NEUR0SP0RID1UM. 81

On Neurosporidium cephalodisci, n.g., n.sp.,a Sporozoon from the Nervous System of

Cephalodiscus nigrescens.

ByW. Cr. Ridcwood, O.Sc,

Lecturer in Biology at St. Mary's Medical School, University of Loudon ;

JI. B. Fantliain, B.Sr., A.R.C.S.,University College, London; Demonstrator in Biology at St. Mary's

Medical School.

With Plates 6 and 7-

INTRODUCTION.

The organism described in this paper occurs in thenervous layer of the ectoderm of Cephalodiscus (Idio-thecia) nigrescens, Lank.j a large and massive form ofCephalodiscus dredged by the " Discovery " on January13th, 1902, in 100 fathoms, off Coulmann Island, nearVictoria Land, in the Antarctic Ocean.

The specific name nigrescens was given fco this Cephalo-discus by Prof. Ray Lankester,1 the polypides being to thenaked eye of a sooty black colour; a detailed description ofthe polypides and the tubarium or "house" of Cephalo-discus nigrescens is given in the " Reports of the'Dis-covery ' Expedition," published by the British Museum.

The Sporozoon that occurs in Cephalodiscus nigres-1 'Proc. Roy. Soc.,' 1905, pp. 400—402.

VOL. 5 1 , PAET 1. NEW SERIES. 6

82 AV. G. KIDEWOOD AND H. B. FANTHAM.

cens is found in relation with the nervous, deeper layer ofthe ectoderm. It does not always occur actually within thatlayer, since it may bulge outwards into the more superficialcells of the ectoderm, or may project inwards into the under-lying tissues, but the new orgauisms seem always to besituated in the nerve tissue before they begin to enlarge andsporulate. On account of this peculiarity we apply to theorganism the generic name Neurospor id ium. Cephalo-d i scus being the host in which the parasite has beenencountered, we base the specific name upon that. Theorganism is thus denominated Neurospo r id ium cephalo-disci.

No unicellular parasites of Cephalodiscus have up tothe present been described. In Cephalodiscus g i l ch r i s t ithere is a Copepod Crustacean occurring parasiticnlly in thestomach of a large proportion of the polypides 1; this parasiteis closely allied to the Copepods that occur in the gut ofTunicates, and belong to the family Ascidicolidas ofGiesbrecht.

Masterman3 has drawn attention to "cyst-like structures"occurring near the anus of Cephalodiscus dodecalophus,which he is disposed to regard as larvae of Cephalodiscus,but which from their position are more likely to be parasites.They are of comparatively large size, and judging by thelarge size of the cells as drawn in the figure, they are moreprobably the young of some metazoon parasite than cysts ofSporozoa.

M'Intosh3 marks by the symbol bp. a series of three spheri-cal bodies occurring in the deeper part of the nervous layerof the ectoderm, which might at first glance be taken forSporozoa. Their position is one commonly occupied by theparasite which forms the subject of the present paper, andtheir size and general shape cari'y the resemblance with

1 Ridewood, W. G., "Cephalodiscus gilchristi," 'Marine Invest.South Africa,' iv, 1906, p. 181.

3 ' Trans. Roy. Soc. Edinb.,' 1898, p. 513, last paragraph ; also pi. 5, fig. 86.3 ' "Challenger" Reports,' part 62, vol. xx, 1887, pi. 6, fig. 3.

NEUROSPORIDJUM. 83

Neurosporidiuin still further; but it is to be noticed thatthe central, uninucleate cells that one would take to be thespores, are enclosed within a wall composed of a single layerof regular cells, and not by a structureless cyst or capsule.In the explanation of the figure, the parts marked bp. are saidto be "sections apparently of the shield pores in their pro-gress outwards," but it is highly improbable that the probosciscanals could, in a section taken at right angles to the buccalshield, be so cut as to present a circular outline, and eventhen the presence of the third body is difficult to explain, forthe proboscis canals are short, and with firm walls of closely-set epithelial cells, and could not, even in the most contortedpolypides, be cut across twice in the same section.

In view of the close relationship which is admitted to existbetween Balanoglossus and Gephalodiscus it is worthyof mention in this connection that Spengel' has found in theccelom of a species of Balanoglossus (Ptychoderaminnta) masses of small, uniform, nucleated cells which heis inclined to regard as of a parasitic nature. Caullery andMesnil2 agree with Spengel that these bodies are parasiticorganisms, and they refer them provisionally to the neworder of the class Sporozoa to which they have given thename Haplosporidia, the order in which we propose to placethe JSTeurosporidium of Cephalodiscus.

MATERIAL AND METHODS.

The specimens of C e p h a l o d i s c u s n i g r e s c e n s obtainedby the " Discovery " were fixed, some in a 5 per cent, solutionof formalin, some in Perenyi's fluid, and some in picric acidsolution. Serial sections of the polypides were cut for the

1 'Fauna and Elora des Golfes von Neapel,' Monogr. 18, 1893, pp. 661,662; pi. 2, figs. 19, 20; pi. 3, figs. 50, 61; pi. 4, figs. 60, 61, 76, 79, 80;pi. 5, fig. 105.

' "Haplosporidies," 'Arch, de Zool. Exp. et G6u.,' iv, 3, 1905, p. 164, andpi. 13, fig. 125.

84 W. G. RIDBWOOD AND H. B. EANTHAM.

pui'pose of investigating the anatomical structure of this newspecies of Cephalodiscus , and these sections, and a fewadditional ones cut more recently, were utilised for the studyof the Sporozoou. The sections were 5ju, 6 n, and 7'5,u inthickness. The majority of them were stained with Ehrlich'shsematoxylin and eosin, the others with hEematoxylin andorange G, or Mayer's hsemalum, or borax carmine. Thestructural details of the parasite were most satisfactorilyshown by the sections that were cut from material fixed in 5per cent, formalin and stained with hasmatoxylin and orangeG, or hasmatoxylin and eosin. The parasites were studiedwith a Zeiss 3 mm. apochromatic homogeneous immersionobjective, used in combination with compensating oculars 4and 8, and occasionally 12. In some of the stages they areoE large size (pi. 6, figs. 1 and 2, Nsp.), and can be readilyrecognised with a -| inch objective.

OCCDREISNCE OF THE PARASITE.

The infection appears always to commence in the nervoustissue, and although the parasites when at their largest mayproject beyond the nerve layer into other tissues (pi. 6,fig. 1, Nsp.), they remain in relation with nerve tissue at onepart of their surface. The parts of the nervous system inