also common. In the majority of cases each isolated trophozoite contains two developing spores, but about 20–25% of the trophozoites are monosporous (Figs. 2, 3). A single residual nucleus remains after spore formation is complete. The many small individuals are apparently formed as a result of exogenous budding. Small trophozoites are multinucleate, usually with a minimum of two nuclei. Nuclear divisions begin early. After from three to five nucler are present (Fig. 2), the trophozortes enter a growth period during which there are relatively few nuclear divisions until spore formation is begun. Trophozoites which are to undergo budding, on the other hand, tend to exhibit a nuclear increase during the growth period. Spore Dimensions. Based on spores from two Thyrsites. Breadth, 9.6–14.2μ (11.8μ); height, 5.0–7.9μ (5.8μ), thickness, 5.4–5.9μ (5.5μ); valvular axes, 4.8–7.8μ (6.4μ) and 4.8–6.8μ (5.7μ), capsules 2.4–3.4μ (2.7μ) by 2.0–3.4μ (2.4μ) Anterior angle, 71–128° (109°); posterior angle, 154–225° (199°); tangential angle, 145–180° (178°). Curvature index, 52°., taper index, 21°; valvular index, 90°. Breadth; height ratio for quartiles of breadth range, 1.67, 1.99, 2.24 and 2.37. Spore Morphology. The stubby spores are in the form of flattened ellipsoids, with large, conspicuous polar capsules. In sutural view the anterior margin is convex, without interruption at the suture, while the posterior margin varies from convex to slightly concave, with a few spores showing a definite curvature (Figs. 4, 6–7) The two equal or sub-equal valves are similar in shape, meeting in a prominent, elevated suture, usually quite straight, but slightly oblique in a few spores. In capsular view the spore is essentially straight, although one margin may be somewhat more convex than the other (Fig. 5) The spore is slightly flattened, the thickness averaging a little less than the height. The large, broadly oval polar capsules are placed near the anterior margin. They converge somewhat on the suture, and in a few spores are strongly convergent. They are opaque, the filament being invisible in fresh spores. The filament measures up to 35μ in length when extruded. The capsulogenous nuclei are usually persistent, and can be seen in fresh spores. In capsular view the capsules appear to be slightly rotated toward the lateral margins. The sporoplasm fills the spore cavity in fresh spores. It is quite transparent, and extends upward between the capsules. A small number of rather large refractive spherules are usually present in fresh spores. These are sometimes in the sporoplasm, but are usually just outside of it, between the sporoplasm and the spore membrane. These spherules disappear in permanent preparations. There are two small, vesicular nuclei of variable position. In a number of the spores the terminal portion of the valves appear to be somewhat thickened, in the region where the valvular nuclei appear during development. This thickened region does not appear in stained preparations, so presumably is persistent after the complete disappearance of the nucleus itself. About 5% of the spore population has a triad form, with three valves and three polar capsules. The valves meet in three sutures, which pass between the capsules and meet at the apex of the spore (Fig. 8). Unlike the triad spores seen in some Ceratomyxa, these are very well formed, with quite normally formed valves. Tetrad spores were not seen. Spore Variability. Although this species shows about the same kind of variability as that seen in strains of C. inconstans, there tends to be a somewhat stronger central tendency with relatively fewer examples near the extremes of the range. The position of the suture with reference to the valvular tips is quite variable, as indicated by the range of the posterior angle (154–225°), and is accompanied by considerable fluctuation in spore curvature However, 60% of the spores from one Thyrsites atun had posterior angles between 203° and 218°,

while 60% from the other Thyrsites had posterior angles between 190° and 214°. In the great majority of cases, the valvular tips are anterior to the posterior end of the suture, and the posterior margin is flattened. Since but a small sample of spores were measured in one of the Thyrsites and the Jordanidia, no careful analysis of the similarities and differences of spore populations in the different hosts is possible. The small sample from Thyrsites, in all linear and angular measurements, had means which differed more from the full sample from Thyrsites than the small sample from Jordanidia. About the same range of spore shapes was observed in all three hosts. The maximal difference between samples was 1.0μ in spore breadth, with all other linear dimensions agreeing within a range of.0.5μ The means of the angular measurements were also very close, the greatest difference in anterior angles being 9°, in posterior angles, 2°, and in tangential angle, 2°. Discussion. The typical form of the spore is such that it falls into groups I-A-1 or I-B-1. It is evident that the only species in these groups which it bears a close resemblance to is C. inconstans and C. gracilis. That it is closely allied to C. inconstans is suggested by its occurrence in a host of the general mackerel type, as well as its small size. The more rounded spore shape places it nearer to the border line between the Leptotheca and Ceratomyxa than the strains of C. inconstans found in New Zealand material. It is considerably smaller than the strain of C. inconstans originally described by Jameson (1929), in which fixed spores varied between 11.2μ and 13.33μ in breadth. The polar capsules, not measured by Jameson, are considerably larger than those found in strains of C. inconstans obtained from New Zealand fishes. Jameson describes C. gracilis spores as oval, with no distinction between anterior and posterior margins. It is evident that the form found in Thyrsites and Jordanidia cannot be C. gracilis. Members of the Gempyllidae have not previously been found to harbour Ceratomyxidae. In the present study Leptotheca annulata has also been found in the same host species, and host organ. The conspicuous suture of L. annulata makes it easy to distinguish from C. minuta. Since it cannot be equated with any of the previously described species, the name C. minuta n. sp., has been proposed for it. Ceratomyxa inconstans Jameson (Text-Figs. 2–3, Figs. 9–29) Habitat. There are a number of strains of small-spored Ceratomyxa which appear to be best referred together to the species C. inconstans, despite differences in their spore from. They have been found in Scomber japonica, Usacaranx lutescens (Richardson), Trachurus novae-zelandiae Richardson, and Helicolenus percoides (Richardson). While not common, they are not particularly rare, insofar as is known. Unfortunately, not a long series of the mackerel group were available for examination, so repeated examples of each strain were not obtained. They were taken at various times during winter, spring and summer from hosts which had been taken by trawlers working out of Wellington. They did not occur in four Helicolenus taken at about 200 fathoms from Cape Palliser. The gall bladders of the Usacaranx and Trachurus were red and looked to be irritated, but there was no signs of abnormality in the remaining hosts. Trophic Stages. In most of the fishes, the trophozoites were deteriorating when the gall bladder was examined. No trophozoites were seen in Scomber. In Trachurus the disintegrating trophozoites are rounded to irregular, with disporous forms containing nearly mature spores measuring about 15μ to 20μ in diameter. The rounded trophozoites found in Helicolenus are from 15 to 30μ in diameter, and include both disporous and monosporous forms. In Usacaranx the trophozoites are similar to those found in Trachurus, and measure about 16 to 18μ in diameter when carrying nearly mature spores (Fig. 9). In all cases the protoplasm is transparent, and carries a number of tiny, refractive spherules measuring from about 0.5 to 1μ in diameter. These inclusions are less conspicuous than the ones seen in C. minuta, and are missing in fixed and stained forms. In Helicolenus there are many trophozoites which are very small, averaging about 4–5μ, and apparently arising from the larger trophozoites by budding. Spore Dimensions. From Scomber japonica. Based on spores fixed in sublimate-acetic and examined in water suspension. Breadth, 8.8–12.2μ (10.2μ); height, 3.4–5.4μ (4.4μ); thickness 3.4–5.0μ (4.5μ) valvular axes. 5.0–6.8μ (5.4μ) and 3.9–5.9μ (4.9μ), capsules,

1.0–2.0 (1.7μ). Anterior angle, 112–140° (125°); posterior angle, 195–276° (228°), tangential angle, 163–180° (178°). Curvature index, 7°; taper index, 50°; valvular index, 103°; Breadth height ratios for quartiles of breadth range, 2.00, 2.34, 2.50 and 2.58. From Usacaranx lutescens. Breadth, 10.7–15.7μ (12.5μ); height, 4.5–6.8μ (5.5μ), thickness, 4.6–5.6μ (5.1μ); valvular axes, 5.6–9.0μ (6.8μ) and 5.6–7.9μ (6.4μ); capsules, 1.7–2.3μ (2.1μ). Anterior angle, 84–132° (105°);, posterior angle, 160–213° (191°); tangential angle, 130–180° (160°). Curvature index, 64°; taper index, 31°; valvular index, 86°. Breadth: height ratios for quartiles of breadth range, 1.91, 2.13, 2.45 and 2.72. From Helicolenus percoides. Breadth, 8.6–15.4μ (12.0μ); height, 4.3–6.5μ (5.4μ); valvular axes, 5.4–8.6μ (6.4μ) and 4.8–7.6μ (5.9μ); capsules, 1.6–2.2μ (1.9μ). Anterior angle, 93–149° (109°); posterior angle, 168–228° (192°); tangential angle, 139–180° (160°). Curvature index, 59°; taper index, 32°; valvular index, 83°. Breadth: height ratios for quartiles of breadth range, 1.93, 2.07, 2.36 and 2.66. From Trachurus novae-zelandiae. Breadth, 9.8–14.1μ (11.8μ); height, 4.5–6.8μ (5.5μ); thickness, 4.5–5.6μ (4.8μ); valvular axes, 5.1–8.5μ (6.5μ) and 5.1–7.3μ (6.1μ); capsules, 1.4–2.0μ (1.7μ). Anterior angle, 71–119° (98°); posterior angle, 150–210° (181°); tangential angle, 117–180° (162°). Curvature index, 81°; taper index, 19°; valvular index, 83°. Breadth. height ratios for quartiles of breadth range, 1.94, 2.12, 2.24 and 2.19. Spore Morphology. The small spores are variable in curvature, and have short, subequal to unequal valves, terminating in rounded tips. In sutural view the anterior margin is always somewhat convex, although a few spores are seen in material from Scomber with the anterior margin nearly flat, and the axes of the valves angled somewhat forward (Fig. 21), and the material in Trachurus included some quite arcuate spores with correspondingly more convex anterior margins (Fig. 25). In all cases, the anterior margin curves smoothly over the suture. The posterior margin varies from convex to very concave. As indicated by the figures quoted in the dimensions section, the range of variability of the strains is somewhat different in the different hosts. In Scomber the posterior margin varies from extremely convex to at the most gently concave (Figs. 21 and 22), with the valvular tips always well anterior to the posterior end of the suture (posterior angle, 195–276°). In Trachurus the posterior margin varies from slightly concave to markedly concave (Figs. 23–25) with the valvular tips slightly anterior to well posterior to the end of the suture (posterior angle, 150–210°), and averaging out immediately opposite the end of the suture. In Usacaranx the spores vary from shorter, more stubby forms, to more clongate ones (Figs. 26–27), usually with the posterior margin somewhat concave. Variations in spore breadth result in placing the valvular tips well anterior to the sutural tip or somewhat posterior to it, with the average somewhat in front of the end of the suture (posterior angle, 160–213°). In Helicolenus the range of shapes is much like that seen in Usacaranx, with similar ranges in the posterior angle (168–228°) and almost identical means (Figs. 28, 29). In the majority of spores the valves are nearly equal in length. In all of the hosts, spores with valves of slightly different shape are not uncommon, one valve usually being more inflated, and the other terminating in a somewhat narrower tip, and often being somewhat more curved. The more slender valve is often evident in capsular view as well as sutural view (Figs. 14, 16). Dissimilarity of valvular shape is especially prevalent in the material from Trachurus. The valves meet in a distinct, but not prominent suture which is slightly elevated. In capsular view the spore is straight, or nearly so, and as the thickness tends to be somewhat less than the height in many spores, may appear somewhat flattened. The two polar capsules are nearly spherical, and while small, are essentially proportionate to the size of the spore. They are placed near, but not at the anterior margin, and are somewhat rotated when seen in capsular view. They are usually somewhat convergent on the suture. They are opaque, the filament being invisible in fresh spores. Capsulogenous nuclei usually persist as small, pycnotic nucler adhering to the capsular walls. The sporoplasm fills the spore cavity in fresh spores, or the greater portion of it. When it does not fill the spore cavity, it may be either symmetrical or asymmetrical in position. It is finely granular, transparent, and usually contains a few small, refractive spherules which disappear in permanent preparations. It is binucleate, the nuclei being vesicular and, in most instances, placed rather close together (Figs. 17–19). Spore Variability. There is a considerable range of variability seen in the spore populations within each host. This variability is not remarkably greater than that seen in other Ceratomyxa (see Tables V and VI), but is unusually extensive if all of these strains belong to the same species. Jameson (1929, p. 63) says of his C. inconstans, obtaind from Scomber japonica, “The front edge is always strongly arched, but the posterior one may be anything from very straight to strongly concave and the ends of the valves vary from very blunt to sharply pointed.

The two valves of a spore may vary in different directions. At first one feels that abnormal spores are being studied, but when different fish on different days are always found to offer the same picture one is forced to believe that we are dealing with a very variable spore.” It should be noted that in none of the hosts was such an extensive series of variations seen as those implied by Jameson, but if all of the hosts are included, then about the same range of shapes have been found in the New Zealand material. In actuality, the spore populations seen in the four host species appear to fall naturally into three discrete strains, those in Usacaranx and Helicolenus belonging to the same strain. This is shown most clearly if we take the central 60% of the various populations for the various dimensions, particularly the angular ones, as these are independent of fixation, to some extent, and the material from Scomber was fixed prior to its study. The range of the anterior, posterior and tangential angles for the median 60% of the spores from the different hosts are, respectively, Scomber, 112–128°; 202–229°; and all 180°; Trachurus, 90–110°; 175–192°; and 142–175°; Usacaranx, 98–109°; 180–205°; and 146–176°; and Helicolenus 104–120°; 102–207°; and 152–176°. Throughout, the Scomber material tends to be more straight, with little overlap with the other three, while the Trachurus material tends to be more curved, although with more overlap with the intermediate strain from Helicolenus and Usacaranx. It is not clear whether these may be thought of as host-induced strains or should be thought of as distinctive. Many more hosts must be examined before concluding that the differences seen in the forms the various hosts are stable ones. Neither is it conclusive that the forms seen here are actually identical with the strain seen by Jameson. Since Jameson measured fixed spores, reporting a breadth range of 11.2–13.33μ, it is evident that the material from California has larger spores than that from New Zealand, and the disparity in spore height is even greater. This may be explained on the basis of a certain amount of flattening of some spores, which often occurs in prepared mounts. On the other hand, Jameson speaks of nearly spherical spores, which were not seen in the material from any of the New Zealand hosts. However, immature spores of this, as of other Ceratomyxa, are very broadly oval (Fig. 20), and if Jameson included any immature spores in his measurements, this may account for some of the discrepancy. In view of the kind of variability seen in the various host species allied to the mackerels, and the existence of Jameson's variable species, it seems most probable that they are identical. The relationship of C. inconstans to C. parva (Thélohan, 1895) is not clear. C. parva is somewhat smaller than C. inconstans, as measured by Jameson, but is much closer to the size recorded in the New Zealand material. C. parva is shown as having a small, arcuate spore, not unlike those seen in Trachurus. There is no information concerning the amount or kind of variability occurring in C. parva, but as it occurs in Scomber scombrus, there can be little doubt of its close kinship with C. inconstans. If it can be shown that the Ceratomyxa of Scomber scombrus are also variable in form, it will be evident that all of these forms can be referred to C. parva. Ceratomyxa faba n. sp. (Text-fig. 3, Figs. 30–33). Habitat. This species was found in the gall bladder of a single Caulopsetta scapha (Forster). It occurred as a relatively heavy infection, mixed with C. laxa and C. torquata. The infected fish was taken in August by a trawler working out of Wellington. TrophiC. Stages. As the trophozoites had largely deteriorated, and there were several species present, no attempt was made to distinguish between the various types. Spore Dimensions. Breadth, 10.7–14.1μ (12.7μ); height, 5.6–6.7μ (6.2μ); thickness, 6.1 7.0μ (6.4μ); valvular axes, 5.6–8.4μ (6.9μ) and 4.5–7.3μ (5.6μ); capsular diameter,

Text-fig. 3.—Figs. 21–29—Ceratomyxa inconstans. Figs. 21–22—Outlines of spores from Scomber. showing range of variation. S. A., U. (e). Figs. 23–25—Outlines of spores from Trachurus, showing range of variation. E. U. (d). Figs. 26–27—Outlines of spores from Usacaranx, showing range of variation. F. U. (d). Figs. 28–29— Outlines of spores from Helicolenus, showing range of variation. F. U. (f). Figs. 30–33—Ceratomyxa faba n. sp. from Caulopsetta scapha. Figs. 30–31—Spores in sutural and capsular view. F. U. (c). Figs. 32–33—Outlines of spores, showing range of variation. F. U. (e). Figs. 34–38—Ceratomyxa castigata n. sp. Fig. 34—Unisporous trophozoite, with young sporoblast. S. A. H. (f). Fig. 35—Spore in sutural view. F. U. (f). Fig. 36—Stained spore. S. A., G. (f). Figs. 37–38—Outlines of spores, showing range of variation. F. U. (f). Figs. 39–40— Ceratomyxa castigatoides n. sp. Fig. 39—Two spore, attached together after leaving the trophozoite. S.A., U. (f). Fig. 40—Sutural view of spore. S.A., U. (f).

2. 0–3. 1μ (2.4μ). Anterior angle, 66–129° (99°); posterior angle, 150–239° (200°), tangential angle, 125–180° (162°). Curvature index, 61°, taper index, 38°; valvular index, 101°. Breadth: height ratios of quartiles of breadth range, 1.82, 1.94, 2.15, and 2.16. Spore Morphology. The small, stubby spores are gently curved toward the posterior margin, and have fairly large polar capsules. The anterior margin is strongly convex in sutural view, and is sometimes slightly indented at the suture. In the majority of spores it is unbroken at the suture, and its curvature grades into the terminal curvature of the valves without any evident inflection. The posterior margin is usually nearly straight, with a very gentle concavity which is sometimes more evident near the suture. In some spores an angle is formed at the suture (Fig. 30). In some spores the posterior margin very nearly parallels the anterior margin, with very little valvular taper, while in others the two are not parallel, and there is a moderate taper (Figs. 32, 33). The valves generally terminate in extremely broad extremities. They meet in a rather indistinct, slightly elevated suture, straight or nearly so. The spore valves may be equal or somewhat unequal in axial length, but are generally rather similar in shape. Despite the concavity of the posterior margin, the valvular tips are usually well anterior to the end of the suture in the majority of spores. In capsular view the valves are broadly rounded, with little taper. The spore is straight or nearly so, and slightly flattened, with the thickness somewhat less than the height. The relatively large polar capsules lie near the anterior margin. They converge somewhat on the suture, and are nearly spherical to somewhat oval. In capsular view the capsules are opposite, or slightly rotated (Fig. 31). The capsulogenous nuclei are usually persistent, although all other remnants of the capsulogenous cells disappear. The sporoplasm fills the spore cavity. It is quite transparent, and contains a few refractive spherules, in addition to two small, vesicular nuclei. It usually extends upward slightly between the capsules. Spore Variability. The spore form shows considerable variability insofar as the range of its angular measurements is concerned. The central tendency in this species is much like that of C. inconstans, which it resembles very closely in the general spore form and angular measurements. The median 60% of the spores lie between 93° and 100° for anterior angle, 190–213° for posterior angle, and 154–174° for the tangential angle. All of these values are very close to those seen in C. inconstans, and quite similar in range. Discussion. Although many species of Ceratomyxa have been reported from flat-fishes, none are in the general size range of this form. The average spores of this species fall into group I-A-1, with a fairly large number in I-B-1. Despite the stubbiness of the spores, the angular measurements would suggest that it might also be compared with the species in group II-A-1. Of the various species found in these groups, it is evident that it is most similar to C. minuta, C. inconstans, C. gracilis and C. obesa. It may be distinguished from C. minuta by its larger size and smaller polar capsules. It is more flattened posteriorly than C. gracilis, and the spore form is shorter and stubbier (Jameson, 1929). The spores of C. obesa are not as high in proportion to their breadth, and are somewhat larger, with a breadth of 13.2–14.8μ in fixed state. There are many similarities between C. faba and C. inconstans. The range and means of the linear and angular measurements are very similar, and the median 60% of the spores exhibit ranges similar to those seen in strains of C. inconstans, and which overlap them to a considerable extent. The difference in any of the linear dimensions between C. faba and C. inconstans is well within the range that may be expected in different host animals of the same species On the other hand, when samples differ to this extent in forms obviously belonging to the same species, they tend to differ proportionately, with the different samples showing a similar breadth: height ratio. C. faba spores have a breadth-height ratio much nearer the border line between the Ceratomyxa and Leptotheca than does C. inconstans, and in view of the marked difference in choice of host, they have been tentatively considered as a distinct species, and designated as C. faba n. sp. Ceratomyxa castigata n. sp. (Text-fig. 3, Figs. 34–8) Habitat This species occuried in the gall bladder of two of five Congiopodus leucopaecilis (Richardson). The infected fishes were taken by trawlers working out of Wellington in December and January. In one host the infection was mixed

with C. gibba, and in the other it was mixed with Sphaeromyxa. The bile was thin and yellowish. In neither organ was there any evidence of inflammation. Trophic Stages. There are many small, rounded to pyriform trophozoites about 5μ in diameter Pyriform trophozoites have a definite posterior process, and a few move sluggishly No clavate small trophozoites were seen. Larger trophozoites, some containing nearly fully developed spores, were also abundant. These were rounded, amoeboid, or clavate, with a few essentially pyriform. Apparently there is a tendency for the trophozoites with developing sporoblasts to forsake the pyriform shape, and assume a clavate or amoeboid form. Larger, rounded trophozoites containing two well-developed spores measure from 18–20μ in diameter Clavate forms may be up to 55μ or 60μ in length. While the majority of trophozoites are disporous, a few monosporous ones are found (Fig 34) The trophozoites are rather transparent, with homogenous to very finely granular ground plasm. There are very few inclusions, these taking the form of small, irregular granules, about 0.25μ in diameter, or larger, refractive bodies, about 1.0–1.5μ in diameter. The latter give an oily appearance, but are not selectively stained by either Janus green or neutral red. There is little evidence of ectoplasmic differentiation in the tiniest trophozoites. In some of the larger trophozoites the ectoplasmic layer is distinct, forming a continuous zone around the body In others it can be seen only in pseudopodia. Spore Dimensions. Based on measurements from two hosts. Breadth, 9.2–15.3μ (13.1μ), height, 5.1–6.9μ (5.9μ), valvular axes, 6.1–8.0μ (7.3μ) and 5.4–7.7μ (6.2μ); capsules, 1.6–3 3μ (2.2μ) by 1.6–2.5μ (2.0μ), and 1.6–2.5μ (2.1μ) by 1.5–2.2μ (1.7μ) Anterior angle, 100–129° (113°); posterior angle, 181–223° (206°); tangential angle, 155–180° (172°) Curvature index, 41°, taper index, 34°, valvular index, 93°. Breadth: height ratios for quartiles of breadth range, 1.93, 2.06, 2.15 and 2.41. Spore Morphology. The small, flattened to slightly curved spores have slightly unequal valves, and often have slightly unequal capsules. In sutural view the anterior margin is convex, with some tendency for the margin of the longer valve to be flattened or slightly concave (Fig. 38). The posterior margin varies from convex to concave (Figs. 35, 38), but in the majority of spores is flattened or slightly concave. One valve is often somewhat more curved than the other. The contour of both anterior and posterior margins is unbroken at the suture. There is at the most a very moderate valvular taper, and the valves terminate in very broadly rounded tips. The spores are not flattened in capsular view, the cross section being approximately circular. The suture is often slightly curved, or oblique, and is narrow. The polpar capsules are equal or slightly unequal. The disparity in size is never strongly marked, but the smaller capsule is often noticeably more slender. They are broadly oval, and opaque, and narrow sharply at the neck. They are slightly rotated when seen in capsular view, and converge slightly on the suture. The filament cannot be seen in living spores. It measures up to 30μ when extruded Capsulogenous nuclei are persistent. In many spores there are somewhat basophilic remnants of the capsulogenous cells. forming a small mass of material near the capsular foramina. The sporoplasm is large, completely filling the posterior part of the spore cavity in fresh spores. It is transparent, finely granular, and contains two vesicular nuclei with no definite locus. There are several small refractive spherules in the majority of the spores. These are sometimes persistent in Giemsa preparations (Fig. 36). Spore Variability. As with other small-spored Ceratomyxa there is a moderate range of spore shapes, associated with differences in the contour of the posterior margin of the spore. Capsules are equal or somewhat unequal, and there are small differences in the contour of the shell valves. The median 60% of the spores have angular measurements of 101–126°, 200–216° and 170–180° for anterior, posterior and tangential angles, respectively. The population samples from the two hosts were very nearly perfect matches. The differences between the two means for the linear dimensions were breadth, 0.2μ; height, 0.0μ: valvular axes, 0.4 and 0.1μ, and capsular diameter, 0.2μ. The differences in the means of the angular measurements were: anterior angle, 2°; posterior angle, 1°; and tangential angle, 4°. It is evident that insofar as the samples measured reflect the populations in the two hosts, they were extremely similar. It is interesting to note that one host was quite small, while the other was unusually large. Apparently the size of the host is not an important factor in spore morphology in this species. Discussion. No Ceratomyxa have been described previously from hosts belonging to the Congiopodidae, although the same host is also infected by C. gibba in New Zealand. The larger size and greater curvature of the spores of C. gibba make it easy to distinguish between them. The typical spores of this species fall into group I-A-1, with some showing sufficient valvular taper to approach group I-B-1. Of the species other than.

C. castigatoides, to which this species will be compared later, it is most like the small-spored species found in mackerels and their allies, C. minuta and C. inconstans, and like C. faba and C. gracilis. It is larger than C. minuta, with relatively smaller polar capsules. The size difference is not sufficient to guarantee that the two are different, however, for samples of spores from the same hosts belonging to the same species of Ceratomyxa occasionally differ as much as these two. Except for capsular size, the spores have much the same proportion in sutural view, but the spores of C. minuta are more flattened in capsular view, and do not commonly show curvatures or obliqueness of the suture. Moreover, the trophozoites of C. minuta contain a number of persistent, refractive inclusions which are not found in C. castigata. There are significant differences in the angular measurements and indices of C. castigata and the strain of C. inconstans seen in Scomber, in which the spores are more curved and have more taper. It is most like the forms seen in Usacaranx and Helicolenus. The differece in linear dimensions is not sufficient to ensure that the two forms are different, and there is considerable overlap in the general contour of the spores and the correlated angular measurements, although differences appear, particularly in the tangential angles of the median 60% of the population samples of the three strains. Despite the similarity, however, the tendency toward a curvature of the suture, and the larger polar capsules, often of slightly different size, in C. castigata makes it improbable that they are actually identical. The spores of C. faba are stubbier than those of C. castigata, with a lower breadth: height ratio. According to Jameson (1929), the spores of C. gracilis, although having a breadth similar to that of C. castigata, are considerably more slender, with a height range of 4–5.6μ. (Mean height, 5.9μ in C. castigata.) The spores of C. gracilis are also oval, without the flattened or slightly concave margin seen in C. castigata. As it does not appear to be identical to any of the previously described species, it has been designated as C. castigata n. sp. Ceratomyxa castigatoides n. sp. (Text-figs. 3–4; Figs. 39–44). Habitat. This species occurs in the gall bladder of Pseudolabrus coccineus (Bloch and Schn). Only a single specimen of this species was examined. It was taken by a trawler working out of Wellington in August. In addition to this species, a light infection with a larger, unidentified species of Ceratomyxa was also present. There was no evidence of irritation of the host organ. Trophic Stages. The trophozoites were not studied exhaustively in life, and as a second infection was not located, there is but little information concerning them. There were many small trophozoites of about 5μ or somewhat less, and fewer large ones. No distinction between ectoplasm and endoplasm was observed in the trophozoites. There were no refractive inclusions, and the protoplasm was very clear and transparent. Most of the larger trophozoites were disporous, at least, although a few isolated spores were seen with small amounts of protoplasm about them. As most of the protoplasm appears to be exhausted during spore formation. It was thought possible that these might be spores arising from monosporous trophozoites. In fixed material it was noticed that there were many spores attached together at right angles, without any constancy in the surfaces which were in contact. Occasionally small groups of three to six spores were found attached together. It is not known whether this indicates that some trophozoites are polysporous, or that they sometimes tend to aggregate. In view of the small size of the trophozoites seen, the latter seems to be the more probable Disporous trophozoites measured about 20μ in diameter. Spore Dimensions. Based on measurements taken of spores fixed in sublimate-acetic, stored in alcohol, and studied in water suspension. Breadth, 9.8–17.8μ (14.7μ); height, 5.1–7.3μ (5.9μ): thickness, 5.8–7.3μ (6.3μ); valvular axes 5.1–9.8μ (6.8μ) and 5.1–8.7μ (6.0μ); capsular diameter, 1.8–2.6μ (2.0μ). Anterior angle, 9–124 (108°); posterior angle, 190–248 (209°); tangential angle, 161–180° (168°). Curvature index, 73°; taper index, 41°; valvular index, 101°. Breadth: height ratios for quartiles of breadth range, 1.64, 2.07, 2.10 and 2.20. Spore Morphology. In general shape and appearance, the spores are rather simiiar to those of C. castigata, but are somewhat larger and have equal polar capsules. In sutural

view the anterior margin is convex, curving over the sutural region without interruption. One valve is not uncommonly less inflated than the other (Fig. 40). In such spores the lateral half of the more slender valve is flattened or slightly concave. In a few spores the posterior margin is flattened (Fig. 44), but in most spores it is gently to moderately concave. The range of spore shapes is essentially that between Figs. 44 and 42. The valvular taper is quite moderate, and the valves terminate in very broadly rounded tips. They are equal to slightly unequal in length, and meet in a straight suture, slightly elevated, and varying somewhat in prominence. In fresh spores it was noted that the suture often sat in a very shallow depression running about the spore, but this could not be seen in fixed spores. In some spores stained with haematoxylin, the suture stained as a double line (Fig. 42). In capsular view most of the spores are straight. In a few there is a slight tendency toward curvature, which is never marked. The disparity in valvular diameter often gives the spores a slightly boatshaped outline. The two equal polar capsules are nearly spherical in fresh and fixed material studied in water suspension, and broadly pyriform in permanent slides, with very narrow necks. They are set some distance from the anterior margin, and are not opposite when seen in capsular view. It appears, however, that they tend to converge on the suture, and to be convergent rather than divergent, slanting toward the main axis of the spore In fixed spores the filament cannot be seen In fresh spores it can be seen, but indistinctly, and the number of coils has not been determined. The capsulogenous nuclei are sometimes persistent, but are very small in mature spores. There are no other remnants of the capsulogenous cells. The sporoplasm fills the fresh spore. It is highly transparent, but contains a few small, refractive bodies which disappear in fixed spores, even prior to dehydration. The two vesicular nuclei usually lie close together. Spore Variability. Although the spore varies markedly in size, the largest spores being nearly twice as broad as the smallest spores, the general contour is relatively stable. Spores are never really convex posteriorly, and are never sharply curved. Sixty per cent. of the sample measured between 10.9μ and 14.1μ had anterior angles between 96° and 117°, posterior angles between 190° and 215°, and tangential angles between 157° and 180°. Discussion. The wrasses are often hosts to the Ceratomyxidae C. linospora, C. inaequalis, and C. arcuata have been reported from various members of the Labridae (see Kudo, 1920), while C. coris and Leptotheca fischeri have been found in members of the Coridae (Kudo, 1920; Jameson, 1931). This species has much smaller spores than those of C. inaequalis (31μ, according to Doflein, 1898); much straighter spores than C. arcuata, and lacks the appendages of C. linospora (Doflein, 1898) Ceratomyxa coris is something of a mystery as Georgévitch (1916) neglected to mention the dimension of its spores, and has drawn a spore which appears to be immature (see Kudo, 1920, Fig. 48). It is not clear whether the spores all retain this juvenile character, and as the drawn spore has the breadth: height ratio of a Leptotheca, its taxonomic position is questionable. In general appearance it is not unlike young spores of C. castigatoides, but until it has been described more fully it will not be possible to reach a final conclusion. Spore shape places this species in group I-B-I, or, when most curved, into II-B-I. Of the species in these groups, it is most similar to C. castigata. Since the spores measured had been fixed and were somewhat shrunken, the spores of C. castigatoides are somewhat larger than those of C. castigata Spore breadth is similar enough to prevent its use for separation of the two species, as are the other dimensions, taken alone. The breadth: height ratio is consistently lower in C. castigatoides, however, and the spores have a greater valvular index. The lack of curved sutures, the equality of the polar capsules, and the tendency of the spores to adhere in pairs, as well as the smaller quantity of residual protoplasm after spore formation in C. castigoides, all seem to indicate that despite the similarity in spore dimensions, these are not identical. C. castigatoides differs from the smaller C. inconstans in size, probably to a significant extent, at least for identification purposes. The capsules are larger, and the spores have a lower breadth height ratio. They are more curved than the spores of C. obovalis (Fantham, 1930), and somewhat less Leptotheca- like.

Of the species in group II-B-I, C. castigatoides differs from C. declivis in being less curved, and in having a more slender shape, as well as in the reaction of the spores to Giemsa Since the form cannot be equated with any of the previously described species, it has been designated C. castigatoides n. sp. Ceratomyxa declivis n. sp. (Plate III, Figs. 45–49) Habitat. This species occurs in the gall bladder of Cyttus novae-zelandiae (Arthur), and probably in the gall bladder of Cyttus australis Richardson as well. Two infected fishes of the former species were seen, both taken by trawlers working out of Wellington, in July and August. In August one very lightly infected Cyttus australis was also taken, apparently with the same species In none of the hosts was the gall bladder or bile duct inflamed. Trophic Stages. The trophozoites, immobile and rounded when examined, are transparent, apparently without the spherical inclusion often seen in Ceratomyxa. Only lobopodia were seen, but as the material was not fresh little can be sad of the range of trophozoite shapes or pseudopodia. The smallest trophozoites measure about 8μ in diameter, and are multinucleate Large trophozoites reach 25μ in diameter. Most of the trophozoites are disporous, although some unisporous ones occur. The small trophozoites are apparently formed as the result of gemmule formation (Fig. 49). A few trophozoites with dividing nuclei were seen. These have four chromosomes (Fig. 45). Spore Dimensions Based on spores from one Cyttus novae-zelandiae Breadth, 13.5–15.2μ (14.4μ); height, 5.1–6.8μ (5.9μ); thickness, 5.1–6.2μ, (5.6μ); valvular axes, 7.2–9.0μ (8.4μ) and 6.8–7.9μ (7.3μ); capsules, 1.7–2.8μ (2.4μ) by 1.7–2.2μ (2.0μ). Anterior angle, 87–105° (97°); posterior angle, 163–186° (175°); tangential angle, 132–155° (145°). Curvature index, 88°; taper index, 30°; valvular index, 78°. Breadth: height ratios for quartiles of breadth range, 2.65, 2.27, 2.51 and 2.57. Spore Morphology. The small spores are plump crescents, with rounded, almost truncated tips. In sutural view the anterior margin is convex, curving smoothly over the suture. The posterior margin of each valve is nearly straight, but meet at an angle at the suture in fresh spores. In permanent slides, however, the posterior margin tends to appear as a concave, curving margin (Figs. 46, 47). The two valves tend to be equal or slightly unequal, the differences between them never becoming prominent. In some spores one of the valves is slightly more slender than the other, and curves more sharply. In capsular view the spores are straight (Fig. 48). The straight suture is slightly elevated, and not very prominent. The spore membrane is very delicate, and rather easily deformed. The two polar capsules are broadly oval to spherical. They lie near the anterior margin. While they appear to be opposite when seen in capsular view, they are slightly divergent, the foramina being more lateral than the body of the capsule. The filament is but indistinctly visible in the fresh spores. It appears to form four or five coils. The capsulogenous nuclei are usually persistent. The sporoplasm very nearly fills the fresh spore, but is greatly shrunk by fixation. In fixed and stained spores its position is variable, but it tends to adhere along the posterior margin of the spore membrane. It is rather granular in fresh spores. A number of small, refractive spherules are seen in fresh spores. These are rarely more than 0.5μ in diameter, and usually a good deal smaller. These bodies disappear in fixed and stained spores. The two vesicular nuclei are small, and variable in position. A prominent feature of the spore is the presence of a number of bodies, oval to somewhat irregular in shape, which adhere to the spore membrane and stain intensely with the eosin colour of Giemsa's stain. These are apparently not identical with the small inclusions seen in the fresh spores, and cannot be seen in haematoxylin preparations In many spores they occur in such numbers that the sporoplasm, and even the capsules, are obscured. They occur in both immature and mature spores. It is interesting to note that similar granules occur in two other species of Ceratomyxa with curved spores, C. arcuata and C. flexa. Spore Variability. The spore varies somewhat in curvature and in the amount of valvular taper. All spores are at least gently curved, however, and the range of variability is less than that seen in many Ceratomyxa For anterior, posterior and tangential angles, respectively, 60% of the sample were in the ranges between 93–102°, 166–183° and 134–150°. Unfortunately, the infection was so light in the Cyttus australis that it was not possible to undertake a careful analysis of any differences between the spores in the two host species. The few spores that were measured fell into the same size range, and exhibited the same general properties. There can be little doubt that the two samples belong to the same species. Discussion. The only previously recorded ceratomyxid parasite from members of the Zeidae is Leptotheca vikrami, described by Tripathi (1948) from Zeus faber.

There are few similarities between L. vikrami and the present form. The spore shape places this species in group II-B-1. Of the species in this group, C. inconstans is significantly smaller, and more variable, and except for the strain from Trachurus, less curved. Some of the spores of C. castigatoides are as bent as this species, but only those near the end of the range of variability. The curvature index, 43° in C. castigatoides, is 88° in C. declivis. The unusual staining reaction in Giemsa also serves to distinguish C. declivis from all of the smaller curved species seen in New Zealand material. It would appear that C. declivis is probably more closely allied to C. flexa and C. arcuata than to any of the other species found in New Zealand material. Both of these species have spores which are significantly larger than those of C. declivis, and are more sharply arcuate, but both also have the prominent eosinophilic granules in Giemsa-stained material. As this form appears to be distinct from all previously described species, the name C. declivis n. sp. is proposed for it. Ceratomyxa intexua n. sp. (Text-figs. 4–5; Figs. 50–57). Habitat. This species occurs in the gall bladder of Jordanidia solandri (Cuv. and Val.). Two of five fishes were infected. They were taken by a trawler working out of Wellington in August. A third hake, taken in July, may also have been infected. In addition to having a few spores of Leptotheca annulata in its bile, there were several spores which closely resemble C. intexua. It may also occur in Plagiogenion rubiginosus (Hutton). In a specimen containing a heavy infection of C. flexa, a few spores were found which appeared to belong to C. intexua. There were not enough to make a positive identification possible, however. The two hake which were heavily infected with C. intexua had gall bladders which were red and inflamed, especially near the junction of the gall bladder and bile duct. Trophic Stages. The smallest trophozoites are very tiny, measuring about 1.5μ in permanent preparations. They tend to be rounded, and have clear, transparent, inclusionfree protoplasm. As the trophozoites increase in size they are less often rounded, becoming irregular, with lobopodia, or extremely irregular, with long, slender, often branching pseudopodia. Like the younger forms, the older trophozoites are quite transparent. It is not until they approach full size that a few granular inclusions, and an occasional spherical inclusion of an inconspicuous type, appear. Although the nuclei cannot be seen in the living organisms, the sporoblasts are visible early in development. Isolated trophozoites carrying two spores are not uncommon. Small clusters of trophozoites, containing spores in multiples of two are also seen. The smallest trophozoites show no tendency to aggregate, but as the sporoblasts appear the great majority of the trophozoites appear to aggregate in smaller or larger groups. Sometimes several hundred individuals are combined in a very large mass. Each unit of the aggregation is joined to the others by the intimate union of slender, intertwined pseudopodia. The limits of the individual organism are apparently lost in these aggregations, for no plasmalemma can be seen crossing the pseudopodial connections between them. As the spores mature the material of the trophozoites becomes more tenuous, and eventually masses of spores, held together temporarily by a delicate reticulum appear. Delicate remnants of the empty reticulum are also found. It is not uncommon for Ceratomyxa trophozoites to aggregate, but the intimacy and permanence of the union in this species is certainly unusual. Spore Dimensions Breadth, 9.3–20.1μ (15.4μ); height, 3.4–5.4μ (4.4μ); thickness, 3.5–4.3μ (3.9μ); valvular axes, 5.0–10.8μ (8.6μ) and 4.5–9.8μ (7.4μ); capsules, 1.2–2.2μ (1.8μ). Anterior angle, 93–135° (120°); posterior angle, 150–205° (180°); tangential angle, 130–180° (167°). Curvature index, 60°; taper index, 13°; valvular index, 60°. Breadth height ratios for quartiles of breadth range, 2.62, 3.03, 3.37 and 3.92. Spore Morphology. The elongated, slightly curved spores terminate in moderately narrow, rounded tips. In sutural view the anterior margin is convex, unbroken at the suture or slightly indented In some spores the lateral two-thirds of one valve is somewhat flattened, with a corresponding inflection of the anterior margin (Fig. 51). The posterior margin varies from flat to somewhat concave. It is usually smoothly curved over the suture, but in some spores it is slightly angled at the suture. The suture is inconspicuous, narrow, and only slightly elevated. In some spores it sits in a slight depression at the anterior face of the spore. The two valves are sometimes equal, and occasionally quite unequal (Fig. 53). In capsular view the spore is straight or slightly bent (Figs. 52–3). The spore is very slightly flattened in profile, but it is not a conspicuous feature.

Text-fig. 4.—Figs. 41–44.—Ceratomyxa castigatoides n. sp. from Pseudolabrus coccineus. Fig. 41—Capsular view of spore. S. A., U. (f). Fig. 42.—Stained spore, S. A., H. (f) Fig. 43.—Outline of spore with triad form S. A., U. (f). Fig. 44—Outline of spore with convex posterior margin, S.A., U. (f). Figs. 45–49.—Ceratomyxa declivis n. sp. from Cyttus novae-zelandiae. Fig. 45—Small trophozoite in early growth phase. S.A., G. (e). Fig. 46—Stained spore, showing cosinophilie bodies S. A. G. (f). Figs. 47–48—Spores from sutural and capsular views. F. U. (d). Fig. 49—Larger trophozoite, containing gemmule S. A., G. (f). Figs. 50–54—Ceratomyxa intexua n. sp. Fig. 50—Trophozoite, attached to other trophozoites by pseudopodia. F. U. (e). Fig. 51—Spore in sutural view F., U. (e). Figs. 52–53—Spores from capsular views, showing variation in bending F., U. (e). Fig. 54—Stained spore, near lower end of range of breadth S. A., H. (f).

The small polar capsules are nearly spherical. They lie near the anterior margin, through which they open, and are opposite, or nearly so, in capsular view. The filament cannot be seen in living spores. Capsulogenous nuclei are persistent, and in some spores remnants of the capsulogenous cells can alse be seen (Fig. 54). The rather small sporoplasm is transparent, and without conspicuous inclusions. It contains two vesicular nuclei. It is variable in position. A few small refractive spherules are found in most fresh spores, but these are missing in fixed and stained spores. In this species spore development is completed after the trophozoites have been placed on the microscope slide, hundreds appearing in depressions. The impression was gained that more spores were formed than the number of trophozoites present in the depressions initially would have formed Davis (1917) reports a similar experience with C. lunata. He states (p. 226–7), “The great variability in the spores was probably due to the fact that they were, for the most part, fomed under abnormal conditions after the trophozoites had been placed on a slide and covered with a cover glass sealed with parafin Spores formed under such conditions are often smaller than those produced under more favourable conditions.” As it happens, a sample of spores had been measured immediately after a depression slide was prepared, on a Friday afternoon. On the following Monday the slide was re-examined, and it was evident that a very large number of new spores had been produced. A second spore sample was measured and compared with the first. The second sample was somewhat more variable, and had a somewhat greater incidence of abnormal spores. The mean dimensions of the two samples were essentially the same. It may be that this species is more stable than C. lunata, less sensitive to the environment in depression slides, or that many of the spores in the original sample had been formed after the death of the host, and were already abnormal. The laboratory temperatures at Wellington are considerably cooler than those at Beaufort during the summer, which may also have been a factor In view of the very marked variability in size of C. intexua, it seems scarcely likely that it is a highly stable species, and it is more probable that the original population of spores were somewhat abnormal, or that the conditions favoured a more normal development. A small number of spores were found in a Plagiogenion rubiginosus which were very like the spores of C. intexua. The host was heavily infected with C. flexa, a species with strongly curved spores having relatively slender valves, and it is not certain whether they were unusually straight spores of C. flexa, or actually belonged to another species Spores like them did not occur in other Plagiogenion infected with C. flexa, however. Although the length of the valvular axis is about the same in C. flexa and C. intexua, the sutural diameter is greater in C. flexa, and the polar capsules are somewhat larger. The few spores which were relatively straight had smaller polar capsules, and a somewhat smaller sutural diameter (Figs. 56, 57). A positive identification, however, could not be effected with so little material. Spore Variability. There is a considerable variation in the size of the spores from a single host, and some spores have more tapered valves than others. There is a considerable difference in the relative equality of the valves in different spores. The most prominent variation, however, is in spore breadth In this species the spores are quite transparent, and the capsulogenous cells are very difficult to see in fresh spores. As the valvular nuclei disappear early, it is not impossible that some of the variability has resulted from including immature spores in the samples. The wide range of spore breadths is indicated by the fact that the median 60% of the spores vary in breadth between 13.2μ and 18.1μ. Sixty per cent of the spore population have anterior, posterior, and tangential angles of 116–128°, 173–194°, and 160–177°, respectively. It is evident that despite the variation in spore breadth, he spore shape is not unusually variable. The spread covered by the median 60% is essentially comparable to that seen in other Ceratamyxa species Spore height and thickness are not as variable as spore breadth. The breadth height ratios of the various quartiles of the breadth range reveal a rapid increase in breadth, proportionate to height. It seems possible that this may be one of the factors which tends to favour an unusually variable spore breadth However, in some of the larger, slender species of Ceratomyxa in which there is also a great difference in the breadth, height ratios of the various quartiles, no such marked variability in breadth was seen. Discussion. No Ceratomyxa have been reported previously from the Gempyllidae. In the New Zealand material, C. minuta has been seen in Jordanidia as well as Thyrsites atun The small size and conspicuous inclusions of the trophic stages of C. minuta make recognition of the two species easy. C. intexua falls into group II-A-1 or II-B-1, depending on the amount of valvular taper. Of the species in these groups, none are very similar to C. intexua The aggregative activities of the trophozoites are not reported for any of the similar species. The spores can be distinguished from those of C. inconstans by their greater

breadth and more slender shape. The same criteria distinguish them from C. faba, and while the spores of C. obesa are more similar in breadth, they are also much less slender (Jameson, 1929). While there is a considerable range of dimensions and shapes recorded for C. obovalis (Fantham, 1930), none of these have a breadth: height ratio comparable to that seen in C. intexua. Since the species cannot be equated with any of the previously described species, it has been designated C. intexua n. sp. Ceratomyxa recta n. sp. (Text-fig. 5, Figs. 58–62). Habitat. This species occurs in the urinary bladder of Genypterus blacodes (Bloch and Schn.). It is not a common form, as it was seen in but a single ling, taken in the Wellington area in September. Trophic Stages. The rounded, amoeboid, or clavate trophosoites reach a maximum size of about 35μ. Only lobopodia have been seen. No moving trophozoites have been observed, but the pseudopodia are apparently used, in part, at least, for attachment to the host organ. The protoplasm is dark, and rather coarsely granular. No spherical or other prominent inclusions are present. Despite the darkness of the protoplasm, no definite ectoplasmic layer can be seen, even in pseudopodia. The trophozoites are unisporous. A considerable quantity of protoplasm remains at the end of spore formation. There appears to be a single residual nucleus (Fig. 58). Spore Dimensions. Breadth, 14.7–16.7μ (15.6μ); height, 6.8–8.8μ (7.8μ); thickness, 6.8–8.3μ (7.8μ); valvular axes, 7.3–8.8μ (8.1μ) and 7.1–8.3μ (7.6μ); capsular diameter, 2.0–3.4μ (2.6μ). Anterior angle, 120–134° (120°); posterior angle, 219–240° (230°); tangential angle, 180°. Curvature index, 10°; taper index, 50°; valvular index, 90°. Breadth: height ratios for quartiles of breadth range, 2.07, 2.04, 2.05, 1.90. Spore Morphology. The spores have an elongated oval shape, with somewhat flattened anterior and posterior margins. In sutural view the valves are equal, and have similar shapes. They meet in a low, rather conspicuous, straight suture. Except for the position of the capsules and sporoplasm, the spores are similar in capsular and sutural views. In some spores the suture sits in a very shallow depression. The two nearly spherical capsules are placed a short distance from the anterior margin. They are slightly rotated, although they are opposite when seen in sutural view. The filament is visible indistinctly, apparently in the form of four or five coils. Capsulogenous cells persist in spores free from the trophozoites (Fig. 60, 61). At first they are rather dense, but as they age they become vacuolated, and eventually form a delicate membrane, in which the remnants of the capsulogenous nucleus lie, similar to the spores of Leptotheca pinguis. In this species, however, the spore eventually loses the membranes around the capsules. The sporoplasm is large, nearly filling the spore cavity. It is rather granular and opaque in fresh spores. A few refractive spherules sometimes occur in or near the sporoplasm. The two vesicular nucle are usually placed close together, and have rather conspicious peripheral chromatin. Immature spores have very conspicuous valvular nuclei, located at the tips of the valves, and embedded in the thickened substance of the membrane (Fig. 61). These nuclei persist, and may be seen in fresh spores, as well as in stained ones, giving the spore a permanently juvenile appearance. Spore Variability. The spore is remarkably constant in shape Lacking any tendency toward curvature, and with no valvular taper, it appears to be devoid of the traits which are most often variable in other species. The range of variation is about that seen in Figs. 59 and 62. Discussion. The only Ceratomyxa previously reported from a member of the Ophidiidae is C. arcuata, which occurs in the gall bladder of Ophidium vasalli (Kudo, 1920). It bears very little resemblance to this species. C. hokarari and C. inversa have also been found in Genypterus in the Wellington area. Both occur in the gall bladder, and one is strongly curved, while the other is very slender. The species belongs in group I-A-1. The only species in its general size range occurring in that group is C. dubia (Dunkerly, 1921), a biliary parasite of Cottus bubalis. The greater breadth: height ratio of C. dubia, as well as its choice of host organ, make it extremely unlikely that the two forms are identical. Immature spores bear a strong resemblance to the figures of Georgévitch depicting C. coris (see Kudo, 1920, Fig. 48), and to a lesser extent, C. herouardi

Text-fig. 5.—Figs. 55–57—Ceratomyxa intexua n. sp. Fig. 55—Stained spore, near upper extreme of breadth range. S. A., G. (f). Fig. 56—Outline of spore from Plagiogenion rubiginosus. F., U. (d). Fig. 57—Outline of spore with extruded capsules, from Plagiogenion. F., U. (d). Figs. 58–62.—Ceratomyxa recta n. sp., from Genypterus blacodes. Fig. 58—Trophozoite of clavate form. F., M. G. (e). Fig. 59—Slightly oblique view of spore. F., U. (e). Fig. 60—Capsular view of somewhat immature spore. F., M. G. (e). Fig. 61—Young spore S. A., H. (f). Fig. 62—Outline of Leptotheca-like spore. F., U. (e). Figs. 63–67.—Ceratomyxa flexa n. sp. from Plagiogenion rubiginosus. Fig. 63—Disporous trophozoites. S. A., H. (e). Fig. 64—Spore in sutural view. F., U. (d). Fig. 65—Stained spore. S. A., H. (d). Fig. 66—Stained spore. S. A., G. (f). Fig. 67—Outline of straight spore, showing range of shapes. F., U. (d). Fig. 68.—Ceratomyxa insolita. n. sp. from Dactylopagrus macropterus. F., U. (e). Fig. 68—Fresh spore in sutural view. U. (e).

(Kudo, 1920, Fig. 49). Georgévitch may have been dealing with species which, like this one, retains a more or less permanent juvenile appearance. Although no dimensions are given for these species, it is highly improbable that either are identical with C. recta, as both are biliary parasites. This species lies at the border line between the Leptotheca and the Ceratomyxa. It has been placed in the Ceratomyxa as a result of a statistical accident, as it were, the mean breadth being just over twice the mean height. It is interesting that in this species there was no evidence of an increasing breadth: height ratio in the various quartiles of the breadth range. The thickness of the valvular tips, containing the remnants of the valvular nuclei, may tend to prevent the lateral expansion of the spore valves during the later part of spore formation. For whatever reason, the relationship of height to breadth is somewhat different in this species than in most members of the group. Its similarity to the Leptotheca calls for examination of that genus for possible identification. Of the members of that genus, it is perhaps closest to L. pinguis, from which it differs in being considerably broader in comparison to its height, and L. lobosa (Davis, 1917) It differs from L. lobosa in having a straight suture, and anterior and posterior margins which are alike As this form is evidently undescribed, it has been given the name C. recta n sp. Ceratomyxa flexa n.sp. (Text-fig. 5, Figs. 63–7) Habitat. This species occurs in the gall bladder of Plagiogenion rubiginosus (Hutton). All three examples of this fish were infected; one possibly contained C. intexua as well. All of the infected fishes were taken by a trawler working out of Wellington in July. There was no evidence of irritation of the infected organs. Trophic Stages. The trophozoites, immobile when studied, are rounded to amoeboid, and measure up to about 25μ in diameter. Small granules, less than 0.5μ in diameter, and larger refactive spherules occur in the protoplasm. The latter do not persist in permanent preparations. Ectoplasm is never distinct In permanent preparations a number of tiny basophilic bodies, which appear to be more abundant than the granules seen in living organisms, can be seen (Fig 63). All of the sporulating trophozoites appear to be disporous. The two spores are variously oriented in the trophozoite, but not uncommonly are arranged with their posterior faces in contact, and the lateral portions of the valves curved around each other. In some young; spores just released from the trophozoites, there appears to be a membranous remnant of the sporoblast investing the spore. In young spores the valvular nuclei deteriorate before the capsulogenous cells have disappeared. Spore Dimensions. Based on spores fixed in sublimate-acetic, stored in alcohol, and measured in water suspension Breadth, 13.6–16.9μ (15.9μ); height, 5.6–7.9μ (5.6μ); valvular axes, 8.2–11.9μ (10.1μ) and 6.5–11.3μ (8.6μ); capsules, 2.3–3.4μ (2.6μ) Anterior angle, 5.3–1.13° (82°); posterior angle, 95–192° (147°); tangential angle, 95–170° (129°). Curvature index, 131°; taper index, 18°; valvular index, 65°. Breadth: height ratio for quartiles of breadth range, 2.17, 2.49, 2.76 and 3.24. Spore Morphology. The arcuate spores usually have unequal valves, with moderately narrow tips. In sutural view the anterior margin is strongly convex, the curvature passing smoothly over the suture. Occasionally a spore is seen with the lateral third of one valve somewhat flattened anteriorly. The posterior margin is moderately to strongly concave, usually curving smoothly along its whole course, but occasionally slightly flattened in the central region (Fig. 65). The suture is rather inconspicuous in most spores. It is scarcely elevated, and in the majority of spores takes a straight course over the spore surface In a few spores, however, it passes obliquely over the anterior face of the spore, and is slightly curved in sutural view. The polar capsules are equal, and lie at the anterior margin. They are broadly oval to spherical, with short, slender necks. The capsules sometimes converge a little on the suture, but this is never a conspicuous feature of the spore. In capsular view the sutures are usually opposite, opening at the midline. In spores with an oblique suture, however, the capsules are somewhat rotated in position, with the foramina located about half-way between the spore axis and the lateral margins. The filament is only imperfectly visible in fresh spores. It appears to coil four or five times. The longest filament to be extruded measured 25μ. Although the capsulogenous nuclei are persistent in mature spores, they become very small and inconspicuous (Fig. 64. 66).

The large sporoplasm does not fill the spore cavity. It is rather granular in fresh spores. From one to about six refractive spherules are present in spores. These are often outside of the sporoplasm, which is slightly withdrawn from them (Fig. 64). They are not persistent in stained preparations. The two sporoplasmic nuclei are small, with a small endosome. They usually lie close together. The spores of C. flexa, like those of C. declivis, contain many bodies of irregular shape which stain deeply in Giemsa preparations and cannot be seen in haematoxylin preparations (Fig. 66). In C. flexa they tend to be less numerous than in C. declivis, to cluster together and from larger masses more often, and to be less exclusively associated with the spore membrane, occurring in the sporoplasm as well. They are not seen in very young spores, just taking form, and are somewhat less abundant in the mature spores. Spore Variability. The spores vary primarily in culvature, the range being essentially that between Figs. 67 and 65. The valves vary somewhat in taper, but otherwise, except for curvature, their shape is relatively constant. The suture is somewhat variable, and the capsular position is also somewhat variable. Although occasional spores with much more curved or much straighter shapes are seen, the majority of the spores fall in a fairly narrow range. Sixty per cent. of the spores have anterior, posterior and tangential angles of 74–92°, 131–167°, and 116–142°, respectively, a range somewhat larger than most Ceratomyxa species, but not outstandingly so. Discussion. No Ceratomyxa have previously been reported from the Theraponidae. Perhaps the closest record is the occurrence of C. arcuata in Pagellus centrodontus, one of the Sparidae (see Kudo, 1920). C. arcuata is larger than C. flexa, with considerably longer valvular axes, although in other respects the two are rather similar. The configuration of the average spore fits group II-C-1. It is significantly larger than C. parva, and smaller than C. lunata, the only previously described species adjacent to it in this group. Of the species found in New Zealand material, the closest in size is C. gibba, which differs from it in the breadth height ratio, the inequality of the valves, as well as in lacking the eosinophilic material in the spores stained with Giemsa. In many ways it is more similar to C. declivis, from which it differs in its more slender valvular contour, and greater breadth height ratio, as well as in being more curved. As this form is not identical with other previously described species, it has been termed C. flexa n. sp. Ceratomyxa insolita n. sp. (Text-figs. 5–6, Figs. 68–9) Habitat. This species occurs in the gall bladder of Dactylopagrus macropterus (Bloch and Schn). It is not common, having been found in but a single host. The infected fish was taken by a trawler working out of Wellington in August. Trophic Stages. A few trophozoites were seen, rounded and immobile when examined, and probably dead. They are large, oval organisms, reaching up to 300μ in greatest diameter, and had no pseudopodia when examined. They are covered by a rather tough outer covering, apparently ectoplasmic, which in some individuals had become loosened and wrinkled. The endoplasm is very dark, and finely granular, with a tendency toward a fibrillar structure near the extremities. There are many small, granular inclusions and larger, refractive bodies. The latter are not conspicuous. Many spores can be seen within the trophozoites. Although they are numerous, details of structure cannot be seen in whole trophozoites because of the opaquepess of the cytoplasm. It was estimated that some of the trophozoites had as many as 50 spores. Spore Dimensions. Breadth across widest part of spore, 13.0–20.0μ (16.2μ), height, 10.0–13-μ (11.5μ). valvular axes, 19.0–26.0μ (23.8μ). and 17.0–23.0μ (21.5μ), capsules 10.0–13.0μ (11.5μ) by 3.0–4.0μ (3.2μ), and 7.0–9.0μ (8.5μ) by 2.5–4.0μ (3.4μ) Anterior angle, 16–41° (28°), posterior angle, 22–55° (41°, tangential angle, 13–48° (32°) Curvature index, 291°, taper index, 9°, valvular index, 13° It is evident that breadth height ratios are meaningless in a spore so strongly furcate. Spore Morphology. The unusual spore is furcate, with large, elongated, unequal polar capsules and asymmetrically placed foramina. The shape of the spore results in all assuming the same orientation in slides, with the sutural view presented. The anterior margin is sharply angled, with a large prominence extending upward at the foramen for the larger capsule, and a smaller one at the foramen for the smaller capsule. The anterior margin of the valves are somewhat convex lateral to the capsular foramina, the convexity extending to the valvular tips in some spores. or being interrupted at a level approximately opposite the posterior end

of the suture in others. The posterior margin tends to be nearly parallel to the more lateral part of the anterior margin, but there is some valvular taper in the most lateral parts of the spore. The valves are sometimes slightly constricted at a point just beyond the posterior end of the suture, where the sporoplasm terminates, but no septum is present. The two valves are always somewhat unequal, due to the differences in the capsular foramina, and usually differ by one or two microns in axial length. Otherwise they are quite similar. They meet in a heavy, prominent suture, which is straight. The polar capsules are unlike those seen in most Ceratomyxa, being much larger, more elongatedly pyriform, and having the filament loosely coiled, and more or less longitudinally arranged in the smaller capsule. In the larger capsule the filament is tightly coiled, up to 12 coils being present. A central, longitudinally arranged portion of the filament passes from the foramen to the lowermost coil, at least in many spores. It cannot be seen in some. The foramina are very prominent, and project unequally from the anterior margin, the larger capsule always projecting further forward. In some spores the capsules may project beyond the foramina a short distance. In most spores some remnants of the capsulogenous cells can be seen, surrounding the capsules and containing a capsulogenous nucleus. They are rather dark and granular in fresh spores (Fig. 68). The sporoplasm is large, extending about halfway toward the valvular extremities. It is rather finely granular in fresh spores. It is more or less symmetrical in position, and contains two rather small, vesicular nuclei. Spore Variability. The spore variability is not unusually great, tending to exhibit some of the tendencies seen in elongated species with empty valvular prolongations. The empty portions of the valves are sometimes deformed, and occasionally cross, although in such spores the membrane is usually wrinkled or otherwise abnormal. The anterior portion of the spore, containing the sporoplasm, capsules and foramina, is relatively stable, varying a little in size, but otherwise similar in all spores. Sixty per cent of the population have angular measurements between 20–32°, and 32–48° and 23–40° for anterior, posterior and tangential angles, respectively. The range for this portion of the population is about that seen in other species. Discussion. None of the previously described Ceratomyxa have a form very reminiscent of that seen in C. insolita. It evidently is a species which has differentiated in a direction not shared by most of the members of the genus. Nevertheless, the basic architecture is that of the Ceratomyxa, and it seems proper to place it in that genus. The very large capsules with the prominent, unequally elevated foramina are unique. The form has, therefore, been named C. insolita n. sp. It is the first Ceratomyxa to be described from a member of the Cheilodactylidae. Ceratomyxa gibba n. sp. (Text-fig. 6, Figs. 70–74) Habitat. This species occurs in the gall bladder of Congiopodus leucopaecilis (Richardson). It is not a very common species, and has been seen in but a single host, in which it was mixed with C. castigata. There was no evidence of irritation of the host organ. Trophic Stages. The trophozoites are almost indistinguishable from those of C. castigata. Young forms have not been sorted out. The larger, sporulating trophozoites attain a size somewhat larger than C. castigata, reaching a diameter of about 25μ when rounded. They are very transparent, although a little less so than C. castigata, and have a few inconspicuous granules. An occasional small, spherical inclusion, about 0.5–1.0μ in diameter, are found. These are apparently oily in nature. Only disporous trophozoites have been seen. Spore Dimensions. Breadth, 14.2–18.9μ (17.0μ), height, 5.6–8.0μ (6.9μ), valvular axes, 7.6–15.2μ (10.2μ) and 6.9–10.2μ (8.3μ), capsules, 2.5–3.3μ (2.8μ) by 1.8–2.5μ (2.3μ) and 1.9–2.9μ (2.3μ) by 1.5–2.2μ (2.0μ). Anterior angle, 78–113° (96°); posterior angle, 151–189° (170°), tangential angle, 134–162° (150°) Curvature index, 94°; taper index, 20°; valvular index, 74°. Breadth height ratios for quartiles of breadth range, 2.11, 2.59, 2.42 and 2.35. Spore Morphology. The curved spore is made up of unequal, different shaped valves and has unequal polar capsules. In sutural view the spore has a convex anterior margin, often curving outward toward the tips of the valves (Fig. 72). The two valves are often different in this respect, as in Fig. 72, although some are seen with both valves curving outwards terminally, when the spore assumes a broader. flatter shape (Fig. 73). One valve is usually larger in diameter than the other, and this tends to give the spore a hump-backed appearance. The valvular tips, on the average, are posterior to the suture. The valvular taper is quite variable, the shorter valves often being quite rounded in shape, while the more elongated.

valves sometimes terminate in quite narrow tips. The suture is much broader in immature spores (Fig. 74). In mature spores it is rather narrow, and slightly elevated. Few spores orient themselves so that a capsular view is obtained. Those seen in capsular view are straight, or nearly so, with the capsules opposite or slightly away from the midline. The polar capsules are equal to somewhat unequal. They are broadly oval structures, opaque, and situated near the anterior margin of the spore. They open near the midline of the anterior margin by way of very narrow, short necks. Capsulogenous cells are conspicuous in younger spores, and often remain as irregular, basophilic masses associated with the capsulogenous nuclei. The number of coils in the filament could not be counted. The sporoplasm is relatively large, but does not completely fill the spore cavity. It is variable in position, finely granular in structure, and contains two vesicular nuclei, which are rather small. In most spores several refractive spherules are seen, mostly outside of the sporoplasm itself. Spore Variability. Variability in this species, other than the usuall differences in spore size itself, centres on the differently shaped spore valves. Some valves are quite stubby, while others are elongated and more curved, while still others are elongated and less curved. Various combinations of valves produce spores with differing outlines, the range of which are suggested by Figs. 70, 72 and 73. Sixty per cent of the population sample measured have anterior, posterior and tangential angles ranging between 88–110°, and 145–162°, respectively. This breadth of the range containing 60% of the population is only moderately greater than that seen in most species, as most of the spores with nearly equal valves do not lie in the median 60% Discussion.. The only species of Ceratomyxa known from the Congiopodidae is C. castigata, described previously in this report. The two species differ most significantly in size, curvature and valvular taper (13.1μ and 17.0μ, 41° and 94°, 34° and 20°), insofar as the measured parameters are concerned. They are evidently not identical, as C. castigata has been seen in a host with C. gibba, and one without C. gibba. There was an almost perfect agreement in the two hosts of the C. castigata. populations. Of the various species which have been found in the closely allied scorpaenids, none are very similar to C. gibba C. yoichiensis (Fujita, 1923) has spores which are straight, and C. starksi (Jameson, 1929) has symmetrical, larger spores. The form of the typical spore fits with group II-B-1 or II-C-1, with the majority of spores falling into the latter group. In the former group, C. declivis and C. intexua are similar in breadth, as is C. obovalis (Fatham, 1930) Spore shape excludes C. intexua and C. obovalis at once, the former having straighter, more slender spores, and the latter more oval spores. The spores of C. declivis are only slightly smaller, the various dimensions alone not being sufficient to provide a dependable method of separation. However, the spores of C. declivis are less broad, relatively, and have prominent eosinophilic granules in Giemsa preparations, which are not seen in this species. The slight differences in angular measurements, like those in linear dimensions, are insufficient to provide dependable methods of separation, the means being no more than 5° different. However, the taper index (30° for C. declivis and 20° for C. gibba), which tends to be quite a stable index, seems to be dependably different. Actually, the differences between the two forms, when compared, is considerably greater than the raw data of measurement would suggest, as reference to the figures will show. Of the species in II-C-1, this form most resembles C. flexa It differs from C. flexa in being much less curved (curvature index 131° and 88°), in the inequality of the spore valves, as well as in lacking eosinophilic granules when stained with Giemsa. As this from does not correspond with other named species of Ceratomyxa, it has been designated C. gibba n. sp. Ceratomyxa gemmaphora n. sp. (Text-fig. 6, Figs. 75–82). Habitat. This species occurs in the gall bladder of Caesioperca lepidoptera (Bloch and Schn). Only a single host of this species was available for study. It was taken by a trawler working out of Wellington in September, and was heavily infected. The wall of the gall bladder appeared thickened, and the upper part of the gall

Text-fig. 6—Fig. 69—Ceratomyxa insolita. n. sp. from Dactylopagrus macropterus. Fig. 69—Stained spore. S. A., H. (f). Figs. 70–74—Ceratomyxa gibba from Congiopodus leucopaecilis Fig. 70— Fresh spore in sutural view F., U. (f). Fig. 71—Stained spore, S. A., H. (f). Figs. 72–73—Outlines of spores showing range of shapes F. U. (f). Fig. 74—Immature spore. F., J.G. (f). Figs. 75–82—Ceratomyxa gemmaphora n. sp. from Caestoporea lepidoptera Figs. 75–76—Trophozoites as seen in life. (b) Fig. 77—Outline of trophozoite releasing gemmae. (c) Figs. 78–76—Sutural and capsular views of fresh spores. M. G. (e). Fig. 80—Stained, somewhat flattened spore. S. A., H. (f). Figs. 81–82—Outlines of spores, showing range of shapes. F., U. (c). Figs. 83–87—Ceratomyxa subtilis from Coelarhynchus australis Fig. 83—Fresh spore M. G. (e). Figs. 84–85—Outlines of spores to show variation in curvature F. U. (e). Fig. 86—Outlines of spore in capsular view F., U. (e). Fig. 87—stained spore. S. A., H. (f). Fig. 88—Ceratomyxa lava from Caulopsetta scapha Fig. 88—Stained spore S. A., H. (f).

bladder and bile duct was red and inflamed. The bile was yellow, and apparently of normal viscosity. Trophic Stages. The trophozoites are extremely variable in shape and size, ranging from about 3μ to about 40μ in diameter Some are rounded, but the majority have long, slender, tapering pseudopodia, which are only rarely branched. The smallest forms are more frequently rounded but some are actively motile, when they assume a pyriform shape, with a trailing process Larger trophozoites, of about 5–7μ, tend to be more active than the smallest forms. They usually develop one or several small, sharply-pointed pseudopodia anteriorly, moving in the manner described by Davis (1917, p. 207) for a form from Cestracion zygaena. Ordinarily, a single caudal prolongation is present, although small remnants of the originally anterior pseudopodia are also seen Moving forms often incline the posterior pseudopodium, and without changing direction Occasionally, however, forms with no anterior pseudopodia, and two approximately equal posterior pseudopodia are seen (Fig. 76). Smaller and larger trophozoites often assume a floating posture. The majority of floating trophozoites are more or less triangular, with two anterior pseudopodia and a long posterior one (Fig. 75). These pseudopodia are often drawn out to a great length, some measuring 150μ from the tips of the anterior and posterior pseudopodia. Floating forms often occur with more than two elongated pseudopodia, sometimes four or five are present extending in different directions. These forms do not exhibit progressive movement, and in the few instances when they were seen to change from floating to motile forms, the pyriform shape was assumed before movement began. In addition to the motile and floating forms, some clavate individuals were seen This form is not common, apparently, and does not occur in motile individuals In some other. Ceratomyxa the clavate form appears to be used primarily for organisms attached to the wall of the gall bladder. Except in the pseudopodia, which consist entirely of ectoplasm, the ectoplasm is narrow and difficult to distinguish. It shows no visible structure. The endoplasm is relatively transparent, but less so than the ectoplasm, being shadowed in a finely granular pattern Small, more or less angular inclusions occur in small numbers, and a few larger, refractive spherules, apparently of an oily material, also occur. Nuclei cannot be seen in living organisms, but sporoblasts can be distinguished. The pseudopodia often develop small bleb-like enlargements of homogeneous material. There is little tendency of the trophozoites to aggregate, although masses of them may be attached to the wall of the gall bladder in the same vicinity, and small groups of floating forms are sometimes entangled together by the pseudopodia. There is no evidence of union between them. Rounded buds are produced internally. They are eventually released, sometimes in groups of up to five or six. It is not known whether this is a constant feature of this species, or whether the infection was in a state favouring their formation. However, they were quite abundant, more so than in any of the other forms that have been seen. The bud formation is undoubtedly the same phenomenon as that referred to by Davis (1917), who remarked that gemmules represented a common form of reproduction in the Ceratomyxidae. Spore Dimensions. Based upon the measurement of spores fixed in sublimate acetic, stored in alcohol and measured in water suspension. Breadth, 14.2–23.0μ (19.2μ); height, 5.9–8.3μ (7.0μ), thickness, 5.6–6.6μ (6.4μ); valvular axes, 7.8–13.2μ (10.8μ) and 7.3–12.2μ (9.4μ); capsular diameter, 2.0–3.4μ (2.1μ). Anterior angle, 94–130° (116°); posterior angle, 33°, valvular index, 82°. Breadth height ratios of quartiles of breadth range, 2.37, 2.65, 2.89, and 2.82. The presence of a few spores considerably more curved than the others resulted in bringing the mean figure somewhat below the medians for the angular measurements. The medians of 128°, 205° and 172° for anterior, posterior, and tangential angles, respectively, more closely describe the average spore. Spore Morphology. The spore is a broad ellipsoid, flattened to somewhat curved, with subequal valves and relatively small polar capsules. In sutural view the anterior margin is arched, with a slightly inset sutural line. The lateral portions of the valves, in some spores, are slightly flattened anteriorly (Fig. 82). The posterior margin is straight to gently concave, with a few spores having a moderate concavity of the posterior margin. The two valves are equal to sub-equal. but the average difference between them stands at the threshold of dependability. Valvular taper is moderate, and the valvular tips rather broadly rounded. The distinct, elevated suture is straight or nearly so. In capsular view the spore is flattened (Fig. 79) and straight. It is common for one valve to be somewhat narrower than the other in capsular view, giving the spore a tapered appearance This is also evident in slightly flattened spores in permanent preparations (Fig. 80).

The relatively small polar capsules are nearly spherical, and lie near, but not at, the capsular margin of the spore. In capsular view they are opposite or slightly rotated. In immature spores the capsulogenous cells are conspicuous, with large nuclei. The cells disappear entirely as the spores mature, but the nuclei are usually persistent. The rather large sporoplasm is somewhat granular and alveolar in fresh spores. It does not fill the spore cavity, and may be symmetrical or asymmetrical in its arrangement. In many spores it is drawn out into prolongations which adhere at their tips to the inner surface of the spore membrane (Figs. 80, 82). There are two small, vesicular, sporoplasmic nuclei. Spore Variability. As with most Ceratomyxa spores, there is some variability in spore curvature, the range being essentially that from the spore shown in Fig. 81 to a curvature a little greater than that shown in Fig. 78. It is not a particularly variable species, 60% of the population having anterior, posterior and tangential angles between 120–30°, 195–212° and 170–175°, respectively. Discussion. It is curious that no members of the Ceratomyxa have been reported from the Serranidae, although other closely related percomorph fishes are known to harbor several species. Of the species encountered in New Zealand, the rather closely allied Plagiogenion rubiginosus harbors C. flexa. The spores of C. flexa differ from this species in a number of ways, being smaller, more curved, and having eosinophilic granules when stained with Giemsa. C. gibba from Congiopodus leucopaecilis is also about the same size, but has more oval capsules, more unequal valves of a different shape, and is more curved. The average spore of C. gemmaphora falls near the borderline between groups II-B-1 and I-B-1. Of the species in group II-B-1 C. declivis is significantly smaller, and C. blennius is significantly larger. Of the species between these extremes in the list, C. intexua and C. scatophagi have spores of considerably less height, C. vepallida has spores of greater height, and C. gibba has already been mentioned. C. obovalis (Fantham, 1930) has a range of shapes and sizes, most of which are distinctly smaller than C. gemmaphora, and is generally more Leptotheca-like in appearance. The broader strains are not as high in relation to their breadth, and the smaller strains are significantly smaller. C. lata (Dunkerly, 1921) has more crescentic spores, with larger polar capsules. Of the species in I-B-1, C. castigatoides is significantly smaller, and C. starksi (Jameson, 1929) is significantly larger. Of the remaining species, C. maenae (Georgévitch, 1929) is a bit of a puzzle. It is described as producing microspores and megaspores, a trait which is certainly not seen in this species. If C. maenae should prove to be a mixture of two species, C. gemmaphora is not unlike the larger strain, but appears to differ in not having fusiform sporoblasts No polysporous trophozoites were seen in C. gemmaphora, another trait which Georgévitch gives for C. maenae. It seems quite improbable that C. maenae and this form are the same. C. yoichiensis (Fujita, 1923) has spores which are straight from all views, and also somewhat larger than those from Caesioperca. As this form cannot be equated with any of the previously described species, it has been designated C. gemmaphora n. sp. Ceratomyxa vepallida n. sp. (Text-figs., 6–7, Figs. 88–92). Habitat. This species occurs in the gall bladder of Caulopsetta scapha (Forster). It is rather common, having been found in at least five Caulopsetta, taken in winter, spring and summer. It is probably more common than appears superficially as the trophozoites are often present with few spores, and it is probable that it is sometimes present without being noticed. It occurs in combination with various other Ceratomyxa species. Trophic Stages. The trophozoites are very clear and transparent in life. They are usually attached to the wall of the gall bladder, often in clusters. Unlike most of the Ceratomyxa, attached forms do not tend to develop a clavate shape in most cases, and the attached trophozoites are shaped much like the motile forms. Motility is slow, and the pyriform body has a distinct, but relatively short, posterior process. There are few pseudopodia, slender in shape, with sharp tips. They very rarely branch. The largest forms reach a maximal diameter of about 25μ. The ectoplasm is clear, transparent, and without inclusions. The endoplasm is also very transparent, and nearly without inclusions. In bright field illumination, a few rather small

inclusions are sometimes seen. These are not highly refractive, and are always inconspicuous In dark contrast phase the nuclei become visible as dark objects, with differing optical densities and some evidence of interior structure, but details of structure cannot be seen. An occasional bright-rimmed inclusion, apparently fatty in nature. is seen In dark field illumination the trophozoites are almost invisible, except for the bright outline of the plasmalemma, and the outline of any sporoblasts which may be present. The transparency of the trophozoites, and their lack of visible differentiation in dark field illumination make these trophozoites easy to recognize in mixed infections. Trophozoites containing sporoblasts are easily recognized. They are round to oval, and are easily seen in bright field, dark field, or phase illumination. All sporulating forms have been disporous (Fig. 88). In some instances a portion of the endoplasm is more highly basophilic in haematoxylin mounts, retaining the stain after the remainder of the protoplasm is destained. Spore Dimensions. Based on small samples of spores from four hosts Breadth, 16.0–21.4μ (18.7μ); height, 7.8–9.6μ) (8.6μ); thickness about 8.1μ; valvular axes, 8.8–11.6μ (10.4μ), and 7.8–10.4μ (9.2μ), capsules, 2.9–3.6μ (3.2μ) by 2.5–3.2μ (2.9μ) Anterior angle, 94–117° (104°), posterior angle, 177–215° (194°); tangential angle, 142–180° (158°) Curvature index, 62°; taper index, 44°; valvular index, 90°. Mean breadth: height ratios for quartiles of breadth range, 1.88, 2.04, 2.26, and 2.30. Spore Morphology. Despite its common occurrence, most hosts contain very few mature spores. As spores do not mature well in depressions, relatively few are seen, and as a result dimensions are based on a series of small samples, as are the observations on spore morphology In all, however, several hundreds of spores have been examined. The small spore has rather stubby valves, equal to subequal in length, and terminating in broadly rounded tips. In sutural view the spores have a convex anterior margin, not interrupted at the suture and a posterior margin varying from flattened to moderately concave. The broadly rounded valvular tips are turned toward the posterior margin, in most spores. The two valves are usually nearly equal. They meet in a somewhat elevated suture, which is sometimes slightly oblique in its course. In capsular view the spore is straight or somewhat curved (Fig. 91). As in a number of other species of Ceratomyxa, the curvature of the spore is about equal in sutural and capsular views. The measurement of a small sample of 6 spores for horizontal angles resulted in means of 105°, 195° and 165° for the convex angle, concave angle, and horizontal tangential angle, respectively. These values are essentially similar to those found in the angles measured in sutural view. The capsules are somewhat variable in size, but are usually equal. Seen in capsular view they are opposite in position, or slightly rotated. In sutural view they are placed near, but not at, the anterior margin of the spore. They are broadly oval, with sharply narrowed necks. The filament cannot be seen in fresh spores. In immature spores, the capsulogenous cells are large and distinct (Fig. 90), but in mature spores the capsulogenous nuclei are usually missing. The sporoplasm is usually more or less central in position. It contains a number of small inclusions, some of which are quite refractive, and two vesicular nuclei usually lying close together. The sporoplasm fills but a small part of the spore cavity, and in the space between the sporoplasm and the spore membrance, a variable number of refractive spherules of apparently an oily nature are usually visible. Where trophozoites have disintegrated, spores are sometimes seen surrounded by membranous remnants, apparently of the sporoblast (Fig. 90). This is not uncommon in Ceraiomyxa from flat-fishes, and sometimes appears to add to the variety of shapes that the spores assume unless care is exercised. Triad spores are sometimes seen (Fig 92), especially in samples taken in the winter months. Discussion. This species falls into group II-B-1, with some of the spores approaching I-B-1 and II-A-1. The species in II-B-1 of simar breadth are C. gemmaphora, C. lata, and C. scatophagi C. gemmaphora is broader in relation to its height, and has trophozoites which are somewhat different, especially in the smaller size range C. scatophagi (Chakravarty, 1943), also, has spores with a considerably lesser height, as well as somewhat greater breadth C. lata, which occurs in Pleuronectes microcephalus (Dunkerly, 1921), has spores of the same breadth but somewhat less in height (19x7μ). It is somewhat more crescentic in shape. It seems certain that the form from Caulopsetta is closely related to C. lata. It is quite possible that when more information is available concerning the variability of the spore in size as well as shape, and the trophozoite of C. lata has been found, the two may be found to be identical. There seems to be too little evidence to make a positive identification, however. The spores in the New Zealand material which are about 19μ in breadth have a height of about 8.3μ, on the average. This is a greater difference in breadth height ratio that can ordinarily be expected in two.

samples belonging to the same species, according to the results obtained here, and it is tentatively concluded that the two forms are not identical. No other species in the bordering groups are very closely related to this species in the morphological sense. As the form has not been identified as belonging to previously described species, the name C. vepallida n. sp. has been given to it. Ceratomyxa subtilis n. sp. (Text-fig. 6, Figs. 83–87). Habitat. This species occurs in the gall bladder of Coelorhynchus australis (Richardson). It is not a very common species, having been found in a single host taken in the Wellington area in August, as a heavy infection, and in another taken in December as a very light infection. There was no evidence of abnormality in the host organ. Trophic Stages. No trophozoites were seen in the two hosts in which spores were found Trophozoites were seen in but a single host, and in this case no spores were seen. It is probable that these were trophozoites of C. subtilis. The trophozoites are slenderly pyriform, with a long posterior process terminating in a region of clear ectoplasm. The endoplasm is clear and transparent, containing a large number of rounded, inconspicuous inclusions, about 1 2μ in diameter. The inclusions tend to appear in a linear arrangement in the posterior portion of the body. The larges trophozoites measured about 25μ in diameter, and were not mature. No sporoblasts were seen. Spore Dimensions. Breadth, 15.7–26.0μ (21.5μ); height, 3.4–4.5μ (3.9μ), thickness, 3.9–4.5μ (4.3μ), valvular axes, 7.8–14.2μ (11.9μ) and 6.8–12.2μ (10.4μ), capsules, 1.5–2.0μ (1.8μ). Anterior angle, 68–150° (128°); posterior angle, 99–183° (164°), tangential angle, 98–180° (158°) Curvature index, 68°, taper index, 6°, valvular index, 36°. Breadth: height ratios for quartiles of breadth range, 3.94, 5.37, 5.20 and 6.57. Spore Morphology. The rather small, slender spore has equal to unequal valves, terminating in narrow, rounded tips. In sutural view the anterior margin is convex, with a rather regular curvature, either without interruption or with a very slight constriction at the suture. The straightest spores have the anterior margin of each valve nearly flat, and meeting at a rounded angle at the suture (Fig. 84). The posterior margin is nearly straight in the least curved spores, and distinctly concave in the more arcuate forms (Fig. 85). Although the position of the valvular tips is quite variable, it is generally somewhat below the end of the suture (posterior angle, 164°). The two valves may be equal or somewhat unequal (Fig. 84). They may be similar in shape, or one valve may be somewhat more curved than the other (Fig. 83). There is very little gradient of taper, but the spore height is so small that the valves terminate in quite narrowly rounded tips In capsular view, the spore may be straight slightly sigmoid, or somewhat bent. The spore tends to be slightly flattened horizontally, with the thickness a little in excess of the height. The relatively small polar capsules lie near the anterior margin, through which they open. They are oval, with narrow necks, and converge on the suture, quite sharply in fixed and stained spores (Fig. 87). The filament cannot be seen in fresh spores. Capsulogenous nuclei are usually missing in mature spores. The small sporoplasm is placed in the centre of the spore, sometimes extending but a short distance beyond the capsules, and sometimes extending nearly half-way along the valves. It is finely granular in fresh spores, and contains a number of very small inclusions, about 0.5μ or less in diameter, which are tinted by Bismarck brown, but not by methyl green. They disappear in fixed and stained spores. There are two small, vesicular nuclei. Spore Variability. The spores of this species are quite variable in curvature. The extremes of curvature are approximately those shown in Figs. 84 and 85 Spore breadth is also variable, partly because of differences in curvature, and partly because of differences in valvular length. The relative sizes of the two spore valves are also variable Sixty per cent of the spores have a breadth between 20.0 and 23.9μ, and angular measurements of 120–138°, 150–178°, and 148–171° for anterior, posterior and tangential angles, respectively This range of angular measurements for the median 60% of the spore population is rather wider than for most species. Discussion. No Ceratomyxa have been previously reported from the Macrundae. The typical spore form places this species in group II-C-1. Only C. lunata, of the previously described species, is similar in breadth. This form, however, is much more curved, and broader at the suture, with a height of 7–9μ (Davis, 1917). While there are a number of species with a slender spore form, all of these are much larger than C. subtilis, except for C. agglomerata, C. pallida and C. truncata Until.

C. pallida has been recovered and more fully described, it will be most difficult to recognize. According to Thélohan (1895), the trophozoite is 16–20μ in diameter, somewhat smaller than those seen in Coelorhynchus, and apparently without the inclusions which occur in the Coelorhynchus form. The spore is somewhat higher, also, and considerably greater in breadth. C. truncata appears to be significantly larger, and has much broader valvular tips (Thélohan, 1895) C. agglomerata is somewhat larger in all dimensions, with the sporoplasm very asymmetrically placed in most spores (Davis, 1917), and a very marked disparity in valvular shape. It is concluded that this form is previously undescribed, and it has been designated. C. subtilis n. sp. Ceratomyxa hama n. sp. (Text-fig. 7, Figs. 93–7). Habitat. This species occurs in the gall bladder of Caulopsetta scapha. (Forster). It is one of the most common of the Ceratomyxa occurring in its host, and is found all year round. It usually occurs in combination with other Ceratomyxa, but was seen alone in one host. It does not cause any visible abnormality in the host organ. It has been seen in fishes taken in the Wellington and Napier region. Trophic Stages. Good trophozoites were not available in the only host containing only. C. hama, and since there is generally a confusing mixture of species present, it is only with the greatest difficult that they can be sorted out. The problem is made more complex because there are several species of about the same size, with curved spores, difficult to distinguish in the immature state while still in the trophozoite As several of the types of trophozoites are quite distinctive in appearance, it would be valuable to make a much more thorough study of the. Ceratomyxa fauna of Caulopsetta scapha. The trophozoite which is believed to belong to this species has been recognized, primarily, by a comparison of the frequency of mature spores of three types which occurred together so often, and a comparison of the frequency of the spores. The trophozoites are relatively large some reaching a length of 125μ when extended, although rounded forms are much smaller. They have a tendency to occur in two different forms, a motile form and an attached form. The motile form is a modified pyriform shape, usually with a definitely enlarged region, in which any spores that are present occur (Fig. 93). The protoplasm extends backward in a large, elongated region, rather like the posterior process of smaller Ceratomyxa species, although so much larger in size. The shape gives the effect of clavateness rather than that of a pyriform organism, because of the large quantity of protoplasm in the trailing portion Movements are never fast, and many trophozoites appear not to move at all. Attached trophozoites are clavate, usually with an enlarged region containing any developing sporoblasts. They are attached by one or several short, slender, tapering pseudopodia at one end. The elongated portion of the body is usually gently curved, and extends outward into the bile. At the free end of the trophozoite one or several slender pseudopodia may be present. A common phenomenon is the presence of three or four enlargements on the pseudopodia. These enlargements are quite distinct, and are about the size of the smallest trophozoites. It is thought that they are buds, for the production of new individuals. The ectoplasm is transparent, and without inclusions. It is without structure in bright field, dark field, or phase contrast. The endoplasm is also quite transparent in bright field illumination. It contains two kinds of inclusions, in addition to nuclei and developing sporoblasts. The most numerous inclusions are indistinct, spherical bodies, about 1.0–1.25μ in diameter. These occur in large numbers in mature trophozoites. They are not highly refractive and are rather inconspicuous despite their large numbers. They are much more conspicuous in dark contrast phase, where they appear as dark bodies. They cannot be seen in dark field. They often occur in rather regular lines in the elongated portion of the trophozoite. They are persistent in fixed and stained slides, even surviving the Feulgen technique. In permanent preparations they are coloured by the acidic dye. Among these inclusions there is a small, variable number of irregular dark granules. They are brownish in colour, and usually less than 0.4μ in diameter In smaller trophozoites they occur in very small numbers, about 5 to 7 of then being the usual complement. As the sporoblasts appear and develop, they become a little more abundant, but are never a very conspicuous portion of the endoplasm. These granules appear as bright objects in dark field illumination. Their presence makes it possible to distinguish the trophozoites from C. vepallida, which lacks them, and C. laxa, which has a great abundance of them.

Text-fig. 7—Fig. 89–92— C. vepallida in sp. from Caulopsella scapha Fig. 89—Slightly oblique view of spore F. U. (c). Fig. 90—Immature spore in membranous sporont F.J.G. (e). Fig. 91—Outline of spore in capsular view F. U. (b) Fig. 92—Outline of spore with triad form F., U. (e). Figs. 93–97—C. hama n. sp. from Caulopsetta scapha Fig. 93—Trophozoite as seen in life Motile form × 300 Fig. 94 Sutural view of fresh spore U. (d). Figs. 95–96—Outlines of spores in capsular and sutural views F., U. (a). Fig. 97—Outline of spore with triad form F. U. (a). Fig. 98—C. lava n. sp. from Caulopsetta scapha Fig. 98—Trophozoite as seen in life ×300.

The trophozoites are monosporous or disporous. No polysporous forms have ever been encountered. In this respect they differ from the trophozoites of C. laxa, which are sometimes polysporous. Spore Dimensions. Based on spore samples from three hosts. Breadth, 18.5–29.8μ (23.4μ), height, 6.8–8.3μ (7.4μ), thickness, 7.0–8.3μ (7.7μ), valvular axes, 9.3–17.5μ (14.4μ) and 8.3–16 1μ (11.9μ); capsules, 2.0–3.5μ (3.2μ). Anterior angle, 59–117° (101°); posterior angle, 107–205° (160°); tangential angle, 98–179° (136°) Curvature index, 99°, taper index, 24°; valvular index, 59°. Convex angle, 63–113° (91°), concave angle, 103–179° (139°), tangential angle, 89–152° (111°). Bending index, 130°, horizontal taper index, 28°, valvular index, 48°. Breadth height ratios for quartiles of breadth range, 2.57, 3.04, 3.46, and 4.08. Spore Morphology The thin-walled spore is curved, bent, or both curved and bent. The anterior margin is convex, gently or sharply, depending on spore curvature, and the posterior margin flattened or concave, again depending on the direction and sharpness of spore curvature (Figs. 100–102). The valves in the majority of spores are similar in shape although often slightly unequal in length In some spores, however, one valves is more inflated. and the other terminates in a narrower tip (Fig. 94). The two valves meet in a straight suture, slightly elevated, and rather narrow. Because of the variety of curvatures affecting the spore, the overall spore shape is extremely variable In spores with a curvature in both directions, the essentially spiral shape can be seen when seen in valvular view. The capsules are nearly spherical, but narrow sharply as they approach the anterior margin, through which they open. They converge sharply on the suture. In fixed and stained preparations the capsules are more pyriform in shape. The filament is easily seen in the capsules of fresh spores, making from four to six coils. When extruded, the filament reaches a maximal length of about 60μ. Capsulogenous nuclei are usually persistent in stained spores (Fig. 102). The sporoplasm is fairly large, and may be symmetrical in position or placed far off centre, nearly all occurring in one valve. It is finely granular, and contains a few refractive spherules. The spore cavity usually contains a few oily spherules. When spores are first exposed to neutral red in depression slides, the spherules are not stained. During the first 24 hours they grow, and become deeply tinted with the dye. After about two more days, the colour begins to fade In a week or so the large spherules break down into many smaller ones, and the sporoplasm is noticeably smaller in size. It would appear, therefore, that some metabolic activity goes on in the spores. At the end of a month, many spores have no sporoplasm left at all, although the spore shape is still maintained by the membrane, while others have a small, rounded mass of sporoplasm containing one or two nuclei Still others, apparently having died earlier, have large sporoplasms in which changes in refractivity have occurred. Spores stained in Giemsa (Fig. 99), contain a number of bodies which are deeply stained with eosin. These are, in some ways, similar to those seen in C. declivis, C. flexa and C. arcuata, but differ in that they can be seen, if indistinctly, in permanent mounts stained with haematoxylin (Fig. 102), as well as in having a quite different form. While some small, rounded or irregular bodies occur, in each spore there are one or two centres about which the material tends to be disposed. These are placed laterally usually beyond the end of the sporoplasm. Similar bodies are sometimes seen in spores of C. laxa (Fig. 104), although they are less conspicuous in that species In mature spores, the main centres of the eosinophilic material tend to disperse, and the substance is dispersed in the spore cavity. Here it forms more or less filamentous structures, often spiral in form, as well as amorphous masses. The spiral form of some of the inclusions is very evident. No conclusions have been drawn concerning the nature of this material. Spore Variability. There is a very extensive variability of the spores of this species largely associated with the amount and direction of spore curvature. With the wide range of spore shapes resulting from the variable curvature, as well as the occurrence of somewhat similar curved species of about the same size, difficulties are constantly encountered in orientation of spores and in being sure that no errors are made in placing them in the appropriate categories. As a result, it is felt that the measurements, particularly of angles, are somewhat less dependable in this species than in most. As a result, no extensive analysis of variation has been undertaken. Despite all of the difficulties, however, the measurements recorded proved to be relatively reproducible. In two hosts containing large numbers of spores of this species. and in which neither C. torquata nor C. laxa were seen, and in which more confidence can be felt in the samples the two samples had identical breadths, mean height and thickness within 0.1μ mean valvular axes differing by 0.5 and 0.1μ, and capsules differing by 0.4μ in diameter. Differences in anterior. posterior and tangential angles were 6°, 7° and 1° in the two samples. while differences in means of convex. concave and horizontal tangential angles were 16° 15° and 12°. Other smaller samples, taken from hosts containing more different species and in which more possibility of error occurred matched less well but differences were not greater than those found in other curved species of Ceratomyxa. Despite these results however, the impression

was gained that the consistency of angular measurement was more a matter of chance than of accuracy in measurements, and that errors in spore orientation were more or less comparable in the various samples measured. Where spores may be both curved and bent, or either curved or bent, the angular measurements are much more difficult to take, and the technique yields results in which one feels rather less confidence. Perhaps the most definite statement that can be made is that about the same range of sizes and shapes were seen in each infected host, and that there was no evidence of a greater frequency of any kind of shape in one sample, as compared to another, as might be anticipated if the differences in spore shape were characteristic of different strains. There are many species of Ceratomyxa reported in various species of flat-fishes. Over 20 have been described previously from different localities. It is not impossible that these forms, like limpets and some other animals, have a tendency to occur as local species, with relatively restricted geographical distribution areas. In general, there seems to be a strong tendency for the species to be restricted to flat-fishes, but little tendency toward a host specificity within the flat-fish group. It is also possible that we are dealing with a much smaller number of extremely variable species. At the present time there is insufficient evidence to make firm conclusions possible. In the New Zealand material, there has been difficulty in sorting out the spores found in Caulopsetta scapha into meaningful categories which showed any consistency. It seems that there are several rather variable strains, all having curved spores of about the same breadth, forming a complex which is most confusing. As spores are sorted out into categories, it is found that it is quite easy to have more categories than species, and, perhaps, no less easy to have more species than categories. Eventually, a set of categories was developed which appeared to more or less consistently provide places for the spores that were seen, and to correspond to the trophic stages which were also seen, and some degree of confidence in those categories developed. It is, neverthless, possible that there are more than the number of species recorded. Fujita (1923), found material indicating that truncations of the valvular tips, and directions of spore curvature provided dependable criteria for species groupings. In the New Zealand species, the attempt to include valvular truncation in the categories resulted in considerable confusion, and would have required at least three more species groups, which were always associated with other species, so this criterion was not used. Direction of curvature, also, appeared to be independable in this species, as spores with essentially similar valves exhibited horizontal, vertical and combined curvatures. and appeared always to occur together. It is true, however, that much more constant species groups could be described if the horizontally, vertically, and doubly curved spores were separated into different species. On two bases, I am inclined to have no faith in this kind of distinction in the material from Caulopsetta. The three types of spores occur again and again in combination, and, in addition, there is a certain similarity in the angular measurements of spores seen in sutural and valvular views. Thus the means for the three angles measured in sutural view were, 101°, 160° and 136° and the three measured in capsular view were 91°, 139°, and 111°. These values are similar, as if the spores tended to be bent about 10–15° more in capsular than in sutural view. It has been noticed that when spores are bent in both sutural and capsular views, there is a tendency for the angular measurements to be similar in the two views, as if the factors resulting in spore curvature acted independently of those determining capsular position. Discussion. The spores of this species fall into group II-B-1, with some spores having so little valvular taper that they fall in II-A-1. In the latter group. C. arcuata is most similar in breadth, but differs in having spores which are more slender. In II-B-1, C. truncata, C. elegans, C. constricta, C. moenei, and C. amorpha are similar in spore breadth, as well as C. laxa and C. torquata, to be described subsequently from Caulopsetta. In none of these species is the spore both curved and bent as in this one. While some of Fujita's (1923) species from Japanese flat-fishes have somewhat similar form, none are as small as this species. It may be that if clines can be shown connecting the Japanese forms with those from New Zealand, and resulting in a consistent change of size, some may prove to be identical with the New Zealand species, but the difference in size is much too great to permit their being considered identical without such evidence. It must, therefore, be concluded that this species is not previously described, and it has been named C. hama n. sp. Ceratomyxa laxa n. sp. (Text-figs. 7–8, Figs. 98, 103–107) Habitat This species occurrs in the gall bladder of Caulopsetta scapha (Forster). It is relatively common, particularly during the warmer months, and has been seen in at least seven Caulopsetta, taken by trawlers in the Wellington area. There is no evidence of abnormality in the host organ.

Trophic Stages. Motile trophozoites of this species have not seen seen, even in cases where other trophozoites were in fairly rapid motion. Most of them are attached to the wall of the bladder, or float freely in the bile. Attached forms are sometimes nearly clavate, but the majority of trophozoites are essentially oblong in form, with prominent, branching pseudopodia. In attached forms, slender pseudopodia are often thrust between the host cells. In floating forms, pseudopodia tend to be formed over most of the body surface, and extend outward in all directions (Fig. 98). They sometimes branch, and it is not uncommon to see a small tuft of delicate, needle-like pseudopodia at the end of a larger pseudopod. The clear, transparent ectoplasm is without inclusions, and forms a continuous zone about the body. The endoplasm is highly granular, and darkened by hundreds of small, irregular, brown bodies which mask the remaining inclusions, and make it easy to recognize the trophozoite in mixed infections. In both dark field and dark contrast phase illumination, the granules appear as bright bodies. They are similar in colour and general appearance to the granules seen in C. hama, but are much more numerous. Among the granules, but usually obscured by them, are a few spherical inclusions, appearing as white-rimmed bodies in dark contrast phase and in dark field. These are apparently fatty in nature, and disappear in permanent slides, while the granules persist. The nuclei, which cannot be seen in dark field or bright field illumination, can be seen in dark contrast phase. They are difficult to make out because of the masking effect of the granular inclusions. Sporoblasts and developing spores are conspicuous, and as they lack the granular material, appear as lighter areas. In the majority of trophozoites from two to eight spores are formed. An occasional unisporous form is also seen This is the only polysporous species that has been observed in Caulopsetta. The largest trophozoites reach a length of about 80μ, not including the pseudopodia. The pseudopodia are often very long, reaching a maximum length about equal to that of the body mass. Spore Dimensions. Based on samples from two hosts. Breadth, 20.8–30.3μ (25.6μ), thickness, 7.8–9.8μ (8.7μ), valvular axes, 9.8–19.0μ (15.7μ) and 8.2–15.6μ (14.5μ), capsules 2.9–3.9μ (3.4μ) Convex angle, 60–116° (88°), concave angle, 103–204° (147°); horizontal tangential angle, 93–173° (130°). Bending index, 125°; taper index, 17°; valvular index, 59°. Breadth thickness ratios for quartiles of breadth range, 2.50, 2.80, 3.22, 3.54. Spore Morphology. The spore is bent, with relatively large, round polar capsules, and is flattened horizontally. Seen in capsular view, one border is convex, sometimes with a slight constriction near the suture (Fig. 105), and sometimes passing the suture without interruption (Fig. 103) The two valves are broad at the suture, and taper moderately to rather broadly rounded tips, in most spores, but are rather stubby in others, terminating in very rounded tips. The principal variability appears to be in the length of the valvular axes, some spores having much longer valves than others, relatively speaking. The extent of variability is about that indicated in Figs.105 and 104. The two valves are equal to distinctly unequal in axial length, and meet in a prominent, straight suture. In sutural view, which is seen but rarely, the spore is flattened, with a straight or slightly curved axis. The two capsules are round, nearly spherical, and with short, narrow, inconspicuous necks. The filament is prominent in life, and is coiled about five or six times, insofar as could be seen in the few spores seen in sutural view. The capsules sometimes converge somewhat on the suture, and are sometimes slightly rotated. While the capsulogenous nuclei are sometimes seen in stained spores, they are not usually persistent. There is no remnant of the capsulogenous cells in mature spores. The filament, when extruded, extends up to about 55μ in length. The sporoplasm is relatively large, and is usually somewhat asymmetrical in position. It is quite transparent, showing a finely granular structure. It contains a few refractive spherules in fresh spores, usually small in number and in size (Fig. 103). These disappear in permanent preparations. The two nuclei are vesicular, and usually placed close together. As spores age in depressions, the sporoplasm shrinks in size, and tends to form a spherical mass located in one valve. Spores stained in Giemsa have a small amount of eosinophilic material, usually in two centres, located near the end of the sporoplasm in each valve (Fig. 104). This material is much less abundant than in C. hama, but is evidently of a similar nature, and tends to form spiral filaments as it disperses. Spore Variability. While the spores have been seen in a number of hosts, they were usually mixed with other species, particularly C. hama and C. torquata, so that measurement was deemed likely to include spores of other species as well as this one. In the two hosts in which spore identification could be most certain, and in which relatively heavy infections occured, the samples proved to differ somewhat in linear dimensions, but to have very similar angular measurements. The two samples differed by 3.6μ in spore breadth, 0.2μ in spore thickness, and the valvular axes differed by 2.3μ and 1.4μ The differences were occasioned,

primarily, by the occurrence of a larger number of spores in one sample in which the valves were longer, and somewhat more tapered. The capsules were identical in size. The convex, concave, and horizontal angular measurements differed by 4°, 2° and 5° respectively. While the smaller samples measured in other hosts differed in the means from the two larger, more dependable samples, all were between the extremes set by the two larger samples insofar as linear measurements are concerned, and while the angular measurements were somewhat more variable, the spores all fell within the extremes set by the larger samples, and had means within 15° of those seen in the larger samples. It would appear, therefore, that this species varies primarily in valvular length, and that differences in angular measurements are, on the whole, relatively slight insofar as means are concerned. Discussion. Of the various species of Ceratomyxa occurring in flat-fishes, this species is most like C. hama and C. torquata. Until it is understood that this species is horizontally flattened, whereas the other two are nearly circular in cross section, it is most difficult to sort them out, at least under lower magnifications. Like C. hama, C. laxa falls into a size range which is well below the species found by Fujita (1923) in Japanese flat-fishes, except for C. toitae. C. toitae, however, has spores which are much larger at the sutural line, measuring 13μ in sutural diameter, and has small capsules, 2μ in diameter, with filaments but 20μ in length. This species falls into group II-B-1. None of the species in its general size range is bent rather than curved, except C. torquata, from which it differs in its horizontal flattening, and in its straight suture. It is concluded that it is a previously undescribed species, and has been given the name C. laxa n. sp. Ceratomyxa arcuata Thélohan (Text-Fig. 8, Figs. 108–110) Habitat. This species occurred as a light infection in one specimen of Anthias pulchellas. The host fish was taken by a trawler working out of Wellington in July. Two other Anthias taken at the same time were not infected. There was no evidence of damage to the host organ. Trophic Stages. These have not been seen. Spore Dimensions. Based on spores from permanent slides, after fixation in sublimate acetic, and staining in haematoxylin or Giemsa. Breadth, 16.2–36.7μ (25.6μ); height, 4.5–7.8μ (5.7μ); thickness, 5.4–5.9μ (5.7μ); valvular axes, 13.2–24.4μ (18.8μ) and 10.3–21.1μ (15.0μ); capsules, 1.5–2.4μ (1.9μ). Anterior angle, 45–90° (74°); posterior angle, 63–115° (99°); tangential angle, 62–112° (96°). Curvature index, 187°; taper index 3°; valvular index, 25°. Breadth height ratios of quartiles of breadth range 3.34, 3.75, 5.40 and 5.33. Only a small sample of spores were measured for thickness. Spore Morphology. The slender, arcuate spore has small polar capsules and rounded valvular tips. In sutural view the anterior margin is strongly convex, and the posterior margin strongly concave (Fig. 108). A very moderate valvular taper is present, and the valves terminate in relatively broad to somewhat narrow valvular tips. In some spores, the lateral extremity of the valve is slightly inflected toward the anterior margin, breaking the otherwise smooth curvatures of anterior and posterior margins (Fig. 108). While the two valves are equal in some spores, they are often quite unequal in length, with a mean difference of over 3μ in axial length. They meet in a suture which is scarcely elevated, and is inconspicuous in stained slides. The polar capsules are quite small, somewhat oval in shape, averaging about 0.1–0.2μ less in breadth than in length. They are somewhat convergent on the suture. In capsular view, they are opposite (Fig. 109). The capsulogenous nuclei are usually persistent, and in many spores remnants of the capsulogenous cells can also be seen. The fairly large sporoplasm is usually centrally placed, and is quite transparent in fresh spores. It contains two nuclei with rather conspicuous endosomes. The spores contain a number of eosinophilic bodies which are deeply stained in Giemsa (Fig. 110). These are essentially like those seen in C. declivis and C. flexa, being generally distributed in the spore, and not showing the spiralled form seen in C. laxa and C. hama. They tend to be somewhat larger than those seen in C. declivis, and to be more frequently located in the sporoplasm. Discussion.. In size and shape this species is very like C. arcuata. While Thélohan (1895) refers to the valves as being of equal length, Parisi (1912) figures spores with as much difference in valvular length as seen in the New Zealand.

Text-fig. 8.—Figs.99–102—C. hama n. sp. from Caulopsetta scapha Fig. 99—Spore stained with Giemsa showing eosinophilic bodies. S.A., G. (f). Fig. 100—Spore bent, but not curved. F., U. (d). Fig. 101—Slightly oblique view of spore curved, but not bent. F. U. (d). Fig. 102—Spore, stained with haematoxylin. S. A., H. (f). Figs. 103–107—C. laxa n. sp. from Caulopsetta scapha Fig. 103—Fresh spore with elongated valvular axes. F., U. (d). Fig. 104—Spore, stained with Giemsa, showing eosinophilic bodies. S. A., G. (f). Fig. 105—Spore, with stubby valves. F., U. (f). Fig. 106—Sutural view of spore. F., U. (f). Fig. 107—Spore, stained with haematoxylin. S. A., H. (f). Figs.108–110—C. arcuata from Anthias pulchellis Figs.108–109—Outlines of spores. S. A., G. (d). Fig. 110—Spore, stained with Giemsa, showing eosinophilic bodies. S. A., G. (d). Figs.111–114—C. constricta n. sp. from Centriscops humerosus Fig. 111-Spore with triad form. F., U. (d). Fig. 112-Capsular view of spore. F., U. (d). Fig. 113—Outline of spore with slender form. F., U. (d). Fig. 114—Sutural view of spore. F., U. (d).

material. Parisi also shows spores with a tendency to be curved outward at the lateral extremities of the valves. Lateral appendages were not seen in spores from Anthias. They may or may not be present in C. arcuata. Judging by Parisi's sketches, the agreement in shape is as nearly perfect as the agreement in size. While C. arcuata has not been found in members of the Serranidae, it has been reported from the Moronidae by Dunkerly (1921), who observed it in Roccus labrax. It seems highly probable that the New Zealand form is identical with the strains that have been described as C. arcuata. Ceratomyxa constricta n. sp. (Text-Figs.8–9, Figs.111–115) Habitat. This species occurs in the gall bladder of Centriscops humerosus (Richardson). It is apparently not a common parasite, having been found in but one of six fishes examined. The infected fish was taken by a trawler working out of Wellington in August. There was no evidence of damage to the host organ. Trophic Stages. These are unknown. Spore Dimensions. Breadth, 23.6–29.3μ (26.9μ); height, 5.6–9.0μ (8.1μ); valvular axes, 13.5–15.8μ (14.7μ); and 11.9–15.2μ (13.7μ); capsules, 2.8–3.9μ (3.5μ); by 2.8–3.4μ (3.2μ); Anterior angle, 105–131° (113°); posterior angle, 150–193° (178°); tangential angle, 128–169° (145°); Curvature index, 69°, taper index, 33°; valvular index, 65°. Breadth: height ratios for quartiles of breadth range, 2.88, 3.12, 3.68 and 3.54. Spore Morphology. The rather plump, slightly curved spores have approximately equal valves with a definite medial constriction in which the suture lies. In sutural veiw the anterior margin is gently convex, forming a continuous curve broken only by the sutural constriction. The posterior margin of each valve is flattened or slightly convex, but the two valves meet at an obtuse angle at the suture (Fig. 114). The valves are nearly equal in axial length, and taper very moderately to broadly rounded tips. They meet in a prominent, narrow, elevated suture. In capsular view the spore is straight or somewhat curved. In a few spores the valves are bent in opposite directions, so that the spore has a slightly sigmoid shape in capsular view. The capsules are nearly spherical, with very short, inconspicuous necks. They are placed close to the suture, often touching each other. They are usually flattened at the point of contact. In capsular view they are opposite. They converge slightly on the suture. The filament is not conspicuous, but four or five coils may be seen in fresh spores. Capsulogenous nuclei are usually persistent, and lie near the base of the capsules. The sporoplasm is fairly large, and arranged either symmetrically or asymmetrically. It is quite transparent, but contains a few refractive spherules as well as two vesicular nuclei, with prominent endosomes. The space between sporoplasm and spore membrance usually contains a few refractive spherules, which disappear during the preparation of permanent slides. Spore Variability. The spores of this species are not particularly variable, the principal differences being in the amount of bending of the spore as seen in capsular view, and the spore height. There is quite a significant difference between the spore height at the two extremes, the range being approximately that shown in Figs. 113 and 114. There are no apparent relationships between the spore height and amount of curvature seen in capsular view. For the angular measurements, 60% of the spore population fell between 106–121°, 165–185° and 136–158° for anterior, posterior and tangential angles, respectively. Discussion. No Myxosporidia have been reported from members of the Macrorhamphosidae, or, for that matter, from members of the Order Aulostomi. The form of the spore is such that it falls in group II-A-1 or II-B-1, depending on valvular taper. Of the species in the former group, C. intexua is significantly smaller, and C. hopkinsi significantly larger. C. arcuata and C. hama differ from this species in being more slender and more curved, and C. torquata has a sinuous sutural line. C. crassa (Jameson, 1929) has similar dimensions, but the spore has less valvular taper and lacks the medial constriction in the sutural area. Of the species in group II-B-1, C. scatophagi (Chakravarty, 1943) appears to be significantly smaller, and C. moenei significantly larger, insofar as breadth is concerned. Most of the species occurring between these extremes may be eliminated on the basis of the degree of spore bending. Of those remaining, C. monospora has more taper to its valves, and the valves are dissimilar in shape (Davis, 1917). C. undulata

(Davis, 1917) has more slender spores, with narrower valvular tips C. elegans (Jameson, 1929) has spores without a medial constriction, and with a less angled posterior margin, as well as valves of somewhat differing shape. Since it is not possible to equate this form with any of the previously described species, it has been given the name C. constricta n. sp. Ceratomyxa torquata n. sp. (Text-Fig. 9, Figs. 116–7) Habitat. This species occurs in the gall bladder of Caulopsetta scapha (Forster). It is not uncommon, having been seen in at least five Caulopsetta. It apparently occurs throughout the year, although rather more of it has been seen in the summer months. No evidence of damage to the host organ has been observed. Trophic Stages. The trophozoites have not been accurately separated from those of C. hama, which they very closely resemble. It appears that they may have rather fewer of the brownish granules than the trophozoites of C. hama in organisms of the same size, but distinctions between them are most difficult to observe. Spore Dimensions. Based on a full sample from one host, and small samples from three other hosts Breadth, 20.5–30.3μ (26.9μ); height, 7.8–9.5μ (8.1μ); thickness, 8.3–10.1μ (9.2μ); valvular axes, 14.0–22.6μ (16.7μ); and 10.8–18.0μ (14.4μ); capsules, 2.8–4.7μ (3.8μ); Anterior angle, 85–134° (99°); posterior angle, 126–207° (151°); tangential angle, 115–178° (135°). Curvature index, 110°; taper index, 16°; valvular index, 52°. Convex angle, 64–116° (91°), concave angle, 103–204° (148°), horizontal tangential angle, 95–173° (133°); Bending index, 121°; taper index, 15°; valvular index, 57° Breadth height ratios for quartiles of breadth range, 2.54, 2.97, 3.41 and 3.76. Spore Morphology. The thin-walled spores are usually bent more than curved, and have nearly equal valves meeting in a sinuous suture. Spores are more often seen in capsular view, as they are slightly flattened horizontally, but as the valves often curve in different directions, it is somewhat difficult to specify spore orientations with accuracy. In the spore shown in Fig. 116, the valve at the right is lying horizontally, while the valve at the left is curving downwards. Seen in valvular view, as one focuses along the length of the spore, the valves tend to follow a spiral course in their curvature. The end result is that one tends to see one valve from a capsular aspect, and the other from a sutural aspect, as it were, so that one valve tends to be somewhat narrower than the other. The two valves meet in a conspicuous, rather narrow, elevated suture which takes a sinuous course. In stained slides, the suture often appears as a double line, but this cannot ordinarily be seen in fresh spores (Fig. 117). The capsules are nearly spherical, with a narrow neck region. They lie near the suture, sometimes touching, and flattened at the point of contact. In some spores the capsules are somewhat rotated, but in most they are opposite. They converge very little on the suture. Capsulogenous nuclei are usually missing in mature spores. The filament is conspicuous in fresh spores and forms about five or six coils. It reaches a maximal length of about 60μ when extruded. The sporoplasm is rather large, and variously situated in the spore. It is finely granular, containing a few small inclusions, and usually several inly droplets lie outside of the sporoplasm, between its margin and the spore membrane. These react to neutral red, and show changes like those recorded for C. hama. The two sporoplasmic nuclei, usually placed close together are vesicular, and contain rather prominent peripheral chromatin. Spore Variability. There is considerable variability in the extent of spore curvature, and in the extent to which the valves extend in opposite directions in their curvature. The difficulty of obtaining satisfactory spore orientation for the measurement of angular dimensions makes it uncertain whether they are accurate representations of the spore shape. On the whole, there was good agreement between the smaller samples and the full sample, with means for the angular measurements within about 15°. Discussion. The spores of this species fall in groups II-A-1 or II-B-1, depending on the amount of valvular taper, the majority belonging in the latter group. In general size and shape of the spore, this species is far more like C. hama and C. laxa than like other species from other regions of the world. It differs from C. hama in being somewhat less slender at the suture, and in having a sinuous suture. It differs from C. laxa in having much less flattening of the spore in the horizontal plane, and in having a sinuous suture, as well as in having trophozoites which resemble those of C. hama. It does not very closely resemble any of the other species in these

two groups, and is considerably smaller than most of the species found by Fujita (1923) in Japanese flat-fishes. It is concluded that this species is previously undescribed, and the name C. torquata n sp. is proposed for it. Ceratomyxa renalis n. sp. (Text-Fig. 8, Figs. 118–120) Habitat. This species occurs in the urinary bladder of Caulopsetta scapha (Forster). It extends upwards into the ureters, but it was not recovered from the kidney, and presumably does not reach the renal tubules. There was no evidence of abnormality in the host organ. The infected fish was taken by a trawler working out of Wellington in December, and was heavily infected with Sphaerospora undulans, to be described in a subsequent section of this report. It is apparently not a common species, as it was seen in but a single host. Trophic Stages These have not been seen. Spore Dimensions Breadth, 22.5–33.4μ (27.7μ); height, 7.3–8.5μ (8.0μ); thickness, 7.7–9.3μ (8.9μ); valvular axes, 14.7–20.1μ (17.4μ); and 8.1–19.4μ (16.0μ), capsules, 2.7–3.2μ (3.1μ) Convex angle, 67–95° (84°); concave angle, 105–143° (128°); horizontal tangential angle, 90–134° (118°); Bending index, 148°, taper index, 10°, valvular index, 44°. Breadth thickness ratios for quartiles of breadth range, 2.75, 3.20, 3.50 and 3.92. Spore Morphology. The crescentic spore is bent rather than curved, and is made up of subequal valves which taper moderately to moderately narrow valvular tips. The spore is ordinarily seen in capsular view, when the convex margin curves smoothly, without interruption at the suture. The concave margin is smoothly curved in some spores, but is frequently somewhat flattened along each valve, and forms an obtuse angle at the suture. The two valves are often unequal in axial length (Fig. 120), and sometimes are different in shape as well, one being broader than the other. The longer valve is usually the more slender in these spores. The valves meet in a distinct, slightly elevated, straight suture. The spores seen in sutural view appear to be somewhat flattened, but this orientation is not long sustained by spores. It appears that the posterior margin is somewhat concave in sutural view, but as the spores were moving it may be that they were somewhat oblique when this impression was gained. The suture is straight when seen in sutural view. The capsules are set near the suture, and open near the centre of the anterior face of the spore. They are nearly spherical, but slightly elongated in sutural view, the length being about 0.2μ in excess of the diameter. They converge slightly on the suture of some spores. Capsulogenous nuclei are persistent, but capsulogenous cells are, at the most, very small. The filament is not entirely obscured by the capsule wall, but the number of coils could not be determined. The fairly large sporoplasm does not fill the spore cavity. It is finely granular, with a few dark inclusions about 0.5μ in diameter. A few refractive spherules, either in or around the sporoplasm, are usually present. The two vesicular nuclei are usually placed close together. Spore Variability. As the infection was not a heavy one, the variability is not as fully known as for some species. It apparently centres upon differences in curvature, and differences in valvular contour, associated with inequalities in valvular length. The least curved spores are still strongly arcuate, Fig. 119 representing about the extreme of straightness observed. Some spores are somewhat more sharply curved than the one shown in Fig. 118. The extreme of valvular disparity is about that seen in Fig. 120. Discussion. Although occurring in the urinary bladder, this species is rather like C. laxa and C. hama found in the gall bladder of the same host. The spores of C. laxa are somewhat thicker, have a little less valvular taper, and are more flattened horizontally. Differences in linear dimensions are just about at the upper limit of differences seen in samples of the other species from different hosts, or somewhat below it. The spores differ from C. laxa by having a greater breadth and a lesser thickness. The breadth: thickness ratios in C. renalis differ from those of C. laxa to a somewhat greater extent than is normally seen in samples of the same species, with a maximal value of 3.54 in the upper quartile of C. laxa and 3.92 in the same quartile for C. renalis. Differences in angular measurements are similar, in that the sample of C. renalis does not differ more than any single sample might from a species with mean values like those seen in C. laxa. The maximal difference in means is in the concave angle, which is 147° in C. laxa and 128° in C. renalis. The relationship of the angular means, however, is a little different in the two species, so that

the difference of 7° in taper index, and 15° in valvular index are essentially significant, confirming the greater length and taper of the valvular axes in C. renalis. The difference in habitat is also, in all probability, significant. The highly specialized environment of the gall bladder is not likely to be occupied by the same species as inhabit the urinary system. Differences in measurements between C. hama and C. renalis are probably significant, even on a merely numerical basis. The difference in breadth of the samples (over 4μ) is somewhat more than the difference usually seen in samples of a species from different hosts, representing about 17% of the mean breadth of C. hama, and the difference in taper ratios (28° and 10°) also appears to be significant. Of the remaining Ceratomyxa in groups II-B-1 and II-C-1, in which this species falls, those in the general size range of this species do not inhabit the urinary system. This species is rather unusual as a urinary parasite, as most Ceratomyxa with a urinary habitat have shorter, more oval spores. It is concluded that this is a hitherto undescribed species, and is accordingly given the name C. renalis n. sp. Ceratomyxa moenei n. sp. (Text-Fig. 9, Figs. 121–125) Habitat. This species occurs in the gall bladder of Polyprionum moene Phillips. It has been seen in but a single fish, taken by a trawler working out of Wellington in August. The infection was extremely heavy, masses of spores having formed to plug the bile duct. The gall bladder was somewhat swollen, and appeared red and irritated. Trophic Stages. No trophozoites were present. Spore Dimensions. Breadth, 25.3–35.5μ (30.2μ); height, 4.5–7.1μ (5.7μ); thickness, 5.1–6.8μ (5.9μ); valvular axes, 14 1–18 6μ (16.1μ); and 11.3–17.8μ (14.7μ); capsules. 2.3–3.4μ (2.7μ). Anterior angle, 128–147° (135°), posterior angle, 166–184° (176°), tangential angle, 156–179° (167°). Curvature index, 49°, taper index, 9°, valvular index, 41° Breadth height ratio for quartiles of breadth range 4.51, 5.60, 5.60 and 5.83. Spore Morphology. The slender spore is slightly curved, and terminates in rounded valvular tips. In sutural view the anterior margin is gently convex, often with a slight depression at the suture. In about 10% of the spores the lateral half of the valves are slightly flattened, and the anterior margin is somewhat depressed in this region. The posterior margin is nearly flat along each valve, and forms a very obtuse angle at the suture. The valves taper moderately to tips which are not very narrowly rounded for such a slender spore. In sutural view (Fig. 122), the spores are straight. The polar capsules are large in relation to spore height. The filament is arranged in four or five coils. They are not strongly convergent on the suture, and the capsules are often surrounded by remnants of the capsulogenous cells (Fig. 125). Capsulogenous nuclei are persistent. The sporoplasm is usually placed near the centre of the spores. It is quite transparent, and contains two vesicular nuclei. A few granular inclusions are sometimes visible. These are more prominent after staining with methyl green, and are missing in haematoxylin preparations. Spore Variability. Variability in this species is manifested primarily in differences in valvular taper, and the amount of flattening in the lateral portions of the valve. Spore curvature is relatively constant, and the ranges of the angular measurements are quite small. Sixty per cent of the spore sample had anterior, posterior and tangential angles in the ranges 131–139°, 173–182°, and 161–173°, respectively. Discussion. No Ceratomyxa have been previously reported from members of the Epinephelidae. The only Ceratomyxa recorded from very closely related fish is C. arcuata, which is quite unlike this species. The spore form places this species in group II-C-1. The only species of about the same size in this group is C. urophycis (Fantham, Porter and Richardson, 1940), which has one spore valve more curved than the other, a slightly curved sutural line, and small polar capsules. It appears, therefore, that this is a new form and it has been designated C. moenei n. sp.

Text-fig 9.—Fig. 115—Stained spore of C. constricta S.A., H. (f). Figs. 116–117—C. torquata n. sp. from Caulopsetta scapha. Fig. 116—Capsular view of spore, showing oily inclusions stained with neutral red. The valve at the left curves downward, while the valve on the right curves horizontally. F., N. R. (f). Fig. 117—Stained sport. S. A., H. (f). Figs. 118–120—C. renalis n. sp. from Caulopsetta scapha. Fig. 118—Capsular view of fresh spore. F., U. (f). Figs. 119–120—Outlines of less curved and less curved and less slender spores. F. U. (c). Figs. 121–125—C. moenei n. sp. from Polyprionum moene Figs. 121–122—Sutural and capsular view of spores. F., M. G. (e). Figs. 123–124.—Outlines of fresh spores. F. U. (e). Fig. 125—Stained spore. S. A. H. (f). Figs. 126–127—C. uncinata n. sp. Fig. 126—Spore from Caulopsetta scapha. F. U. (d). Fig. 127—Spore from Pelotretis flavilatus. F. U. (f).

Ceratomyxa uncinata n. sp. (Text-Fig. 9, Figs. 126–7). Habitat. This species occurs in the gall bladder of Pelotretis flavilatus Waite, and Caulopsetta scapha (Bloch and Schn.). It is apparently common in the former host, having been seen in the only two examined from the Wellington area. It is not common in Caulopsetta, but was found in a single host taken in the Napier area. Infected fishes have been seen in winter and summer. Trophic Stages. These have not been identified. Spore Dimensions. Based on samples from two Pelotretis Breadth, 27.8–36.0μ (32.2μ); thickness, 10.9–14.1μ (11.7μ); valvular axes, 18.5–25.1μ (22.1μ); and 16.7–21.8μ (19.7μ); capsules, 3.3–4.4μ (4.0μ); and 3.3–4.0μ (3.6μ). Convex angle 58–83° (72°), concave angle, 100–135° (120°); tangential angle, 86–124° (108°). Bending index, 168°, taper index, 12°, valvular index, 48° Breadth thickness ratios for quartiles of breadth range, 2.35, 2.36, 2.75 and 3.00. A small sample from Caulopsetta. Breadth, 26.5–35.2μ (30.8μ); thickness, 11.5–14.6μ (12.6μ); valvular axes, 17.2–27.7μ (24.5μ); and 17.2–22.5μ (20.4μ); capsules 2.9–3.9μ (3.6μ); Convex angle, 50–70° (59°), concave angle, 84–125° (98°), tangential angle, 75–116° (88°) Bending index, 203°, taper index, 10°; valvular index, 39° Breadth thickness ratios for quartiles of breadth range, 2.13, 2.33, 2.62 and 2.36. Spore Morphology. The spore is sharply bent, shaped like an inverted U, and orients itself so that it is seen in capsular view. None have been seen in sutural view. The convex margin is smoothly curved, unbroken at the suture. In some spores it is slightly inflected near the valvular tips, where the valves are slightly flattened. The opposite margin is concave, or flattened along each valve, which meet at an angle at the suture (Fig. 126–7) As the two margins are not parallel, the valves taper to moderately narrow tips. Although the valves are sometimes somewhat unequal in length, they are similar in shape. They meet in a rather broad, prominent, straight suture. The capsules, seen from above, are nearly spherical, and lie near the suture. The capsular foramina are opposite, and there seems to be little convergence of the capsular axes on the suture. The filament is easily seen, but as it is seen in end view the number of coils cannot be counted. One filament is a right-handed spiral, while the other is a left-handed spiral. Capsulogenous nuclei are sometimes persistent, but the capsulogenous cells disappear entirely in mature spores. The rather large sporoplasm is either symmetrical in position or asymmetrical. In the material from Pelotretix, the sporoplasm tended to be smaller, more rounded, and more asymmetrical in position. The protoplasm is finely granular, and quite transparent. It contains a few small, refractive bodies, about 0.5μ in diameter. Spore Variability. The spore shape is relatively constant, although the outlines are affected by somewhat differing valvular lengths, and somewhat different spore curvatures. On the whole, the effect given by a spore population in a host is one of relative uniformity. Spore populations seen in different hosts of the same species are relatively constant insofar as observations go. In the two Pelotretis, the means of the linear dimensions differed by 2.8μ in breadth, 1.1μ in thickness, 1.9 and 0.7μ in valvular axes, and 0.1μ in capsular diameter. The convex angles differed by 8°, the concave angles by 2° and the tangential angles by 3°. Because of differences in valvular length, the breadth height ratios were somewhat more variable in this than in most species, the maximal ratio for the upper quartile differing by 0.41. The spores seen in Caulopsetta differed from the mean between the Pelotretes hosts by 1.4μ in breadth, 0.9μ in thickness, 2.4 and 0.7μ in valvular axes, and 0.4μ in capsular diameter. The spores from Caulopsetta were somewhat more bent than those in Pelotretis, the convex, concave and tangential angles differing by 13°, 22° and 20°. This difference is about the same as the maximal angular differences seen in samples from hosts of the same species in the case of C. hokarari, and C. polymorpha. In the Pelotretis material, the polar capsules were often somewhat unequal. This was rarely seen in material from Caulopsetta. It seemed unlikely that these two strains were not conspecific, despite the small differences seen in the samples. Discussion. Of the various species of Ceratomyxa obtained from New Zealand flat-fishes, this species is most like C. laxa, but there was little evidence of transitions among spore populations of the two species suggesting that the two could be placed together. The spores of C. uncinata have a mean which is somewhat above the upper limit of the size range of spores of C. laxa, and there is a consistent difference in spore thickness. Until there is some definite evidence of transitional material between them, they must be considered as distinct strains, from the morphological sense, and presumably in the biological sense.

Of the other species of Ceratomyxa in group II-B-1, the most similar is C. toitae (Fujita, 1923), which also occurs in flat-fish hosts. Although it is bent like C. uncinata, however, the valves are unequally curved, and the capsules very small (2μ) and strongly convergent. It seems, therefore, that this form is previously underscribed, and it is designated as C. uncinata n. sp. Ceratomyxa polymorpha n. sp. (Text-Fig. 10, Figs. 128–137) Habitat. This species is a common parasite of the gall bladder of Physiculus bachus (Bloch and Schn.). It occurs in about 80% of the red cod taken by trawlers working out of Wellington, throughout the year. Even where the infection is heavy, there are no obvious signs of abnormality in the host organ, although the bile becomes yellow and cloudy. Trophic Stages. The trophozoites include very small, hyaline forms, about 5μ in diameter or a little less, apparently arising as a result of budding. The larger trophozoites reach a maximal size of 40μ, and are rounded or amoeboid in shape. They have narrow, elongated, blunt lobopodia. None have ever been seen in rapid movement. They often mass together as they are attached to the walls of the gall bladder, but do not fuse together. The ectoplasm is sometimes quite distinct, especially in the pseudopodia. It is clear, colourless, and highly transparent. The endoplasm contains a number of granules, rather small and appearing dark in dark contrast phase. These are mixed with another type of granule of about the same size which is bright in dark contrast phase. Both granules are somewhat less than 0.4μ in diameter. In dark field illumination, the whole endoplasm appears as bright, granular material. A number of small spherules, less than 1.0μ in diameter are also seen. These appear dark in dark contrast phase. They are not stained with Janus green B. Spore formation may be monosporous or disporous, with the latter more common. The sporoblasts begin as nearly spherical bodies, and as they enlarge, the capsulogenous cells become quite vacuolated, with meshy protoplasm (Fig. 129). In the young sporoblasts the suture is usually somewhat sinuous, straightening as the sporoblast reaches the diameter of the mature spore. In most cases the capsulogenous cells disappear early (Fig. 130), but occasionally, mature spores will still contain remnants of the capsulogenous cells (Fig. 131). The valvular nucler are stationed at the valvular extremities. They, like the capsulogenous cells, may disappear rather early, or may still be visible in otherwise mature spores (Fig. 132). Spore Dimensions. Based on samples from three hosts Breadth, 23.0–44.5μ (33.7μ); height, 11.1–16.4μ (14.6μ); thickness, 11.1–18.0μ (14.3μ); valvular axes, 12.0–25.5μ (18.9μ); and 9.0–20.5μ (15.4μ); capsules, 4.4–5.5μ (4.9μ); by 4.0–4.5μ (4.1μ); and 1.6–4.0μ (3.3μ); by 1.6–4.0μ (2.9μ) Anterior angle, 90–142° (114°); posterior angle, 163–244° (203°), tangential angle, 117–180° (158°) Curvature index, 43°, taper index, 45°, valvular index, 89° Convex angle, 93–124° (105°), concave angle, 150–231° (192°), tangential angle, 104–180° (156°). Bending index, 63°; taper index, 36°; valvular index, 87°. Breadth height ratios for quartiles of breadth range, 2.19, 2.25, 2.43 and 2.49. Spore Morphology. The bulky, somewhat curved spore has broad valves terminating in broadly rounded tips. Seen in sutural view the anterior margin varies from nearly straight to sharply convex. The posterior margin very nearly parallels the anterior margin in most spores, although in some the posterior margin is convex and the anterior margin nearly straight (Fig. 133). Some spores are bent rather than curved .In these the posterior margin is generally straight (Fig. 134). In many spores, the anterior and posterior margins are interrupted by a constriction at the suture. The valves are often unequal in axial length, and ill-matched in shape, giving one the impression of valves assembled separately and put together at random. They meet along a distinct, somewhat elevated, rather broad suture which is usually nearly straight, but is sometimes somewhat oblique. The oval to sligtly pyriform capsules are often quite unequal in size. The filament is distinctly visible, and is arranged in from four to seven tight coils. In some of the smallest capsules no filament can be seen. They are difficult to extrude in unkinked condition. It appears that the maximal length is about 45μ to 50μ, although many seem distinctly shorter. The filament from the smaller capsule is shorter, and it seems that the length of the filament is roughly proportional to the capsular size. The capsulogenous nucler differ considerably in their tendency to persist. The rather large, binucleate sporoplasm is variable in position. When it is asymmetrically arranged, the bulk of the sporoplasm is found in the larger valve. It contains a number of small, irregular inclusions, about 0.3–0.5μ in diameter, and larger ones averaging about 1.5μ in diameter. The larger inclusions often tend to adhere to the spore membrane, and occur in the spore cavity as well as the sporoplasm.

Text-fig 10.—Figs. 128–137 —C. polymorpha from Physiculus bachus Fig. 128—Small trophozoitc S. A., H. (f) Fig. 129–130— Younger and older developing spores. S. A. H. (e). Fig. 131—Capsular view of spore. P. U. (e). Fig. 132—Sutural view of spore. F. M. G. (d). Fig. 133—Sutural view of spore. F. U. (e). Fig. 134—Sutural view of curved spore. F. U. (e). Fig. 135—Staired spore. S. A. H. (f). Fig. 136— Spore with triad form F., U. (a). Fig. 137— Spore with tetread. F., U. (a). Figs. 138–144 —C hokarari from Genypterus blacodes Figs. 138–139—Sutural and capsular view of spore. F., U. (d). Fig. 140—Stained spore. S. A., H. (f). Figs. 141–144—Outlines of spores, showing range of curvature. F., U. (b).

Spore Variability. This species is quite variable, the valves being inconsistent in shape, curvature or bending, and often quite different in shape as well as size. Tri-valved and tetravalved spores are sometimes common. During the winter months populations of spores with about 60–75% of the spores in triad from occur. During the summer months, however, relatively few occur, and in one host taken in January, not a single triad spore was seen. Tertrad spores are never as common as triad spores, but sometimes make up to about 5% of the spores population. The tetrad spores are usually ill-formed, and where the valves are quite unequal in size, are almost amorphous. No spores with a differing number of valves and capsules were seen. Although the species is common, it was found in several hosts in which there were few mature spores, and eventually, full samples for measurement were taken from but three of the infected hosts, one taken in winter, one in spring, and one in summer. The sample means for breadth varied from 29.8μ to 38.3μ The mean of all samples was 33.7μ, so that the maximum deviation of a sample mean from the mean of all samples was 4.6μ, representing 13.6% of the mean breadth. On a similar basis, the maximal deviation of a sample mean from the mean of all samples for spore height was 6.9% of the mean height, and the maximal deviation of spore thickness 5.6% of the mean thickness. Sample means had maximal deviation from the mean of all samples of 14.3% of the longer valvular axis, and 8.6% of the shorter valvular axis. It is evident that a significant difference in dimensions of part of the spore did occur in these different hosts. Using small sample methods, it was easy to demonstrate differences in the samples with a p. value of 0.01 or less. In the case of angular measurements, the anterior angle was more stable than either of the others. The sample means agreed with a total difference of but 3° from smallest to largest for the anterior angle. The maximal deviation of a sample mean from the mean of all samples was 15° for the posterior angle, and 10° for the tangential angle. There is considerable compensation in the relationship of the various angular dimensions. If the total range of the posterior angle is divided into quartiles, the mean anterior and tangential angles show a continual change between the quartiles. The results are summarized in Table II. There was no evidence in the three samples, nor in the smaller samples from other hosts, to support the idea of a changing size of the spore with season. The only thing that can be said with some certainty is that the frequency of spores with triad and tetrad form was highest in the samples taken during the winter months. Discussion. Although the family Gadidae, to which Physiculus bacus belongs, is one of the most receptive to Myxosporidia, there are not a large number of reports of Ceratomyxa from among its numbers. C. arcuata has been reported from Gadus merlangus (Dunkerly, 1921), and C. acadiensis and C. urophycis from Urophycis chuss. (Ellis, 1930; Fantham, Porter and Richardson, 1940). Dunkerly (1921) also reports an unidentified Ceratomyxa from Molva molva. None of these species bears a close resemblance to C. polymorpha. This species falls into group I-A-1 or II-A-1, depending on the spore curvature. Of the species in group I-A-1, none approach the size of C. polymorpha. Of those in group II-A-1, none with a breadth comparable to that of C. polymorpha have a spore height which is nearly as great. It is evident, therefore, that it is a hitherto undescribed form, and the name C. polymorpha n. sp. is proposed for it. Table II —The Mean Anterior and Posterior Angles of C polymorpha Spores from Different Quartiles of the Range for the Posterior Angle Posterior Angle Anterior Angle Tangential Angle 159–180° 93° 131° 181–201° 119° 142° 202–222° 120° 172° 222–243° 127° 180° Ceratomyxa hokarari n. sp. (Text-Fig. 10, Figs. 139–145) Habitat. This species occurs in the gall bladder of Genypterus blacodes (Bloch and Schn.). It is a common form, having been found in 7 of 10 fishes examined during various seasons of the year. Although some of the infections were quite heavy,

there was no evidence of damage to the host organ. All of the infected fishes were taken by trawlers working out of Wellington. Trophic Stages. The trophozoites are apparently rather delicate, for they are usually missing entirely, and have never been seen in a motile condition. They are irregular in shape, with long, slender, narrowly rounded pseudopodia, usually located near the ends of a somewhat elongated body. The transparent endoplasm contains a variable number of refractive spherules, about 1. 0μ in diameter. The ectoplasm is not easily distinguished, but the spherules do not extend into the pseudopodia. Although an occasional monosporous trophozoite is seen, the majority are disporous, with the maturing spores carried in tandem. Spore Dimensions. Based on samples from six hosts Breadth, 24.2–48 4μ (35.6μ); thickness, 11.0–13.2μ (11.9μ); valvular axes, 16.5–35.2μ (27.1μ); and 15.4–33.0μ (23.8μ), capsules, 2.2–4.4μ (3.9μ); Convex angle, 32–106° (70°); concave angle, 35–155° (106°); horizontal tangential angle, 25–144° (94°); Bending index, 184°, taper index, 12°, valvular index, 36°, thickness ratios for quartiles of breadth range, 2.33, 2.69 2.98 and 3.69. Spore Morphology. The sharply bent spores have sub-equal valves, tapering gently to rather narrowly rounded tips. They usually orient themselves so that they are seen in capsular view, and relatively few are seen in sutural view. In sutural view the spore has a relatively straight anterior and posterior margin (Fig. 140), not quite parallel, and the suture is straight In capsular view the spore is crescentic, with rather long, slender, gracefully curved valves. The two valves are usually almost equal in size, but occasional spores are seen with considerable disparity in valvular length (Fig. 144). They meet in a slightly elevated, distinct suture, which in some stained spores appears as a double line. The suture is almost circular. The capsules are oval in sutural view, the length slightly exceeding the breadth. They terminate in short, slender necks. About four or five coils can be seen. In capsular view the filament is distinctly visible in the circular capsules. It does not appear to converage much on the suture. The capsulongenous nucler may or may not be persistent. No other remnants of the capsulogenous cells are seen in mature spores. The capsules may be centred on the anterior face of the spore, or be placed toward the convex or concave border. The sporoplasm is finely granular, with a few small, refractive inclusions. The two nucler are relatively small, and are usually placed close together. In some spores they show a structure almost identical to that shown for C. blennius by Noble (1942) Spore Variability. The species shows a considerable range of spore shapes, the principal variation being in relative length of the valvular axes, and in the amount of curvature. The approximate range of variation is shown in Figs. 142–145. The spores in any one host show a somewhat smaller range than is indicated by the diagrams. A total of six samples were compared, to obtain a picture of the amount of variation that might be found in samples from different host animals. The maximal difference of a sample mean from the mean of all samples was 15% of the mean in the case of breadth. The valvular axes exhibited differences in the same order of magnitude, the maximal deviation of a sample being 12.5% for longer valve, and 12.6% for the shorter valve. The maximal divergence of the spore thickness from the mean of all samples was 7% of the species mean. Like the linear dimensions, the angular measurements were subject to considerable variation. The greatest deviation of a sample mean from the species mean was 16° for the convex angle, 23° for the concave angle, and 22° for the tangential angle. The differences between the means of the various samples were often highly significant from the statistical point of view. In a number of instances the differences proved to be significant well beyond the. 001 level. It is evident that in the complex of factors correlated with life in different host animals at different seasons, conditions which result in real differences in size and contour of the spore are encountered. These are often sufficient to produce statistically significant differences in linear or angular dimensions. The statistical methods are, evidently, too delicate an instrument to be used for species recognition. This might be anticipated, as statistical methods have their routine use in the genetic studies of populations within the species group. Although the individual angles measured vary considerably in different samples, the relationships between the angles are relatively stable. The bending index is quite variable having been as high as 217° in one sample, and as low as 143° in another sample. However, the valves which are more or less curved retain a similar shape, and the valvular index varied only from 34° to 43°, while the taper index varied only from 9° to 17°. It would appear, therefore, that in comparing spore populations, considerable weight should be given to the relationship of the angles, perhaps more than to the angles themselves. A much more extensive series of infected hosts would require examination before firm conclusions could be drawn concerning the effects of season upon spore shape and size. There was no evidence of a relationship of the season to spore curvature, or other angular measurements. There was, however, a definite relationship between spore breadth and season, insofar

as the samples studied are concerned. All of the samples taken during the winter season had means below the mean of all samples, while all of the samples taken during the warmer months had means well above the mean of all samples. While this may have been forturtous, the chances of this occurring in all of six hosts is not high. The evidence would certainly suggest that it would be worth while looking for a correlation between spore breadth and such seasonal factors as water temperature, nourishment of the host, etc. Discussion. It is interesting to note that the only other Ceratomyxa to have been reported from a member of the Ophidiidae is C. arcuata, with spores sharply curved rather than bent, with smaller capsules and a more slender shape. C. inversa, also found in Genypterus, has straight spores which are considerably larger. C. hokarari falls into group II-C-1. Of the species in this group, only C. urophycis is very similar. This species, however, has one valve somewhat more curved, a slightly curved sutural line, and small, pyriform polar capsules. As the form cannot be considered identical with any of the previously described species, the name C. hokarari is proposed for it. Ceratomyxa angusta n. sp. (Text-Fig. 11, Figs. 145–147) Habitat. This species occurs in the gall bladder of Hypoplectrodes semicinctus (Cuv. & Val.) and Helicolenus percoides (Richardson). It is apparently not a very common form, having occurred in but a single example of each host Both infected hosts were taken by trawlers working out of Wellington. In neither host was there evidence of damage to the host organ. Trophic Stages. Trophozoites have not been seen. Spore Dimensions. From Hypoplectrodes, based on spores measured in water suspension, following fixation in sublimate acetic and storage in alcohol Breadth, 37.0–47.4μ (42.1μ); height, 5.1–6.2μ (5.5μ); valvular axes, 19.7–26.5μ (23.5μ); and 13.5–22.5μ (18.2μ); capsules, 2.3–3.4μ (2.7μ). Anterior angle, 143–160° (153°), posterior angle, 176–193° (183°), tangential angle, 169–180° (179°) Curvature index, 24°, taper index, 4°, valvular index, 30° Breadth height ratios for quartiles of breadth range, 7.5, 7.5, 7.8 and 7.6. From Helicolenus. Breadth, 34.2–57.8μ (48.1μ); height, 5.4–6.5μ (5.7μ); thickness, 5.4–5.9μ (5.7μ); valvular axes, 21.6–30.2μ (25.5μ); and 17.3–28.0μ (23.0μ), capsules, 3.2–3.8μ (3.3μ); by 2.7–3.2μ (3.1μ). Anterior angle, 143–160° (153°), posterior angle, 177–186° (181°), tangential angle, 172–179° (175°) Curvature index, 26°, taper index, 6°, valvular index, 28° Breadth height ratios for quartiles of breadth range, 7.0, 8.5, 8.5 and 8.65. Only a small sample were measured for spore thickness. Spore Morphology. The slender, straight spore terminates in narrow, rounded valvular tips In sutural view the anterior margin of each valve is nearly straight, sloping upwards to meet in an obtuse angle at the suture. The posterior margin is nearly straight, but is slightly curved in some spores (Fig. 147). The valves are elongated, and may be equal or quite unequal. They are approximately the same in general shape, even when unequal. They meet in a rather narrow, slightly elevated suture, which is often set in a slight depression. Although the gradient of the valvular taper is slight, the valvular tips are narrowly rounded. The thin spore membrane is deformed easily by pressure and the forces acting during dehydration. It is not uncommon to see a spore with the valves collapsed shut over the whole length, except for the central region containing the sporoplasm In capsular view (Fig. 146) the spore is straight and the amount of valvular taper about the same as that seen in sutural view. The polar capsules are nearly spherical to broadly oval, with very short, narrow necks. They converge somewhat on the suture, and open on the anterior margin In capsular view they are opposite, or nearly so. The filament is rather indistinct in fresh spores, and appears to be very delicate. It forms from five to seven coils Capsulogenous nuclei are persistent in the majority of spores, and in many there are some remnan's of the capsulogenous cells. The sporoplasm is restricted to the central region of the spore, lying beneath the capsules and extending upwards around and between them Occasionally there are rounded masses of sporoplasm which have become separated from the main mass (Fig. 146) and lie in the more lateral portions of the valves. The sporoplasm is finely granular, with few inclusions. It contains two vesicular nuclei. Spore Variability The most variable feature of the spore is the axial length of the valves, which may be equal or unequal, and vary considerably in actual length. The difference in the spore size of the forms found in Hypoplectrodes and Helicolenus may be attributed, in large part, to the fact that the former were fixed when measured. The approximately 20% loss in spore breadth is not far out of line with the normal shrinkage to be expected during

Text-fig 11 — Figs. 145–147 — C. angusta n sp Fig. 145–6 — Spores from Hypoplectrodes semicinctus F, U (d) Fig. 147 — Spore from Helicolenus percoides F, U (f) Fig. 148 — C. aggregata from Dactylopagrus macropterus Stained spore S. A., H (e) Fig. 149 — C. nitida n sp from Peltorham phus novae-zelandiae Fresh spore F, U (e) Fig. 150 — C. inversa n sp from Genypterus blacodes Fresh spore I. U (f) Figs. 151–154 — C. elongata n sp from Lepidopus caudatus Fig. 151 — Sutural view of spore F, M, G (e) Fig. 152 — Outline of spore in capsular view F, U (a) Fig. 153 — Central portion of spore in capsular view S. A., H (e) Fig. 134 — Central portion of spore in sutural view S. A., H (e)

fixation with sublimate acetic. It seems probable that the spores shrink somewhat more in breadth than in sutural diameter during fixation. It is evident that the angular measurements are less affected by fixation than the linear measurements. The close agreement of the angular measurements in the two samples is, no doubt, related to the rather constant spore shape, and to the fact that the ranges observed in the angular measurements were relatively short. It appears that there is somewhat more variability in the angular measurements of the fixed material, but the mean values remain essentially unchanged. Discussion. The hosts belong to the families Serranidae and Scorpaenidae. Except for species reported in this paper, no Ceratomyxa have been previously reported from the Serranidae. Fujita (1923) reported C. yoichiensis and Jameson (1929) C. starksi from Scorpaenids. The spores formed in both of these species are much smaller than those produced in C. angusta The spore form is such that the species belongs in group I-C-1. The only species in this group which resembles this form is Fujita's (1923) C. japonica, which has a spore with a much greater height (11–13μ). Other elongated species have spores which are more curved, more sharply tapered, or are quite different in dimensions As this appears to be previously undescribed, it has been designated C. angusta n. sp. Ceratomyxa aggregata Davis (Text-Fig. 11, Fig. 148) Habitat. This species occurs in the gall bladder of Dactylopagrus macropterus (Bloch and Schn.). It is not common, having been seen as a light infection, mixed with C. insolita, in a single host. The infected fish was taken in August, by a trawler working out of Wellington. There was no evidence of damage to the host organ. Trophic Stages The trophozoites have not been seen. Spore Dimensions Based on a small sample of 6 spores Breadth, 41 4–60 6μ (51 5μ), height, 5 9–7 8μ (7 1μ), valvular axes, 22 0–32 2μ (26 8μ) and 21 6–29 4μ (25 5μ), capsules 2 4μ by 2 0μ Angular measurements were not taken. as the slender valves may be bent or twisted in any direction. Spore Morphology The central region of the spore is enlarged, and the lateral portions of the valves narrow, often flattened, and often bent or twisted In sutural view the central region of the spore has a convex anterior and posterior margin, the curvature being broken, in most cases, as the more lateral portions of the valves are reached. These empty parts of the valves may be straight, curved forward or back, or bent to one side. They are often bent in different directions, and are sometimes collapsed Despite differences in contour of the empty portions of the valves, they are approximately equal in axial length. There is no evidence of a septum at the base of the empty portions of the valves. The valves meet in a fairly prominent, straight suture. The capsules are set near the anterior margin, and near the suture, on which they converge Most of the spores appear a little immature, with prominent capsulogenous cells. It seems probable that the capsules might increase in size somewhat, as they mature Valvular nuclei, however, cannot be seen. The rather large sporoplasm fills the central region of the spore. It contains a few refractive granules, and two vesicular nuclei. Discussion. The original description of C. aggregata (Davis, 1917) is based on material from Leiostomus xanthurus and Micropogon undulatus, members of the Sciaenidae One of the more characteristic traits of C. aggregata is the aggregative tendencies of the trophozoites, and a final identification of this form must await the comparison of the trophic stages from Dactylopagrus with those seen by Davis. There are some differences in the spores described by Davis and those seen in the New Zealand material. The most outstanding difference, perhaps, in the larger size of the capsules in Davis's material, which may be explained on the basis of immaturity of the spores seen in Dactylopagrus There also appears to be somewhat more sharp separation of the more lateral parts of the spore valves in the New Zealand material However, until more material can be examined, the differences appear to be insufficient to permit separation of the two forms. They have spores of about the same size, and general shape, and in both strains the spores tend to collapse laterally. It seems preferable, at the present, to consider the two forms conspecific.

Ceratomyxa nitida n sp. (Text-Fig. 11, Fig. 149) Habitat This species was seen in a single Peltorhamphus novae-zeelandiae Gunther. The host fish was taken by a trawler working out of Napier in September. There was no evidence of damage to the host organ. The infection was mixed with Leptotheca pinguis Trophic Stages These are unknown. Spore Dimensions Based on a sample of 13 spores Breadth, 52.9–63.6μ (58.3μ); height, 10.8–13.7μ (12 1μ); valvular axes, 31.7–34.8μ (32.9μ) and 26.0–31.8μ (28.6μ); capsules, 5.4–6.4μ (5.9μ) by 3.9–4.9μ (4 4μ). Anterior angle, 115–131° (124°); posterior angle, 155–173° (165°); tangential angle, 144–162° (155°). Curvature index, 71°; taper index, 10°; valvular index, 41°. Mean breadth: height ratio, 4.81. Spore Morphology The spore is straight, or nearly straight, in capsular view, the capsules slightly rotated in position In sutural view the two bulky, moderately tapered valves curve slightly, and meet in a heavy, broad, elevated suture. The two valves are often somewhat unequal in length, but have essentially the same shape Despite the length of the valvular axes, the valvular tips are broadly rounded. The large, pyriform polar capsules have narrowed necks, which are relatively long and tapered. They converge sharply on the suture, and are somewhat rotated. The filament appears to form a double coil in the capsule. The coils are set in a diagonal direction when the spore is seen in sutural view. The filament is quite delicate, but from five to seven coils can be seen Capsulogenous nuclei are persistent, and in many spores vacuolated remnants of the capsulogenous cells can also be seen. The large sporoplasm is essentially central in position, but often extends farther along one valve than the other In a few spores it extends nearly to one valvular tip. It is rather coarsely granular, and contains a number of small, refractive granules, and a few refractive spherules 1μ in diameter. The sporoplasm contains two rather large, vesicular nuclei Small amounts of finely granular and rather transparent material and an occasional refractive spherules occur in the spore cavity outside of the sporoplasm. Discussion Although the infection was relatively light, and only 13 spores were properly oriented for measurement, the spores seen were relatively stable in form, and it seemed possible to give a reasonable characterization of their form. The species falls in group II-B-1 Of the species in this group, it most resembles C. protopsettae (Fujita, 1923) and C. drepanopsettae Awerinzew (1908) in spore breadth. Both of these species occur in flat-fishes. In both of these species, however, the two valves are dissimilar in shape, with one valve more slender and more elongated. It is, on the whole, more similar to C. protopsettae than to C. drepanopsettae, but as Fujita states that the suture is indistinct in that species, it seems evident that the two cannot be the same As this species cannot be equated with previously described forms, it is given the name C. nitida n sp. Ceratomyxa inversa n sp (Text-Fig. 11, Fig. 150) Habitat This species occurs in the gall bladder of Genypterus blacodes (Bloch and Schn.) It has been found in but a single host, taken by a trawler working out of Wellington in March. The bile was so heavily infected that it was a milky yellow colour, and seemed less viscous than usual. There was no evidence of any abnormality in the host organ, however. Trophic Stages Trophozoites are extremely varied in size and shape Small forms about 8μ in diameter are common, as are large, sporulating forms of about 150μ in length. The motile organisms move very slowly. They are basically pyriform in shape, with the anterior end rounded and tapering to the region of maximal breadth about one-third of the way back from the front end. The remainder of the body narrows, forming a large and bulky posterior process In small forms, the posterior process is relatively short and narrow, and terminates in a sharp point, while in larger forms the process is wider, longer, and more clavate in shape Sporoblasts are carried in the anterior, expanded part of the body Small slender, tapering, pointed pseudopodia occur at the anterior end of the motile forms Floating trophozoites and those attached to the wall of the gall bladder assume a clavate form, with an enlarged region containing the sporoblasts located somewhere along the body. It is not uncommon to see the attached forms with the enlarged region in contact with the wall of the gall bladder, looking rather like the motile forms, but not indulging in movement Floating

or attached forms develop pseudopodia at various points on the body surface. These pseudopodia are larger than those seen in the motile forms, and often branch. The trophozoites contain few inclusions, and these are never prominent. The endoplasm is finely granular, containing a number of very small spheres, which are either black or light in dark contrast phase, but never bright-rimmed. They are scarcely visible in bright field illumination Small, spherical inclusions, averaging 1μ in diameter, appear as dark spheres in dark contrast phase, and are not seen in dark field. They persist in fixed and stained slides where they are acidophilic. Despite the large size of the trophozoites, none were seen with more than two sporoblasts. The young spores become prominent early in development As they increase in size, they become more elongate, and tend to lie side by side, without crossing Unlike some of the species with long, slender valves, this one does not tend to have the lateral extremities of the spore valves curled about the central region of the spore during development. Spore Dimensions Breadth, 51.1–73.7μ (61 5μ); height, 6.9–9.3μ (8.5μ); thickness, 7.7–10.0μ (9.0μ); valvular axes, 25.5–38.7μ (33.5μ) and 20.0–37.2μ (30.1μ); capsules, 3.1–4.6μ (3.4μ) by 2.3–3.4μ (3.1μ). Anterior angle, 111–188° (152°), posterior angle, 141–215° (180°), tangential angle, 137–180° (167°) Curvature index, 28°; taper index, 13°; valvular index, 28°; Convex angle, 112–153° (132°); concave angle, 134–173° (164°); tangential angle, 130–180° (158°). Bending index, 64°; taper index, 6°; valvular index, 32° Breadth height ratios for quartiles of breadth range, 6 33, 6 40, 7 90 and 9.05. Spore Morphology The slender spores are somewhat curved and bent, and normally orient themselves to float upside down, with the capsular surface down As a result, relatively few spores are seen in either capsular or sutural views, most being seen in basal view. Seen in basal view the spores are straight to somewhat bent, with the straight spores tending to exhibit some vertical curvature Spores are rarely both bent and curved. The two valves, often somewhat unequal in axial length, are of similar shape, and have a moderate taper. They meet in a rather narrow, elevated suture, which is sometimes set in a slight depression Few spores are seen in sutural view, but when they appear in this orientation, the anterior margin tends to be more deeply marked by the sutural constriction than the posterior margin (Fig. 150) In general, as the angular measurements indicate. spores are somewhat more bent than curved. The polar capsules are spherical in basal view and capsular view. They are opposite or nearly so, and usually lie near the axis of the spore Capsulogenous nuclei are usually persistent. The capsulogenous cells are not, however. The filament is distinctly visible, and in spores seen in sutural view, appears to make from four to six coils. The capsules converge on the suture. The sporoplasm is moderately large, extending from ⅓ to ½ the length of one valve, and usually in contact with the bases of the capsules. It is sometimes centred in the spore, and sometimes quite asymmetrically placed. It is rather transparent, and contains a few bright, refractive spherules as well as two vesicular nuclei. Discussion The only other Ceratomyxa reported from the Ophidiidae are C. arcuata and C. hokarari. Both species have been seen in New Zealand material, and neither can be mistaken for C. inversa, both having strongly crescentic spores. It was thought, initially, that C. inversa might represent a straight strain of C. hokarari, but the capsules are somewhat different, and the valvular dimensions are quite different. This species falls into group II-B-1 or II-C-1, with the majority of spores falling into the latter group In the former group, only C. protopsettae and C. drepanosettae have similar spore breadth. Both of these species are parasites characteristic of flatfishes, and differ in spore form, with dissimilarly shaped valves, and somewhat different spore dimensions (Fujita, 1923; Awerinzew, 1908). The species in group II-C-1 include C. ramosa (Awerinzew, 1907) and C. microstomi in the same general size range C. ramosa has trophozoites with a few much ramified pseudopodia which anastomose as well as branch. The spore height is greater than in C. inversa. The spores of C. microstomi (Fujita, 1923) are less curved, and the sporoplasm is marked by coarse, oily spherules Since this form does not agree with any of the previously described species, it has been named C. inversa n sp. Ceratomyxa elongata n sp. (Text-Fig. 11, Figs. 151–154). Habitat This species occurs in the gall bladder of Lepidopus caudatus Infected hosts were taken in the winter and spring by trawlers working out of Wellington.

It is common, occuring in about half of the Lepidopus examined Although some of the infections were very heavy, there is no evidence of damage to the host organ. Trophic Stages. The trophozoites are about 5μ in diameter at their smallest, and about 45μ in diameter when at maximal size. They often gather together in large masses, attached to the walls of the gall bladder. The individual outlines of the trophozoites remain distinct Younger trophozoites are often pyriform in shape, with two or three slender, tapering pseudopodia None of these have exhibited motility Larger trophozoites tend to be more or less clavate to amoeboid in shape. The protoplasm is rather clear, containing a variable number of small, refractive inclusions about 1.0μ in diameter, which do not persist in permanent preparations Ectoplasm can be distinguished only in pseudopodia. The sporoblasts are easily seen in fresh trophozoites Originally quite small, they become oval bodies as differentiation continues At this time they look very much like mature spores of a Leptotheca-like species, surrounded by a clear area, somewhat resembling the membranous remnant of a sporoblast. The clear area, however, is made up of the curved and flattened lateral portions of the valves which are more or less wrapped about the central portion of the spore. The lateral portions of the valves apparently do not expand until the spore is liberated from the trophozoites Only disporous trophozoites have been seen. Spore Dimensions Fresh material from one host Breadth, 72.6–99.0μ (91.0μ); height, 7.7–11.0μ (9 6μ); thickness, 6.6–8.7μ (7.8μ); valvular axes, 36.6–53.9μ (48.4μ) and 36.3–47.3μ (39.6μ); capsular diameter, 2.2–3.3μ (3.0μ). Anterior angle, 137–166° (152°); posterior angle, 159–188° (177°); tangential angle, 155–180° (172°). Curvature index, 31°; taper index, 5°; valvular index, 25°. Breadth: height ratios for quartiles of breadth range, 7.35, 9.40, 10.42 and 9.65. A sample of fixed spores from another host, studied in water suspension. Breadth, 63–89μ (76.0μ); height, 6.5–8.5μ (7.3μ); thickness, 6.5–7.5μ (7.0μ); valvular axes, 32–50μ (40.2μ) and 30–40μ (34.3μ). Anterior angle, 125–180° (152°); posterior angle, 153–201° (175°); tangential angle, 146–180° (166°). Curvature index, 33°; taper index, 9°; valvular index, 23°. Spore Morphology The slender, slightly curved spore is quite large, and has a central, somewhat inflated region In sutural view the anterior margin is nearly straight for the greater part of the length of each valve, with a convexity at the central region adjacent to he suture. The posterior margin is flattened, or slightly concave, and in some instances shows evidence of a somewhat expanded central region also. The margins of the valves are not quite parallel, each tapering very gradually to quite narrowly rounded tips In capsular view (Fig. 152) the spore is straight or nearly so, with similar taper. The two valves meet in a straight, raised suture which is moderately conspicuous. The polar capsules are relatively small, and appear to remain permanently juvenile Capsulogenous cells were invariably present in all hosts, usually forming a dense protoplasmic covering around the capsules In some spores it is not possible to trace the line of separation between the two capsulogenous cells beneath the suture Even in these spores, however, the capsulogenous mass is bilobed. The filament tends to be indistinct, althought about four or five coils appear to be formed. The sporoplasm is a small mass, located centrally, and extending but a short distance along the valves at the most. It is closely associated with the capsules, often appearing to be in close contact with the capsulogenous cells. It contains a few refractive bodies, and two vesicular nuclei. The nuclei are usually placed close together, but are sometimes distant. Spore Variability The variability of the spores is primarily one centred in the length of the valvular axes. They may be somewhat longer or shorter, and the two valves may be equal or quite unequal. The spores vary in curvature from nearly straight to somewhat curved. They are never strongly arcuate, nor are they conspicuously bent. Spores from different hosts are essentially similar in size and shape. Only small samples from most hosts were measured. These fall well within the range of size indicated by the full sample quoted in the dimensions section. The fixed spores give evidence of a marked shrinkage, particularly in breadth. They have a more variable shape, angular measurements indicating this in their greater range, but the mean angular measurements of fixed and fresh spores are almost identical. Discussion C. obovalis (Fantham. 1930) is the only member of the genus reported from the Trichiuridae. It occurs in the same host species, Lepidopus caudatus Its Leptotheca-like spores are in no way similar to those seen in this species.

The spores fall in groups I-C-1 or II-C-1, depending on the amount of spore curvature This species falls in a space between C. mesospora (Davis, 1917) and C. tylosuri (Awerinzew, 1913) in group I-C-1, being significantly larger than the one and smaller than the other In II-C-1 it is larger and more slender than C. microstomi (Fujita, 1923), less bent than either C. osmeri (Fujita, 1923) or C. furcata (Fujita, 1923), and smaller than C. flagellifera As it cannot be identical with any of the previously described species, it has been designated as C. elongata n sp. Table III. —Summary of Means of Linear Dimensions of Ceratomyxa Spores from New Zealand Fishes Species Breadth Height Thickness Larger Valve Shorter Valve Capsule Breadth-Height Ratios tor Q1 Q2 Q3 Q4 minuta 11.8 5.8 5.5 6.4 5.7 2.7 1.67 1.99 2.24 2.37 inconstans (Scomber*)* Fixed, measured in water suspension. 10.2 4.4 4.5 5.4 4.9 1.7 2.00 2.34 2.50 2.58 (Usacaranx) 12.5 5.5 5.1 6.8 6.4 2.1 1.91 2.13 2.45 2.72 (Helicolenus) 12.0 5.4 — 6.4 5.9 1.9 1.93 2.07 2.36 2.66 (Trachurus) 11.8 5.5 4.8 6.5 6.1 1.7 1.94 2.12 2.24 2.19 faba 12.7 6.2 6.4 6.9 5.6 2.4 1.82 1.94 2.15 2.16 castigata 13.1 5.9 — 7.3 6.2 2.2 1.93 2.06 2.15 2.41 castigatoides* 14.7 6.7 5.8 7.9 6.9 2.0 1.64 2.07 2.10 2.20 declivis 14.4 5.9 5.6 8.4 7.3 2.4 2.65 2.27 2.51 2.57 intexua 15.4 4.4 3.9 8.6 7.4 1.8 2.62 3.03 3.37 3.92 recta 15.6 7.8 7.8 8.1 7.6 2.6 2.07 2.04 2.05 1.90 flexa* 15.9 5.6 — 10.1 8.6 2.6 2.17 2.49 2.76 3.24 insolita 16.2 11.5 — 23.8 21.5 11.5 — — — — gibba 17.0 6.9 — 10.2 8.3 2.8 2.11 2.59 2.42 2.35 gemmaphora* 19.2 7.0 6.4 10.8 9.4 2.1 2.37 2.65 2.89 2.82 lepallida 18.7 8.6 8.1 10.4 9.2 2.2 1.88 2.04 2.26 2.30 subtilis 21.5 3.9 4.3 11.9 10.4 1.8 3.94 5.37 5.20 6.57 hama 23.4 7.4 7.7 14.4 11.9 3.2 2.57 3.04 3.46 4.08 laxa 25.6 — 8.7 15.7 14.5 3.4 2.50 2.80 3.22 3.54 arcuata** Fixed measured in permanent slides. 25.6 5.7 5.7 18.8 15.0 1.9 3.34 3.75 5.40 5.33 constricta 26.9 8.1 — 14.7 13.7 3.5 2.88 3.12 3.68 3.54 torquata 26.9 8.1 9.2 16.7 14.4 3.8 2.54 2.97 3.41 3.76 renalis 27.7 8.0 8.9 17.4 16.0 3.1 2.75 3.20 3.50 3.92 moenei 30.2 5.7 5.9 16.1 14.7 2.7 4.51 5.60 5.60 5.83 uncinata (Pelotretis) 32.2 — 11.7 22.1 19.7 4.0 2.35 2.36 2.75 3.00 (Caulopsetta) 30.8 — 12.6 24.5 20.4 3.6 2.13 2.33 2.62 2.36 polymorpha 33.7 14.6 14.3 18.4 15.4 4.9 2.19 2.25 2.43 2.49 hokarari 35.6 — 11.9 27.1 23.8 3.9 2.33 2.68 2.98 3.69 angusta (Hypoplectrodes)* 42.1 5.5 — 23.5 18.2 2.7 7.50 7.50 7.80 7.50 (Helicolenus) 48.1 5.7 5.7 25.5 23.0 3.1 7.00 8.50 8.50 8.65 aggregata 51.5 7.1 — 26.8 25.5 2.4 — — — — nitida 58.3 12.1 — 32.9 28.6 5.9 — — — — inversa 61.5 8.5 9.0 33.5 30.1 3.4 6.33 6.40 7.90 9.05 elongata 91.0 9.6 7.8 48.4 39.6 3.0 7.35 9.40 10.42 9.95

Table IV —Summary of Means of Angular Measurements and Indices of Ceratomyxa Spores from New Zealand Fishes Species Ant Angle Deg Post Angle Deg Tang Angle Deg Curv Index Deg Taper Index Deg Valvular Index Deg minuta 109 199 178 52 21 90 inconstans (Scomber)** Fixed measured in water suspension. 125 228 178 7 50 103 (Usacaranx) 105 191 160 64 31 86 (Helicolenus) 109 192 160 59 32 83 (Trachurus) 98 181 162 81 19 83 faba 99 200 162 61 38 101 castigata 113 206 172 41 34 93 castigatoides 108 209 168 43 41 101 declivis 97 175 145 83 30 78 intexua 120 180 167 60 13 60 recta 120 230 180 10 50 90 flexa* 82 147 129 131 18 65 insolita 28 41 32 291 9 13 gibba 96 170 150 94 20 74 gemmaphora* 116 198 165 46 33 82 vepallida 104 194 158 62 44 90 subtilis 128 164 158 68 6 36 hama 101 160 136 99 24 59 arcuata** Fixed measured in permanent slides. 74 99 96 187 3 25 constricta 113 178 145 69 33 65 torquata 99 151 135 110 16 52 moenei 135 176 167 49 9 41 polymorpha 114 203 158 43 45 89 angusta (Hypoplectrodes)* 153 183 179 24 4 30 (Helicolenus) 153 181 175 26 6 28 nitida 124 165 155 71 10 41 inversa 152 180 167 28 13 28 elongata 152 177 172 31 5 25 Convex Concave Tang Bend Taper Valvular Angle Angle Angle Index Index Index hama 91 159 111 150 28 48 laxa 88 147 130 125 17 59 torquata 91 148 133 121 15 57 renalis 84 128 118 148 10 44 uncinata (Pelotretis) 72 120 108 168 12 48 (Caulopsetta) 59 98 88 203 10 39 hakarari 70 106 94 184 12 36 polymorpha 105 192 156 63 36 87 inversa 132 164 158 64 6 32 Unidentified Species of Ceratomyxa From time to time a small number of spores of trophozoites are encountered in a host, providing too little information to permit accurate characterization of the species While they cannot be described in detail, a brief mention of their known traits may be valuable in directing attention to them and, perhaps, helping to lead to their eventual recovery and proper description. The following notes are based upon sometimes brief observation and in all cases represents less than definitive characterization.

Ceratomyxa sp. from Thyrsites atun Only two spores of a very small, slender, bent species were seen in the gall bladder of a Thyrsites atun taken in the Wellington region in October. The spores were very small, about 10μ in breadth, and 3μ in thickness. The valves were sharply bent, and tapered to very sharply pointed tips. They were seen only in stained preparations One is shown in Fig. 155. Ceratomyxa sp. from Caulopsetta scapha The Ceratomyxa fauna of the gall bladder of Caulopsetta scapha is varied and bewildering. The variety of spore forms is remarkable, and as there are several rather constant types which are nearly of the same size, it is most difficult to sort them out into satisfying clusters. The system used here resulted in the isolation of several spore populations which were too small for proper characterization and yet seemed distinct from the species that have been described. One of these is a rather small, ellipsoid form, somewhat larger than C. faba and somewhat smaller than C. vepallida. It was found for the first time in association with C. faba, and while the data were kept separate from that species, it was thought that it might represent the uppermost portion of the C. faba population insofar as spore breadth was concerned However, it was found later in a host which lacked C. faba, so it is apparently distinct In both hosts it was present only in small numbers, other species forming the greater part of the Ceratomyxa population. Spore Dimensions. Based on 10 spores only. Breadth, 15.2–18.6μ (15.9μ); height, 6.2–7.4μ (6.5μ); thickness, 5.6μ and 7.9μ (2 spores only); valvular axes, 7.4–10.4μ (8.6μ) and 6.7–7.9μ (7.5μ); Capsules, 2.6–3.1μ (2.7μ). Anterior angle 103–134° (122°); posterior angle, 184–230° (206°); tangential angle, 161–180° (175°). Curvature index, 32°; taper index, 31°; valvular index, 84°; Mean breadth: height ratio, 2.45. Spore Morpholoy The ellipsoid spores have a convex or flattened posterior margin. They resemble C. faba, but are somewhat broader, and have larger capsules. They are somewhat less broad than C. vepallida The two subequal valves meet along a distinct, straight, slightly elevated suture. The capsules, placed near the suture, lie near the anterior margin In capsular view the spore is straight or nearly so, and the capsules are opposite. Ceratomyxa sp. from Scomber japonica Several spores of a moderate-sized, bent species of Ceratomyxa were seen in a Scomber japonica, taken by a trawler working in the Wellington area in July. The host was heavily infected with C. inconstans. The spore is sharply bent, with moderately tapering valves, terminating in blunt tips. The small sporoplasm is centred in the spore. The capsules open through the anterior margin, and are accompanied by capsulogenous cells and nuclei Valvular nuclei were faintly visible at the valvular extremities (Fig. 158) Only a single spore was carefully measured. It was 23 5μ in breadth, 8 8μ in thickness, with valvular axes of 17 1 and 13 7μ, and capsules 2 9μ in diameter. The convex angle was 70°, the concave angle, 113° and the tangential angle 103°. The spore is distinctly smaller than those seen in other somewhat similar species As the spores seen may have been somewhat immature, the dimensions may be subject to further change. Ceratomyxa sp. from Caesioperca lepidoptera Several spores of a rather small species with bent or curved spores, often with one valve curved in one direction and the other bent in the other, were found in the gall baldder of Caesioperca lepidoptera The dimensions of six spores averaged. breadth, 15.6μ, height, 7.1μ; thickness, 6.5μ, valvular axes, 9.3 and 8.4μ, capsules, 2.1μ The spore shown in Fig. 156 had the left valve curved downward, and the right valve curved forward, while the one shown in Fig. 157 had both valves curved in the same plane. The rather small capsules were still surrounded by capsulogenous cells in some, but not all spores. They converge on the suture

Text-fig 12.—Fig. 155—C sp? from Scomber japonica Spore in capsular view S. A., G (d) Figs. 156–157—C sp? from Pseudolabrus coccineus Sutural and oblique views of spores, S. A., U (f) Fig. 158 C sp? from Scomber joponica Spore in capsular view F. U. (d) Figs. 159–160—C sp? from Caulopsetta scaphe Fig. 159—Fresh spore F. U (d). Fig. 160—Stained spore S. A. H. (f) Fig. 161—C sp? from Mustellus lenticularis Capsular view of immature spore S. A. U (f) Fig. 162—C sp? from Mustellus lenticularis Stained spore S. A. H. (f) Ceralomyxa sp from Mustellus lenticularis. A single spore, and this immature, was seen in a Mustellus lenticularis heavily infected with Chloromyxum obliquum, which belonged to a species with large, clumsy, bluntly rounded spore valves (Fig. 161). It measured 30.2μ The capsules were 3.3μ in diameter, and were evidently immature as they were surrounded by large, vaculated capsulogenous cells. The sporoplasm

completely filled the spore cavity. The immaturity of the spore makes it impossible to estimate what the shape may be when fully mature. Ceratomyxa sp. from Caulopsetta scapha A species of Ceratomyxa forming large, bent spores with tapering valves was seen in two Caulopsetta The gall bladders were infected with other Ceratomyxa in both cases. The spores resemble those of C. laxa and C. uncinata, somewhat, but are distinctly larger than have been seen in either species. The average of seven spores were breadth, 38.3μ; thickness, 11.2μ; valvular axes, 27.0 and 23.5μ; capsules, 3.3μ Convex angle, 66°; concave angle, 107° and tangential angle, 96° Mean breadth: thickness ratio, 3 43. The spore form (Figs. 159, 160) suggests that it may be a very large, long-valved strain of C. uncinata, if it is not a distinct form. It occurred, however, in a host lacking C. uncinata, so that if it is a strain of C. uncinata it is one which breeds true for its larger spores Since there was too little material to characterize the form, it has been left as an unidentified species. Ceratomyxa sp. from Mustellus lenticularis A few spores and a single trophozoite of a large-spored Ceratomyxa were found in the gall bladder of a Mustellus lenticularis, heavily infected with Chloromyxum obliquum The single trophozoite was rounded, measuring about 40μ in diameter. It contained two spores, curved to permit their lateral extremities to fit within the contours of the trophozoite. The angular measurements of eight fixed and stained spores were: breadth, 59.4μ; height, 10.9μ, valvular axes, 37.6μ and 29.4μ; capsules, 3.8μ Anterior angle, 96°; posterior angle, 128°; tangential angle, 120°. Mean breadth: height ratio, 5.44. The slender spores are arcuate, with large capsules, and a somewhat enlarged central area (Fig. 162). The valves tend to collapse, somewhat, laterally, and the valvular tips of the spores in permanent slides are narrowly rounded. It is probable that the valves would be somewhat broader in appearance in fresh condition. The large, nearly spherical capsules lie near the anterior margin of the spore. They have short necks, and are somewhat convergent and rotated. The sporoplasm tends to be symmetrically placed. It is binucleate. Although there are a number of Ceratomyxa from elasmobranch hosts, and one from Mustellus canis, none are sufficiently close to this form to permit a final identification, and the small amount of material has not permitted it to be fully characterized. It is accordingly left as an unidentified Ceratomyxa species. The Trophic Stages Ceratomyxa trophozoites have but a limited value in taxonomic work, partially because of the difficulty of adequately characterizing them, and partly because of their considerable variability. They may differ in size, shape, abundance and nature of the cytoplasmic inclusions, in the speed and nature of their movements, in their tendency to form aggregations, as well as in the number of spores which they produce Unfortunately, there is no great stock of information concerning the dependability of these various kinds of traits as characteristic of species. The size of trophozoites belonging to the same species often differs enormously. When the population of trophozites includes budding forms, many small individuals may be present Most of these appear to be about 5 to 8μ in diameter, but occasionally much smaller ones are seen, as in C. minuta The size of the sporulating trophozoites containing well-developed spores tends also to be reasonably stable, apparently remaining about the same in different hosts at different seasons. The irregular shape of the trophozoites, however, makes measurement of the sporulating forms difficult. It is probable that the most characteristic measurement is that of rounded trophozoites containing well-developed spores. In a number of species in which the material permitted, an attempt was made to correlate the number of nuclei with the size of the trophozoite No success was met, for apparently the number of nuclei varies somewhat in the small trophozoites, and

there is relatively little nuclear division during the period of growth immediately preceding spore formation. Once the mature trophozoite has reached a size approaching its maximal volume, the sporoblasts begin to develop in earnest, with nuclear divisions leading to the establishment of the sporoblastic cells. It is evident that the period of cytoplasmic growth is not dependent in any direct way on the number of nuclei, and there seems to be no characteristic ratio of nuclear to cytoplasmic mass or surface. Trophozoites of a species come to have an individuality after they have been observed for a time, and when they are seen later, mixed with another species, they may sometimes be picked out with ease This is made considerably more easy if they can be compared in bright field, dark field, and phase illumination, as differences which are distinct in one illumination may be difficult to see in others Davis (1917, p. 205) observed the same indefinite characteristics, remarking “Although the plasmodia of the same species often show considerable variations in form, it is nevertheless true that each species is characterized more or less definitely by pecularities of form and structure, so that in many cases they can be recognized even in the absence of spores” There are instances, as in the complex of species inhabiting Caulopsetta scapha, where the differing types of trophozoites aid in sorting out the spores into categories. In general, the variations of form seen in the trophozoites tend, more or less, toward the establishment of three different types of basic organisation, which are somewhat imperfectly related to three habits of the trophozoites, motility, attachment, and floating. The typical motile form has been described admirably by Davis (1917). It is basically pyriform, with the anterior portion of the body enlarged, and a posterior, tapering region, the posterior process or Stemmpseudopodien” of Doflein (1898). Small, filiform pseudopodia are formed, usually at or near the anterior end, and move back as the animal moves forward, often with movements of the ectoplasmic layer as well Doflein (1898) attributed the forward movement to the extension and enlargement of the posterior process Davis (1917) criticized this point of view, and like Davis, I have seen no instance in which movements appear to have resulted from the increased size of the posterior process On the other hand, Davis (p 206) states, “The sole method of locomotion is by means of pseudopodia, which vary greatly in shape and structure in different species' Certainly, in rapidly moving forms, the pseudopodia appear to be primarily responsible for movement, but in sluggish forms, particularly, trophozoites in which there is no evidence of anterior pseudopodia have been seen in movement This type of movement was particularly evident in the trophozoites of C. gemmaphora and Leptotheca minima Apparently the movement of the ectoplasm alone may propel the organism A few large trophozoites have also been seen moving, apparently in the absence of pseudopodia, in the case of C. hama. The appearance of the motile form, while more or less similar in different species, is subject to considerable variation. The pyriform shape is most evident in small trophozoites This is true, whether one thinks in terms of the smaller members of the same species, or of smaller species as compared to larger ones As the body becomes more bulky, the pyriform shape tends to become less definite, with the posterior process becoming larger, more bulky, more elongated, and eventually producing a rather clavate shape, although with an anterior, enlarged region. Although the pyriform shape is especially related to motility, many pyriform organisms show no movement at all, evidently floating free in the bile, or attached to the wall of the gall bladder Individuals which are in the process of transforming from the pyriform to a clavate or irregular shape are seen, apparently in response to a floating or attached attitude. The floating forms are often more characteristic than the motile forms. They are usually irregular in shape, with pseudopodia extending from both ends, when

elongated, or from all parts of the body surface Transition from the motile to the floating form appears to be accomplished by detachment from the substrate, and the development of one or several elongated pseudopodia at the old anterior end, producing a triangular shape (Fig. 75) Particularly in small trophozoites, only one anterior pseudopodium is formed, when the organism appears as a blob of protoplasm in the middle of a single, long, protoplasmic filament, which tapers as it approaches the enlarged area. In larger floating forms, or in trophozoites which have adapted to a floating attitude for a longer time, pseudopodia often extend outwards from various points on the body surface (Fig. 98). When the trophozoites assume this form, they usually exhibit no movement whatsoever, if the fluid in which they lie is not in motion Occasionally one is seen slowly tilting one pseudopodium to one side This inclining of a pseudopodium is often seen in motile form, also, and is not invariably accompanied by a change in direction. It is in floating forms that branched pseudopodia are especially numerous, and characteristic apperances often develop. When the pseudopodia are fully extended, fortuitous groupings sometimes occur through the pseudopodia becoming entangled. This may or may not initiate a true aggregative activity with more or less intimate union of the various members of the group. Trophozoites attached to the walls of the gall bladder are quite variable. In some cases attached forms retain an essentially pyriform shape. This is true, for example, of C. vepallida, where there is very little difference in the outlines of attached and motile forms Although they may have outlines resembling the motile forms, however, such attached trophozoites usually move very little. Indeed, they are often packed together so closely that movement is next to impossible. In the majority of the species which have been observed here, however, the attached trophozoites tend to be clavate In larger, sporulating trophozoites, the developing spores often lie in a definite enlargement of the otherwise clavate organism. The enlargement may be at any point of the clavate body, although it generally tends to be located rather near to the wall of the gall bladder Attached trophozoites tend to produce few pseudopodia, although in some cases elongated pseudopodia are extended outward into the bile. This is particularly common in the trophozoites of C. inversa. At the end of the body attached to the gall bladder epithelium, it is not uncommon to see one or several filopodia, insinuating themselves between the cells In some cases, as in C. polymorpha, the attaching surface is somewhat flattened, and produces a number of delicate, pointed pseudopodia which pass to the surface of the cells, and, to some extent, between them Surprisingly little host reaction develops in the regions in which pseudopodia extend between the cells. A rounded form with lobopodia rather than filopodia is commonly encountered. As Davis (1917) observed, this is most usually a response to an unfavourable environment, essentially a victory for the forces of surface tension over the forces exerted by the trophozoite and required for extension of the body. On the other hand, in some species motility is seen in the rounded forms with lobopodia, so this form may be normal in some species Perhaps the most useful attribute of the rounded form is that it reveals, more directly than in most shapes, the volume of the body, and it is certainly true that measurements based on the rounded form are more reproducible. It would appear that the rounded form is more commonly normal in inhabitants of the urinary system than the biliary system. One of the features of the Ceratomyxa trophozoites which are promising, from a taxonomic point of view, is the nature of the cytoplasmic inclusions. The smallest trophozoites usually have a homogenous protoplasm, with few or no inclusions. The larger ones develop inclusions, which may be roughly categorized as small, darkish or bright granules, often irregular in shape, larger spherical ones, sometimes inconspicuous and sometimes refractive, and other spherules, somewhat more variable

in size, apparently consisting of an oily substance and appearing as white-rimmed bodies in dark contrast phase and dark field illumination. Trophozoites differ in the kinds of inclusions present, and in the abundance of the inclusions Differences are sometimes consistent, and sometimes inconsistent. Davis also has observed that inclusions tend to appear during growth, and raises the question of whether the trophozoites of the same species may include mature individuals with and without inclusions, as he found (1917, p. 206) “evidence that full-grown trophozoites may sometimes fail to develop spherules” In this, further evidence has been obtained, in the case of C. minuta, the trophozoites of which sometimes possess spheruies and sometimes do not However, in some other species, constancy appears to be the rule, as in C. hama, which always possesses some spherical inclusions, and in C. vepallida, which appears never to develop this type of inclusion. The inclusions, although usually about the same in size and shape, are evidently not all of the same chemical composition In some species the inclusions persist in permanent preparations, while in others they are lost In some species the inclusions are acidophilic, while in others they tend to be basophilic In some species the inclusions tend to be stained darkly by Janus green or neutral red, while in others they are but lightly tinted, or resist staining with these dyes. The inclusions which are darkly coloured by the vital dyes tend also to be more refractive, and in some instances to appear as white-rimmed bodies in dark contrast phase. These are probably at least predominantly lipoidal in nature. In several instances, small nuclei on the side of the inclusion tend to take the dye particularly intensely during the early stages of the staining reaction, thus behaving as though they were lipoidal bodies taking up the stain by flocculation (see Baker, 1958). It is interesting to note that in trophozoites, apparently of the same species, the inclusions are not always consistent in their reaction to vital dyes Thus in C. hama the spherical inclusions are sometimes strongly tinted by Janus green, and in other cases but lightly coloured. It is also evident that some small, apparently oily, droplets, not reactive with either neutral red or Bismarck brown, and apparently derived from the host tissues, occur In some instances Janus green tints these as darkly as it tints the inclusions in the organism. The less refractive spherical inclusions occur in many species of Ceratomyxa. These typically appear as dark bodies in dark contrast phase, and may or may not be visible in dark field illumination. If the less refractive spherules occur, refractive, apparently oily spherules are not abundant. In some species, as in C. laxa, there appears to be some spatial relationship between the dark, granular inclusions and the spherical ones, as if the granules were formed in or about the spherules. The spherules fall into several categories with respect to their reactions to fixatives and stains, as well as to alcohols and clearing agents. It is probable that a careful study of the properties of these bodies might lead to the development of useful criteria in taxonomic work. Small granular inclusions, sometimes, as in C. laxa, very abundant, fall into two obvious categories, darker, rather brownish ones and clear, transparent ones. All of the darker granules that have been seen are resistant to fixation, as well as to water, alcohols and xylene In this respect they behave rather like the colour-carriers found in elasmobranch Chloromyxum (Erdmann, 1917). The colourless granules, on the other hand, are more transitory, usually disappearing entirely during the preparations of permanent slides. Mitochondria are but little known in the Myxosporidia. While they can be demonstrated rather easily in some species, with Janus green B, they appear to have little value from the taxonomic point of view In all species in which they have been successfully demonstrated they have been short rods or spherules, about 0.25μ to 0.50μ in length. and are located, for the most part, in the endoplasm. In

some species, as in C. castigata, they have been more easily demonstrated, but this is probably more a matter of differences in the pH and chemical constitution of the bile than inherent differences in the mitochondria of different species. All in all, the inclusions would appear to be the most promising method of distinguishing some trophozoites on an objective basis, but until much work has been done in identifying them and describing some of their properties, their value is limited. Above all, information on the effect of the biliary environment, its pH, viscosity, and such factors, upon the development and abundance of the inclusions is needed. The tendency of trophozoites to aggregate is not uncommon in the Ceratomyxa. They aggregate in three ways. In some cases they pack together in masses adjacent to the epithelium of the host organ. This appears to be fortuitous, and in most species is not followed by a union of the associated organisms. These clusters are not necessarily monotypic. In cases of mixed infections, trophozoites of different species are found mixed together, without any strong evidence of tendencies for the species to arrange themselves selectively In a count of C. hama, C laxa and C. vepallida trophozoites attached to the wall of the gall bladder of a Caulopsetta scapha, it was found that in different areas there were somewhat different percentages of trophozoites of the different species, and that, on a statistical basis, the probability of the arrangement to have occurred as a result of random chance was about 14. There may, therefore, be some tendency toward selective arrangement, but if so the tendency is not extremely great, insofar as the sample could show Another fortuitous type of aggregative activity occurs in species with long, branching pseudopodia, in which the pseudopodia become entangled. This is seen quite commonly in floating organisms C. laxa, C hama, C gemmaphore, C inversa and C. castigata trophozoites not uncommonly become entangled when infections are heavy. As with the massing at the surface of the epithelium, entanglements may involve members of several species in cases of mixed infections. In the majority of species, there is no subsequent union of the individuals. A third, and more important, type of aggregation occurs in some species, involving the partial or complete union of the separate trophozoites. This is usually effected through the union of pseudopodia In C. intexua this union is extremely intimate, with all traces of the original separation between the individuals disappearing, insofar as can be seen. Generally this type of union is most common in trophozoites with sporoblasts, and the small trophozoites remain isolated. The aggregation of larger trophozoites is not necessary for the production of spores, however, and isolated sporulating individuals with maturing spores are seen as well as the larger aggregations. It would appear that this type of aggregative activities is characteristic of some species, but does not occur in others. It is also probable that it occurs only where infections are relatively heavy, so that in lightly infected hosts a species capable of aggregative activities may not demonstrate it. While the number of spores produced by trophozoites is predominantly two, monosporous individuals are not uncommon in some species. A few preliminary counts suggested that the proportion of monosporous and disporous individuals did not remain constant from host to host. It is probable that in the majority of species with a normally disporous development, monosporous individuals sometimes occur. The polysporous habit, however, appears to be unusual, and to be characteristic only of a few species. It should be noted that in at least some species, as C. laxa, some trophozoites are monosporous, some disporous, and some polysporous. During the course of the study, no evidence to suggest a seasonal difference in the trophic stages has been found. About the same variation in form and size occurs in both summer and winter samples. It would appear, however, that the proportion

of small and sporulating trophozoites does undergo a change during the course of the infection. In some individuals, there are many small trophozoites, and but a few sporulating ones, while in other hosts, there are many sporulating forms, and but few small trophozoites. It would appear that during the earlier stages of an infection, more trophozoites are undergoing endogenous or exogenous budding, with the result that the proportion of small trophozoites is considerbly higher. The point needs confirmation through infection experiments. Spore Variability Spore variability is of keen interest to the taxonomist, for the interpretation of his data depends upon his knowledge of the amount and kind of variability characteristic of species. The parasites occurring in a single host are a local population, for the moment fully isolated from all other members of the species. It is only as spores from many hosts mingle in the water during transmission, or as hosts are repeatedly exposed and infected, that the local population partially merge into the larger geographical population. Where infection rates are high, it may be assumed that repeated infections occur rather commonly, and in such cases the parasites found in a single host may well have a diverse origin. On the other hand, where infection rates are low, most of the hosts are probably infected by coming in contact with one or a few spores on a single occasion, so that the local population is made up of individuals with a common origin. At all events, when we study a sample of parasites from a given host, we are studying this sample as representative of the species. It is, however, but a sample of a sample, as it were, a small fragment of one of the many local populations which, together, constitute the species. To what extent can it be considered as a reasonable picture of the species as a whole? How much difference may be expected in parasite populations from other hosts of the same species? Of different species? Populations from one Host. The Ceratomyxa spores, thin-walled and easily modified by forces acting upon it during development, is admirably suited for an examination of variability. At first glance all of the spores look quite similar, but on closer inspection, one sees many differences, and in the more arcuate species one realizes that the impression created by drawing one spore as typical of the species would make it almost impossible for another investigator to recognize that some other spores which are occurring with it as belonging to the same species. Measurements cannot reflect this variability in any very complete sense, although the inclusion of ranges of size and shape are important as aids in the recognition of species Any objective discussion of spore variability must depend, primarily, upon the kinds of variations which can be measured. It must be borne in mind that other spore features no doubt vary in much the same manner as those which are discussed. Linear dimensions have been based, with some exceptions, on samples of 25 spores, carefully selected for orientation, but otherwise selected at random (see p. 3) It is important to note that the information gathered on variability is based on samples of this size. Had samples of 100 been used, the ranges for the various dimensions would have been somewhat greater. However, the Ceratomyxa generally show but a moderate central tendency, with the result that somewhat more of a sample of 25 is from the more extreme portions of the range. This is not true of all genera. The Myxidium found in New Zealand, for example, have a much more constant spore shape, and a much stronger central tendency in their frequency distribution. Table V summarizes the amount of variation seen in spores taken from a single host, insofar as the principal linear dimensions are concerned. Where a species has been seen in several different hosts, the sample showing the greatest variation was used, for each dimension. It is evident that a range of breadth of 10μ in a

species with spores averaging 20μ in breadth is by no means comparable to a similar range in breadth in a species with spores averaging 100μ in breadth. In order to put the results in a proportionate form, the differences between the smallest and largest examples was expressed as a percentage of the mean. Table V —Variation of Ceratomyxa Spores from the Same Host, Expressed as Percentages of the Mean Sutural Longer Shorter Species Breadth Diameter Valve Valve % % % % minuta 29.8 24.2 37.0 31.0 inconstans (Scomber)* 33.4 45.6 33.1 41.0 (Trachurus) 36.4 41.8 52.5 36.0 (Helicolenus) 56.4 49.8 50.0 47.5 (Usacaranx) 39.8 42.0 50.0 36.0 faba 29.1 19.3 40.6 50.0 castigata 28.1 17.0 26.6 37.0 castigatoides* 34.9 28.4 36.8 36.2 declivis 18.8 28.8 40.6 50.0 intexua 70.2 45.6 67.6 71.5 recta 12.9 25.6 18.5 15.7 flexa* 20.6 41.0 36.8 36.2 insolita 43.0 26.5 29.5 28.0 gibba 27.8 34.9 74.4 39.7 gemmaphora 45.9 34.2 50.0 52.2 subtilis 47.9 28.2 54.0 52.0 laxa 42.9 22.0 38.3 23.0 arcuata** 80.4 57.4 59.6 72.2 constricta 21.4 41.8 17.0 24.1 torquata 23.4 28.4 45.6 46.4 renalis 39.3 27.5 31.1 67.0 moenei 33.8 45.7 44.3 28.0 uncinata (Pelotretis) 21.2 23.7 24.3 19.0 polymorpha 48.0 33.9 31.4 44.6 hokarari 35.6 19.7 68.9 55.7 angusta (Hypoplectrodes)* 49.0 19.6 33.7 46.8 (Helicolenus) 24.7 20.1 25.9 49.4 inecersa 36.8 28.3 39.6 57.4 elongata 29.0 35.4 35.7 27.7 It is evident at once that the samples do not agree very well. Some samples were quite uniform, while others were much more variable. There can be no doubt that some of this fluctuation is a matter of sampling error, but not all of the differences can be attributed to error. The spores of C. recta appear to be uniform, relatively speaking, while those of C. arcuata appear to be variable. The figures for percentage of variability merely bear out what is seen by casual observation, and tend to set some objective limits upon it. Furthermore, although the data do not appear on the table, second samples of the species with a higher variability tend to show a higher variability. It is evident that one may judge fairly well from the amount of variation seen in one host what is likely to be seen in a sample from another host. It would be helpful to find some factor which would tend to predict the amount of size variation that might be expected in a species Every attempt to find such a factor has failed. The search was conducted in but a preliminary fashion, by arranging the species in lists, according to their order in size, angular measurements,

curvature, and breadth: height ratios. The lists were then divided into four parts, with 25% of the species in each part, and the average percentage variation for the groups of species determined. No definite evidence of trends emerged. In general, one cannot expect either more or less percentage variation in species with smaller spores, straighter spores, or more slender spores, although there are a few generalisations mentioned below which developed from this approach. The only conclusion that can be drawn is that there is always a considerable variability in the linear dimensions of Ceratomyxa spores, and that this variability is independent of the other measured features of the spores. For the majority of species, however, one may expect the spore breadth to vary between 20% and 40% of the mean spore breadth. This would indicate that the range of sizes would extend from about 10–20% of the mean breadth above and below the mean. About 62% of the smaples shown on the table fall in this range. The sutural diameter, on the other hand, tends to follow a bimodal pattern, with peaks in the 20–30% range and 40–50% range. About 38% of the species fall in the former and 35% in the latter range. Although there are too many exceptions to make it a valid generalization, the more slender species tend to fall in the former and the more stubby species in the latter group. The length of the valvular axes tend to vary rather like spore breadth. About 60% of the samples have percentage variation between 20% and 40% for the longer valvular axis. Oddly enough, there is more varibility in the percentage variation of the shorter than the longer valvular axis, samples showing no tendency to cluster in the range between 20% and 60% variation. Table VI is a summary of the range of variation seen in spores from a single host in angular measurements and indices. These have not been expressed as percentages, since there is no reason why a spore which is straighter should necessarily have more or less variation in its angular measurements. As with the linear dimensions, it is evident that some samples are much more uniform than others. This difference between samples cannot be entirely a matter of sampling error. Some samples are more uniform than others. In nearly all of the samples, the anterior angles show a variation between about 15° and 60°. The median for all species is about 35°. The posterior angle is somewhat more variable, with a median of about 48°, and the tangential angle intermediate, with a median of about 39°. There is little tendency for the samples to cluster near the median values, however. The curvature index is quite variable, tending to sum up the angular variation of posterior and anterior angles. The taper index is relatively stable. About 62% of the samples have a range of variation between 20° and 40° for the taper index, and the median range is 23°. The valvular index is a little more variable than the taper index, with a median variability of 28°. It is evident that where samples have a maximal spread of 20° or 30° for taper index and valvular index, the mean values will tend to have a high degree of significance. In general, despite their variability, the means of the angular measurements tend to be quite reproducible. The curvature index, also, tends to have a dependable mean, and the mean taper index and valvular index are among the most stable features of spore measurement. In spores with a slender, elongated shape, the taper index becomes extremely stable, as does the valvular index. In the spores of shorter, stubbier species there is considerably more variation. Angular measurements, also, tend to be more stable in the species with slender spores. Measurement of the polar capsules is not very satisfactory. The least difference which can be counted upon is the minimal unit of the micrometer scale, or the millimeter in the case of camera lucida tracings. These do not provide a delicate enough system to permit a discussion of variability, and the factor of foreshortening in capsules which are inclined toward the suture or toward the lateral margins of

Table VI. —Variation in Angular Measurements and Indices in Ceratomyxa Spores from the Same Host. Ant. Post. Tang. Curv. Taper Valvular Species Angle Angle Angle Index Index Index Deg. Deg. Deg. Deg. Deg. Deg. minuta 26 31 15 52 22 27 inconstans (Scomber)* Fixed, measured in water suspension. 28 81 17 101 99 104 (Trachurus) 48 60 73 95 50 36 (Helicolenus) 56 60 41 96 28 45 (Usacaranx) 52 53 50 81 52 50 faba 63 89 55 153 37 31 castigata 19 32 16 60 22 20 castigatoides 33 28 15 55 26 28 declivis 18 23 23 38 21 30 intexua 42 55 50 86 17 77 recta 14 21 0 14 20 23 insoliba 27 33 35 60 11 14 gibba 35 38 39 73 26 32 gemmaphora* 36 64 47 100 23 45 subtilis 82 84 82 152 11 26 laxa 28 43 39 63 15 28 arcuata** Fixed measured in permanent slides. 45 52 50 96 7 18 constricta 41 43 58 76 44 31 torquata 47 73 55 117 21 40 moenei 19 18 23 33 11 24 polymorpha 35 56 60 86 72 39 angusta (Hypopolectodes)* 17 17 20 34 9 7 (Helicolenus) 17 9 7 25 11 9 inversa 77 76 43 153 34 13 elongata 22 17 15 37 8 12 Convex Concave Tang. Bend Taper Valvular Angle Angle Angle Index Index Index torquata 27 41 38 68 18 24 laxa 30 76 54 105 24 39 renalis 28 41 44 65 10 19 hokarari 60 83 83 142 22 32 polymorpha 48 56 66 92 44 45 inversa 41 54 50 95 4 17 the spore make for further error. In general, some species tend to have capsules of different size, while others do not. The species which have unequal capsules do not always produce spores with capsules of different sizes, so that differences are essentially statistical. The origin of the differences in capsular size is not so easy to trace. There appears to be no significant difference in the size of the capsulogenous cells or nuclei, and in some species the impression is gained that the smaller capsule is more juvenile than the larger one, as though it were a difference in rate of development which caused the difference in size. In C. polymorpha, particularly, the smaller capsule sometimes appeared to retain a juvenile appearance. It is not impossible that abnormal development in less than optimal condition following the death of the host may account for some of the discrepancies of capsular size that have been seen in the New Zealand material. Abnormal spores are not uncommon. The most common type of abnormality is membrane deformity. This may involve wrinkling of the spore membrane, or may, in extreme cases, involve the complete collapse of the empty portions of a valve.

It is probable that most of these deformities are related to dehydration, resulting from increased viscosity of the bile, or other similar factors. They are much more common in some species than in others, and are sometimes so abundant as to represent a “normal” abnormality. This is particularly true with some of the thinwalled Ceratomyxa from Caulopsetta, and judging by the illustrations drawn by other investigators, is not uncommon among the Ceratomyxa of flat-fishes, generally. These partially or wholly collapsed spores are often quite characteristic in shape, and, perhaps, would justify more careful investigation In the present paper, all of the effort was concentrated on spores in which no valvular collapse had occurred. The triad form is a common abnormality in the Ceratomyxa, and while less common in Leptotheca, also occurs in some members of that genus. It is a ubiquitous trait, triad spores having been seen in all but four or five of the species studied. The triad form is much more common in some samples than in others. It is probably not a dependable trait for species characterization as a rule, for in at least some of the species studied here, the triad form is common in the cooler months and much less common in the warmer months. This was particularly the case with C. polymorpha, in which species some samples taken during the winter had about 60% of the spores in triad form, while others taken in the summer contained few or no triad spores. The tetrad form is never common, or at least has never been found as a common trait in any of the species so far described. It is not uncommon in C. polymorpha, where several per cent of some of the winter samples have a tetrad form. It is interesting to note that although several thousand such triad and tetrad forms have been examined, every spore with three valves has three capsules, and every spore with four valves has four capsules. It seems evident that these two features of the spore do not vary independently, and that whatever factor results in the formation of a supernumerary valve also results in the formation of a supernumerary capsule. It would seem, from Noble's careful study of the development of C. blennius that this pattern of dependency is not the result of cell lineage, since in that species the valvular and capsular nuclei are segregated early (Noble, 1942). Either C. blennius is exceptional in its cell lineage, or some other factor must be concerned. It is interesting to note that in species in which triad spores are common, trophozoites containing one triad spore usually contain another triad form. Only about half a dozen trophozoites have been observed with one triad and one normal spore. Populations in Different Hosts of the Same Species. In several instances the same species of Ceratomyxa has been recovered from several hosts of the same species. In some instances comparable data were not obtained for the second or subsequent hosts, either because the infection was light and only a small sample of spores were measured, or because material was fixed before measurement. The results which are summarized in Table VII are obtained from samples which were measured under comparable conditions, using only fresh spores. In most instances, the samples were all full samples of 25 spores, but in some instances small samples of 10–20 spores were included in order to get a somewhat more complete comparison. In the table, wherever such small samples have provided the maximal deviation from a mean for one or more of the dimensions, the species name is followed by an asterisk. The number of samples which were compared is shown in parentheses following the species name. The method used was to convert the maximal deviation of a sample mean from the species mean (or mean of all samples) into a percentage of the species mean. It is of little consequence that some samples showed a high agreement. If they are, indeed, samples of the same species, agreement is to be anticipated. It is of more interest to observe how much deviation a sample may show, either as a result

of random chance or of real differences resulting from the different environment provided by a different host animal, or differences in the genetic content of the populations found in different hosts. Although smaller differences are more common, it would appear that occasional deviations in the order of 15–20% of the species mean are to be anticipated. The question immediately arises as to the origin and significance of these differences. Are they the result of random chance, involved in the samples themselves, or are they to be considered as real differences in the nature of the populations in the different hosts? In an attempt to gain some preliminary information on this point, the differences observed were checked by statistical methods. Particularly where small samples are involved, non-parametric methods such as the Wilcoxon or Kendall sums, are particularly useful. In most cases, at least two methods of comparison were used. The results showed that it was rare for two samples to agree so closely that there was not a strong probability of real differences between the two samples. In all but two samples, differences which had a probability of being attributable to random chance in the order of .05 to .02 were found. In a number of cases, differences were found with a probability of .01 or well beyond. While an occasional sampling error may occur, and the statistical methods can only give some estimate of the likelihood of chance causing a difference of a given magnitude and direction, there could be no doubt that samples of spores of the same species quite often show statistically valid differences when taken from different host animals of the same species. One can only conclude that local populations, found in different hosts, are not uniform. They exhibit real differences in one or more of their dimensions in many instances. Since this is the case, it is evident that statistical methods are too delicate to be more than a very rough guide in systematics. These methods are capable of detecting differences which are of less than specific rank. This, of course, is only what might be expected, for the statistical methods have their greatest biological usage in the discussion and description of differences of a genetic or ecological nature within the species group. If the differences seen in different hosts are to be thought of as real, the matter of their origin is of considerable import. There are many possibilities, and it will require a much more thorough and definitive study to determine which of these are the most consequential. The possibilities may be categorized as being intrinsic— that is, dependent on the organism itself as an inherent trait, or extrinsic, dependent on factors external to the organism. The natural hypothesis in the case of intrinsic factors is that there are genetic differences in the local populations occurring in different host fishes. The autogamous life cycle would, on theoretical grounds, tend to produce a relatively homozygous strain, although some heterozygosity might be retained if there were a selection favouring it (see Meglitsch, 1952). If the local populations tend to differ genetically, one might suppose that in species with high infection rates, there would be a greater diversity of forms within a single host, and somewhat greater uniformity of populations in different hosts, as infections might more frequently arise as a result of several exposures to the parasite. In a survey of the type that has been carried out, sufficient data for the purpose cannot be obtained. It may be pointed out that in the more common species, seen in a larger number of hosts, there is a greater opportunity to see differences between samples, while in the less common species, it is difficult to obtain more than two or three infected hosts. Insofar as the percentage variation seen in single hosts is concerned, there was no clear evidence of greater uniformity in less common species, while in the case of populations from different hosts, the greatest deviations were seen in the most common species. All that can be said is that the little data gathered have not favoured the idea that there are significant differences in the genetic strains found in different host animals. There is certainly much too little information to conclude that such differences do not occur.

There are a variety of possible extrinsic factors which may tend to increase or decrease variability in samples. Some of these centre on the host animal, its age, size, sex, and general condition. Few opportunities for observing this type of variation have arisen. One of the two Congiopodus infected with C. castigata was very small, while the other was quite large. There was no unusual amount of difference in the two samples. Larger and small Caulopsetta containing C. laxa and C. hama also tended to substantiate the idea that there is little effect of host size upon the spores of the parasites. There is, likewise, no clear indication of differences being more marked in the case of parasite populations from male and female fishes. Again, while there is far too little data to make a negative conclusion, there has been no indication that the size, age or sex of the host has a significant effect upon the spores of Ceratomyxa. Factors extrinsic to the host are also potential determiners of spore variability Differences in season, accompanied by the various changes in physical factors involved in seasonal changes, might possibly influence the development of spores. Here, in one case, and this the only one in which there was adequate opportunity to observe, there appears to be a factor operating which may definitely influence the spore. In C. hokarari samples, the ones taken in the winter months were composed of smaller spores than those taken in the summer. In other species no such trend was noticed, or the results were indecisive. It is not improbable that water temperature may, through affecting the internal temperature of the cold-blooded host, affect the formation of spores. Whether this is a consequential factor, and whether it always tends to favour the formation of larger spores in warmer conditions, will await further analysis. A still further factor might be of consequence in some instances. Where infections are light, there is less crowding, and under such circumstances spores may be formed which are larger, or otherwise unlike those produced in crowded conditions. There has been little opportunity to observe this point, but where the data can be used to examine this point, the results tend to suggest that this factor is of little consequence. One of the two infections with C. minuta was light, but there is no evidence of unusually great difference between these samples. The smaller spores occurred in the more lightly infected host, in this case. In C. hokarari and C. polymorpha, some heavily infected hosts and some lightly infected hosts were seen, and there was no clear indication of trends toward larger, smaller, or otherwise different spores in the more lightly infected hosts. Table VIII summarizes the deviation of sample means from the species mean for angular measurements and indices. It is evident that one may expect a sample to deviate from 20–25° from a species mean in angular measurements. The shorter, more stubby spores, on the whole, have more variable angular measurements. The taper index and valvular index tend to be more stable than the angular measurements, and differences of more than 10° are to be regarded as unusual, warranting careful examination for other evidence of specific differences. Table VII. —Differences in the Means of Samples, taken from Different Hosts of the Same Species, Expressed as Percentages of the Means of all Samples. Height Breadth Thickness Longer Shorter Valve Valve Species % % % % % minuta (2)* 8.5 1.7 1.8 12.2 10.4 castigata (2) 1.5 0.0 — 5.5 14.5 vepallida (3)* 6.3 19.3 — 15.1 6.5 hama (3)* 9.3 9.9 6.5 1.2 3.4 laxa (2)* 0.4 — 2.6 2.8 3.7 hokarari (5) 16.3 — 7.9 10.0 15.6 polymorpha (3) 12.7 6.9 5.6 17.6 12.3

Table VIII. —Differences in the Means of Angular Measurements and Indices of Samples taken from Different Hosts of the Same Species. Ant. Post. Tang. Curv. Taper Valvular Species Angle Angle Angle Index Index Index Deg. Deg. Deg. Deg. Deg. Deg. minuta (2)* 7 0 2 7 2 7 castigata (2) 1 1 2 4 3 3 vepallida (3)* 5 24 16 22 20 15 hama (3)* 16 25 20 21 7 10 laxa (2)* 2 4 0 6 3 2 hokarari (5) 16 23 22 41 3 8 polymorpha (3) 2 15 10 16 5 11 It is clear that the angular measurements are quite reproducible in samples taken from different hosts. It is not possible to compare them directly with linear measurements by converting them to a percentage figure. It was noticed that statistically valid differences between samples taken from different hosts were much more uncommon in the case of angular measurements. They occurred, especially in the case of C. hokarari, in which there is a very marked difference in the amount of curvature. Even in this species they were not as common, and the probability values were consistently lower than for linear dimensions. This may be interpreted to indicate that linear dimensions are somewhat more delicate than angular measurements, or that form is somewhat more stable than size. The angular measurements are, as yet, too new and insufficiently tested to permit drawing conclusions on this point. Populations from Hosts of Different Species. The survey undertaken was not extensive enough to provide a great deal of information on species found in more than one host species. The few instances recorded here are insufficient to permit any generalizations of consequence. The whole subject is, for that matter, strongly biased because traits considered by some to be of sufficient magnitude to recommend the assignment of strains to different species are, by others, explained on the basis of host-determined strains. As a result, it is almost impossible, without cross-infection experiments, to find an initial point of departure. Among parasites, it would be reasonable to suppose that a species consists of a group of organisms which are capable of infecting the same hosts, ordinarily producing in them an infection with individuals having common morphological features. The form of organisms, however, is determined in part by the environment in which they find themselves. It is not inconceivable, therefore, that when a parasite exhibiting a characteristic form in one species is transmitted to a host of another species, it will assume a form which is consistently and measurably different. Where such a situation is brought about in the laboratory under experimental conditions, but does not occur in nature, it merely provides information about the potentialities of the species concerned. Where, on the other hand, the species normally occurs in both host species in nature, and normally assumes two different forms, depending on the host species in which it finds itself, we are justified in speaking of host-induced strains of the same species. Some have been quite willing to assume that such modifications of parasite form occur frequently and regularly, and where the basic appearance of two strains inhabiting different host species is quite similar, the consistent and measurable differences are ignored at the specific level, being attributed to differing conditions in the two host species. Others have been less willing to assume that such differences are common, and tend to elevate any consistent morphological difference to the specific rank. In the final analysis, of course, it is only through infection experiments that either assumption can be finally confirmed, and in the forseeable future, we may expect that there will be different tastes in this matter.

In the present paper, several instances of a species of Ceratomyxa living in different host species have been reported. These strains may, or may not, be conspecific, of course, and the decision to include them in the same species category merely reflects the personal taste of the investigator. It is, perhaps, more to the point to ask whether parasites do occur in hosts of different species and exhibit no more difference than parasites occurring in hosts of the same species. In this respect, the answer is evidently in the affirmative C. angusta is described as occurring in Hypoplectrodes semicinctus, and in Helicolenus percoides. In this case, one sample was measured in fixed condition, and the other in the fresh condition, so linear dimensions are not strictly comparable. The angular measurements, however, are almost identical insofar as mean values are concerned. In a number of cases samples of the species, found in different examples of the same host species, exhibited differences of greater magnitude. The same close agreement of both linear and angular measurements has been seen in samples of C. inconstans, taken in Usacaranx lutescens and Helicolenus percoides. It is evident that, while in some cases life in a different host species may result in demonstrable differences, it is not always modified. In the case of C. uncinata, the spore population in Caulopsetta scapha was somewhat more bent than that seen in Pelotretis flavilatus. The difference between two samples seen in Pelotretis was less than the difference between either of the Pelotretis samples and the Caulopsetta sample. On the other hand, the extent of the differences in spore curvature was no greater than that seen in different samples of C. hokarari, all of which came from the same host species. In general, however, this was merely the rule which was followed in assigning strains to species rank. In the case of C. inconstans an exception was made, for the extreme varibility of the original material described by Jameson was matched only by the range seen in four different host species in the New Zealand material. Although the material from Trachurus, Scomber and Usacaranx were placed in the same species because of the close relationship of the hosts and the extent of the variability which Jameson noted, it is far from impossible that each of these strains will prove to be distinct. Of the forms which have been described as belonging to different species, and which occur in related hosts, there are few which give the appearance of being different as a result of the effects of living in hosts of different species. There are certain similarities between C. laxa and C. uncinata, and if one were found only in one host species, and the other in another host species, it might be suggestive of host-determined differences. However, recognizable C. uncinata occur in Caulopsetta scapha, the host in which C. laxa is also found. Perhaps the most convincing series of strains are those which occur in fishes of the mackerel type. There appear to be a number of small-spored species inhabiting these fishes, which are morphologically distinct, and which are nevertheles quite similar. C. minuta, for example, is not unlike C. inconstans in many ways, and Leptotheca minima, from Arripis trutta is also quite similar Of the material seen in the present study, these provide the best, and almost the only good, evidence of the possibility of host-determined strains belonging to the same species, and inhabiting fishes which are relatively closely related. If one includes more remotely related hosts, the case of C. castigata and C. castigatoides might serve as a possible example. These two strains are very similar, and while they are not identical in appearance, might well prove to be conspecific if their biological distribution among host animals can be brought closer together, or if infection experiments can be used. Except for these cases, there has been very little evidence of host-determined strains among the Ceratomyxa In most cases there are discrete differences between the morphological traits of one strain and another, while there is little room for doubt about the identity of other strains occurring in different host species. Rather similar results have been obtained with other genera, and will be reported in subsequent sections of this report.

Genus Leptotheca Thélohan Leptotheca is one of the larger genera of Myxosporidia, including at least 30 species at the present time. The various species are coelozoic in the urinary or biliary systems of fishes and amphibia. The genus is cosmopolitan, members having been reported from marine fishes from all parts of the world. Kudo (1933) listed a total of 23 species. Since that time the following species have been added to the list:– 24 L. elegans Noble, 1938 25 L. compressa Noble, 1939 26 L. sphaerula Noble, 1939 27 L. latesi Chakravarty, 1943 28 L. macronesi Chakravarty, 1943 29 L. vikrami Tripathi, 1949 30 L. subelegans Laird, 1953 The relationship of Leptotheca to Ceratomyxa has been discussed in conjunction with the latter genus. The techniques used in the study of Leptotheca have been identical with those used with the Ceratomyxa. Leptotheca are considerably less common in New Zealand fishes than Ceratomyxa. Only three species have been found in sufficient abundance to permit their characterization, and several other forms seen in insufficient numbers to permit a final identification. Leptotheca minima n. sp. (Text-Fig. 13, Figs. 163–168) Habitat. This species occurs in the gall bladder of Arripis trutta (Bloch and Schn). Infected fishes were taken by trawlers working out of Wellington in July and October. There was no evidence of damage to the host organ. Trophic Stages. The trophozoites range from about 4μ to 15μ in diameter. They are rounded to pyriform in shape (Fig. 163) with a short posterior process, frequently with one or two small filiform pseudopodia at the sides of the posterior process. Pseudopodia with one or two blebs are common. No trophozoites in rapid movement have been seen, and none with anterior filopodia have been observed. The protoplasm is very transparent and clear, with no visible distinction between ectoplasmic and endoplasmic regions. A few to several dozen bright, refractive bodies occur in the internal portions of the protoplasm. They are variable in size, and disappear in permanent preparations. Most of the smaller trophozoites are trinucleate. Small trophozoites are formed as a result of budding, with one or two buds being produced by trophozoites of intermediate size. Spores are conspicuous in the mature trophozoites. Only disporous forms have been seen. Spore Dimensions. Breadth, 7.9–11.2μ (9.3μ); height, 5.1–6.2μ (5.2μ); thickness, 4.4–6.4μ (5.3μ); valvular axes, 4.5–7.0μ (5.4μ); and 3.4–6.4μ (4.6μ); capsules, 1.1–2.1μ(1.6μ). Anterior angle, 70–124° (104°); posterior angle, 148–234° (207°); tangential angle, 134–180° (173°). Curvature index, 49°, taper index, 34°, valvular index, 103° Breadth:height ratios for quartiles of breadth range, 1.53, 1.65, 1.93 and 2.05. Spore Morphology. The stubby spore often has unequal valves, and is relatively variable in shape. The anterior margin is usually convex (Figs. 167–8), but in some spores is rather straight along the margin of each valve, meeting at a rounded angle at the suture (Fig. 164). The posterior margin varies from convex to markedly concave, with the majority of the spores having a flattened to very gently concave margin. The valves are essentially circular in cross section, and exhibit a considerable variation in valvular taper, some being markedly tapered and terminating in quite narrow tips (Figs. 164, 167) and some being quite stubby and terminating in very broadly rounded tips (Fig. 165). The valves are generally somewhat unequal, and in some spores are markedly so. They may be of similar or somewhat differing shape. In capsular view the spore axis is straight, and the capsules are nearly opposite. The polar capsules are small, nearly spherical, with very short necks. They lie at the anterior margin, through which they open. There is little tendency for the capsules to converge on the suture, or to be rotated toward the lateral margins. Capsulogenous nuclei are sometimes persistent, but are very small. The sporoplasm fills the cavity of the spore and extends upwards around the capsules. It is quite transparent, but usually contains a small number of small, refractive bodies, which do not persist in stained preparations. The two vesicular nuclei are small and variable in position.

Text-fig 13—Figs. 163–8 —Leptotheca minima n sp from Airipis trutta Fig. 163 Living trophozoite motile from P, U (e) Figs. 164–165—Sutural and capsular view of spores F. U. (d) Fig. 166—Stained spore S. A. H. (e) Figs. 167–168—Outlines of spores showing range of shapes F. U. (c) Figs. 168–174—Leptotheca annulata n sp Fig. 169—Sutural view of spore from Ihyrsites atun F. U. (d) Fig. 170–171—Stained spores in capsular and sutural views S. A. H. (c) Fig. 172—Capsular view of spore with doubled sutural ridge from Thyrsites atun F, U (d) Fig. 173–174—Outlines of spores from danidia solandri F, U (d) Figs. 175–178—Leptotheca pinguis n sp Figs. 175–176—Sutural and capsular views of spores from Peltoramplhus no, ae-zelandiae F, U (e) Fig. 177—Stained spore from Peltorhamphus S. A. II (f) Fig. 178—Stained spore from Caulopsetta scaphe S. V. H. (f)

Spore Variability. The spores are quite variable in shape, with a considerable range in angular measurements. The variability is rather like that seen in Ceratomyxa inconstans, involving considerable differences in the contour of the posterior margin of the spore. Despite the considerable variation in spore shape, the means observed in samples from two different hosts were quite similar. Differences between the means of the two samples were breadth, 0.3μ, height, 0.5μ, thickness, 0.9μ; valvular axes, 0.5μ and 0.5μ; and capsules, 0.0μ. Only a small sample of 10 spores were measured for thickness in each of the two hosts With small sample methods, the measurements for both height and thickness were significantly different, with a p. of about 0.05 Differences in the means of the angular measurements were: anterior angle, 6°; posterior angle, 9° and tangential angle, 2°. Discussion. No Leptotheca have been recorded as living in members of the Airipidae. While most of the Leptotheca have relatively small spores, this species is unusual in having spores less than 10μ in breadth. The only other species with spores more or less in the same size range are L. glomerosa (Davis, 1917) and L. renicola (Thélohan, 1895), both of which are urinary parasites. The globose form of L. renicola, approaching a spherical shape, is sufficient to distinguish the two forms at once. The spores of L. glomerosa are less convex anteriorly, and appear to lack the tendency to curve back, posteriorly. Moreover, the trophozoites have many fatty inclusions which are not seen in the forms from Arripis. Of the remaining species, L. fischeri (Jameson, 1929), originally described as a Ceratomyxa, is somewhat similar, and has spores which are somewhat larger than those seen in Arrpis (breadth, 9.3–13.3μ) but not necessarily significantly larger. It differs, however, in having monosporous trophozoites, and large polar capsules L. latesi and L. macronesi, described by Chakravarty (1943), also have rather small spores, although somewhat larger than those of L. minima L. latesi has capsulogenous cells which persist as membranous elements of the spore, and the capsules are larger than those of L. minima L. macronesi is monosporous, with elliptical spores. It seems probable that this species has closer affinities with Ceratomyxa minuta and C. inconstans than with other species of Leptotheca. It is significantly smaller than C. inconstans insofar as the spores are concerned, and while nearly the same size as C. minuta, has trophozoites which are without the persistent, conspicuous spherules seen in that species. Nevertheless, it seems probable that this is one of a cluster of small-spored species occurring in the mackerels and their allies, among the percomorph fishes. As it cannot be considered identical with any of the previously described forms, however, it has been designated Leptotheca minima n. sp. Leptotheca annulata n. sp. (Text-Fig. 13, Figs. 169–174) Habitat. This species occurs in the gall bladder of Thyrsites atun (Euphnasen), Jordanidia solandri (Cuv. and Val.), and Arripis trutta (Bloch and Schn.) It is not a common parasite in any of these species, and has been seen only in small numbers in each host fish. Infected fishes were taken by trawlers working out of Wellington in July and October. Trophic Stages. These are unknown. Spore Dimensions. Based on one small and one full sample from Thyrsites atun Breadth, 7.9–12.3μ (10.9μ); height, 4.5–6.7μ (5.8μ); thickness, 4.5–5.6μ (5.5μ); valvular axes, 3.9–6.7μ (5.8μ); and 3.3–5.6μ (5.2μ); capsular diameter, 1.7–2.8μ (2.3μ). Anterior angle, 71–125°. (107°); posterior angle, 154–245°. (208°); tangential angle, 158–180° (178°). Curvature index 45°. taper index, 30°; valvular index, 101° Spore Morphology. The small spore has rather cylindrical valves, often somewhat unequal, which meet in a very conspicuous, heavily ridged suture. In sutural view the anterior margin is convex, often only slightly so, and the posterior margin either convex, flattened, or concave, with the great majority of spores flattened (Fig. 169) The valves generally taper very little, being essentially cylindrical, with broadly curved tips. In capsular view the spore is usually straight, but is sometimes somewhat curved. The valves in a very conspicuous suture, which forms a heavy band around the spore. In a few spores a double band is seen(Fig. 172) The sutural prominence is deeply stained by Bismarck brown standing out in

sharp contrast to the nearly unstained spore membrance. The staining is sometimes interrupted in the region of the capsules. The polar capsules are nearly spherical in shape, with very short necks. Their position is variable in spores seen from capsular view, sometimes being opposite, and sometimes rotated far toward the lateral margins (Figs. 172, 174) The capsules are opaque, the filament being completely invisible. The capsulogenous nuclei are usually lost in mature spores. The sporoplasm completely fills the spore cavity, extending upwards about the capsules and between them. It is finely granular with an occasional refractive inclusion. There are two small, vesicular nuclei. Spore Variability. The spore population is variable in several ways, the most striking being the position of the capsules. The prominent suture may be single or double, with the possibility that the doubled suture is a juvenile trait. The range of spore shapes from oval, through flattened, to somewhat arcuate is, as with L. minima and C. inconstans, associated with considerable variation in the contour of the posterior margin of the spore. Unfortunately, infections with this species were always light, and it is not possible to make a careful analysis of the differences found in different hosts. Small samples from Jordanidia and Arripis included spores exhibiting about the same linear dimensions, and with about the same kinds of variations in shape. There were no evidences of peculiarities of the spores associated with particular host species. Discussion. Although there are several species of Leptotheca with small spores, the very heavy sutural ridge seen in this form distinguishes it from all others. In its small size, and general variability, it is reminiscent of the other small-spored species from the mackerels and their allies, but the suture sets it off from them. It is probably allied with this cluster of species, despite the prominent suture. Since none of the Leptotheca which have been previously described are identical with it, this species has been named Leptotheca annulata n. sp. Leptotheca pinguis n. sp. (Text-Fig. 13, Figs. 175–8) Habitat. This species occurs in the gall bladder of Peltorhamphus novaezeelandiae Gunther, and Caulopsetta scapha (Forster). Infected fishes were taken in July and August in the Napier and Wellington area. It was not seen in a number of Caulopsetta examined during the summer months. Trophic Stages. These have not been seen. Spore Dimensions. From Peltorhamphus novae-zeelandiae. Breadth, 13.7–18.6μ (16.4μ); height, 8.3–10.8μ (9.7μ) thickness, 8.3–9.8μ (9.2μ); valvular axes, 6.8–9.3μ (8.3μ) capsules, 2.4–3.9μ (2.9μ). Anterior angle, 90–118° (103°); posterior angle, 196–204° (218°); tangential angle, 150–180° (168°). Curvature index 39°; taper index, 50° valvular index, 115° Breadth: height ratios for quartiles of breadth range, 1.57, 1.73, 1.69 and 1.93. From Caulopsetta scapha, a small sample taken from two hosts. Breadth, 16.0–19.0μ(17.4μ); height, 10.2–11.0μ (10.9μ); thickness, 8.0–8.8μ (8.5μ); valvlar axes 9.8–11.9μ(10.3μ); and 6.4–9.4μ (8.3μ); capsules, 2.8–3.4μ (2.9μ). Anterior angel, 99–108° (104°); posterior angle, 201–237° (215°); tangential angle, 141–180°, (160°). Curvature index, 41°, taper index, 55°; valvular index 111°. Spore Morphology. The stubby, rather inflated spore has a strongly arched anterior margin, and a posterior margin which varies from somewhat flattened to somewhat concave. There is little or no valvular taper, the tips of the valves being very broadly rounded. The stubby and inflated nature of the valves is shown by the high taper and valvular indices (50–55°and 111–115°). On the average the posterior margin is somewhat concave, with the tangential angle between 160° and 168°. In capsular view the spore is a plump oval, often somewhat curved. The two valves are equal in the spores from Peltorhamphus, but somewhat unequal in the spores form Caulopsetta. They have a similar shape in both species. They meet in a distinct, elevated, narrow suture. The polar capsules are oval to somewhat pyriform, and are set on the anterior margin of the spore, through which they open. They converge somewhat on the suture. The filament is indistinctly visible or distinct. It forms a coil about 0.5μ smaller than the capsule, with three to six turns. The longest extruded filament measured 25μ, and was not perfectly straight. In younger spores the capsulogenous cells are distinct (Fig. 177) They become highly vacuolated as the valvular nuclei at the valvular extremities become indistinct, and eventually become membranous remnants. The membranous structure persists for a time, but eventually disappears. The capsulogenous nuclei are small, and often do not persist. The sporoplasm does not fill the spore cavity. It is usually central in position. The sporoplasm is rather granular and dense, containing various regions of differing optical

density. A few refractive inclusions of variable size are usually present, either in the sporoplasm or in the spore cavity. The two sporoplasmic nuclei are usually placed close together. Spore Variability. Except for minor varitions in the linear dimensions of the spores, they vary primarily with respect to the contour of the posterior margin, which varies from somewhat convex to somewhat concave. The spores from Caulapsetta, although not present in such large numbers, appeared to be somewhat more variable, particularly in capsular size and length of valvular axes. Unfortunately the spores were never found in large numbers in Caulopsetta, and a statistical comparison of the species in the two hosts is not possible. The spores from Caulopsetta were somewhat greater in breadth and thickness, and somewhat smaller in thickness, but these differences were well within the limits that might be expected in samples from hosts of the same species. The angular measurements and indices were very similar throughout. The polar capsules were somewhat more variable in size in the Caulopsetta material, in many instances being almost spherical in shape (Fig. 178), and the valvular axes, which were equal in the Peltorhamphus material, were usually unequal in spores from Caulopsetta. Despite these differences, in view of the similarity of the hosts and the overall similarity in the spores from the two hosts it seems probable that the two strains are identical. The study of larger quantities of material from Caulopsetta is needed, however. Discussion. The flat-fishes are very receptive to parasites belonging to the Ceratomyxidae, and in general there seems to be but little host specificity within the flat-fish group, the same species occurring in a number of different flat-fishes (see Fujita, 1923, for example). Four species of Leptotheca have been reported from flat-fishes, L. glomerosa, L. lobosa, L. platichthytis and L. limandae, the first two by Davis (1917) and the last two by Fujita (1923) L. glomerosa is a smallspored, urinary species, with essentially cylindrical spores. L. lobosa, also occurring in the urinary system, has spores about the same size as L. pinguis, but differs in having the two valves dissimilar, a sinuous sutural line, and spores tending to remain united at the suture L. limandae, with much larger spores, is characterized by spores with unequal valves having dissimilar valvular tips. L. platichthytis is characterized by furcate spores. It is evident that this species is not identical with any of those previously reported from flat-fishes. Of the other species of Leptotheca, five are about the same size. These are L. polymorphus (Thélohan) Labbe, 1899; L. informis Auerbach, 1910; L. fusiformis Davis, 1917, L. constricta Fujita, 1923; and L. elegans Noble, 1938. The tendency Text-fig 11 Fig. 179 Leptotheca sp from Caulopsetta ceapha S. A. C. (f) Fig. 180—Leptotheca sp ? from Caulopsetta seapha Fresh spore I. N. R. (f) Fig. 181—Leptotheca sp Congropodus leucopacctlis Fresh spore I. I. (f)

of L. polymorphus to form aggregations of trophozoites and the kinds of variations seen in the form of the vegetative individuals serve to distinguish it from L. pinguis. L. informis is, probably, the most similar of all of the previously described species. It appears to differ, other than in having somewhat larger spores, in having spores which are somewhat more curved and which are more uniformly curved, without the occurrence of the straight individuals so often seen in L. pinguis L. elegans differs in having lateral swellings on the valves, and an indistinct suture. The larger polar capsules and oblique suture of L. fusiformis distinguishes that species from L. pinguis. The spores of L. contricta have a relatively straight anterior margin, and unequal valves and capsules opening at some distance from the anterior margin. These differences appear to be sufficient to warrant considering the form found in Peltorhamphus and Caulopsetta as unique, and it has accordingly been designated Leptotheca pinguis n. sp. Unidentified Leptotheca In addition to the three species described above, several fishes have been seen in which such small numbers of Leptotheca were present that it was not possible to completely characterize the spore and identification was, therefore, impossible. Leptotheca sp.? from Congropodus leucopaecilis (Fig. 180) A few spores of a small, nearly spherical Leptotheca have been seen in the gall bladder of a Congiopodus taken in the Wellington area in January. A typical spore is shown in Fig. 180. The dimensions of a fresh spore were, Breadth, 6.9μ, height, 6.9μ; valvular axes, 4.0 and 2.9μ, capsules, 1.8 by 1.4μ. Anterior angle, 88°; posterior angle, 272°; tangential angle 180°. The spore is almost spherical in shape, with two unequal valves, meeting in a straight, slightly elevated, narrow suture. The polar capsules are spherical, and placed near the anterior margin. They are equal. The filament cannot be seen in fresh spores. The large sporoplasm fills the spore cavity posterior to the capsules. The shape of the spore is such that it might well fall into the genus Sphaerospora. The position of the polar capsules, however, is like that seen in Leptotheca, and the inequality of the spore valves is also typical of the Leptotheca rather than the Sphaerospora. So few spores have been seen that it is not yet known whether the mean spore breadth will eventually prove to be somewhat in excess of the spore height. Its proper taxonomic position will await the examination of more material, and the discovery of the vegetative form. Leptotheca sp.? from Caulopsetta scapha (Fig. 178) A few spores of a Leptotheca somewhat larger than those of L. pinguis have been seen in the gall bladder of two Caulopsetta taken in August. Fresh spores measure about 22–23μ in breadth, and 15μ in height, falling somewhat outside of the size range of L. pinguis. As they have been seen in one host in which no L. pinguis was present, it is probable that they are not members of that species. In most respects they resemble L. pinguis, but have a more oval form, with the posterior margin somewhat flattened. No spores with a concave posterior margin have been seen, but as only a few were present, it is not possible to state the range of variability typical of this form. Leptotheca sp. ? from Caulopsetta scapha. A few spores of a Leptotheca have been seen in the urinary bladder of a single Caulopsetta scapha taken in January. The average dimensions of two spores were: breadth 18.9μ; height, 12.7μ, capsules 4.7μ by 3.6μ. Anterior angle, 115°; posterior angle, 243°; tangential angle, 180° The spores measured were somewhat immature, with remnants of the capsulogenous cells still visible. It is not known whether they retain permanently any juvenilc traits. The two approximately equal valves meet in a narrow, rather inconspicuous suture. They are short, stubby, and have a nearly straight axis. The rather large

polar capsules contain a filament which is easily seen in fresh spores, and is arranged in seven to eight coils. The rounded, binucleate sporoplasm does not fill the spore cavity. Bibliography Auerbach, M, 1910. Die Cnidosporidien (Myxosporidien, Actinosporidien, Microsporidien). Eine monographische Studie. Leipzig. Awerinzew, S., 1907. Ueber Myxosporidien aus Gallenblase der Fische. Zool. Anz., 31. 831–834. — 1913. Ergebnisse der Untersuchungen uber parasitische Protozoen der tropischen Region Afrikas. III. Zool. Anz. 42:151–6. Baker, J. R., 1958. Principles of Biological Microtechnique. Methuen, London. Chakravarty, M, 1943. Studies on Myxosporidia from the common food fishes of Bengal. Proc. Ind. Acad. Sci., B 18:21–35. Cunha, A. M. da, and Fonseca, O. da, 1918. Sobre os Myxosporidies dos piexes do Brazil. Brazil Medico 32:393, 401, 414. Davis, H. S., 1917. The Myxospordia of the Beaufort Region: A systematic and biologic study. Bull. Bur Fish., 35:201–243. Doflein, F., 1898. Studien zur Naturgeschichte der Protozoen. III. Ueber Myxosporidien. Zool. Jahrb., Anat., 11:281–350. Dunkerly, J. S., 1921. Fish Myxosporidia from Plymouth. Parasit. 12:328–333. Fantham, H. B., 1930 Some parasitic Protozoa found in South Africa. 13. South Afri. Jour. Sci. 27. 376–390. — Porter, A., and Richardson, L. R., 1940. Some more Myxosporidia observed in Canadian fishes. Parasit 32:333–353. Fujita, T, 1923. Studies on Myxosporidia of Japan. Jour. Coll. Agr. Hokkaido Imp. Univ. 10:191–248. Georgevitch, J., 1916. Note sur les Myxosporidies des poissons de la baie de Villefranche et de Monaco. Bull. l'inst. ocean, No. 433. — 1929. Recherches sur Ceratomyxa maenae nov. sp. Arch. f Protistenk. 65:106–123. Gray, P., 1958. Handbook of Basic Microtechnique. McGraw-Hill, New York. Jameson, A. P., 1929. Myxosporidia from California Fishes. Jour. Parasit. 16:59–68. — 1931. Notes on California Myxosporidia Jour. Parasit. 18:59–68. Jordan, D. S., 1923. A classification of fishes, including families and genera as far as known. Stanford U. Publ. Biol. Sci., 3:79–243. Kudo, R. R., 1920. Studies on Myxosporidia. A synopsis of genera and species of Myxosporidia. III. Biol. Monogr. 5 (3, 4):1–265. — 1921. On the effect of some fixatives upon Myxosporidian spores. Trans. Am. Micr. Soc. 40:161–7. — 1933. A taxonomic consideration of Myxosporidia. Trans. Am Micr. Soc. 40. 161–7. — 1947. Protozoology. Thomas, Springfield, Ill. Labbe A. 1899. Sporozoa in Das Teirreich. 5 Lief. Laird, M., 1953. The Protozoa of New Zealand Intertidal Zone Fishes. Trans. R. Soc. N. Z., 81:79–113. Mavor, J. W. 1916. On the life-history of Ceratomyxa acadiensis, a new species of Myxosporidia from the eastern coast of Canada. Proc. Amer. Acad Arts and Sci., 51. 551–574. Meglitsch, P. A., 1952. The Myxosporidian fauna of some fresh water and marine fishes. Proc. Iowa. Acad. Sci., 59:480–486. Noble. E. E., 1938 Two new Myxosporidia from tide pool fishes of California. Jour. Parasit. 24:441–2. — 1939. Myxosporidia from tide pool fishes of California. Jour Parasit. 25:359–363.

Noble, E. E., 1941. Nuclear cycles in the life history of the protozoan genus Ceratomyxa. Jour. Morph., 69:455–476. Ray, H., 1933. Preliminary observations on Myxosporidia from India. Curr. Sci. Bangalore. 1. 349–350. Setna, S. A., 1942. Preliminary observations on Myxosporidia from sharks. Curr. Sci. 11:469–470. Thelohan, P., 1892. Observation sur les myxosporidies et essai de classification do ces organismes. Bull. soc. philom., 4:165–178. —1895. Recherches sur les Myxosporidies. Bull. sc. France et Belg., 26:100–394. Professor Paul A. Meglitsch, Ph. d, Biology Department, Drake University, Des Moines, Iowa, U. S. A.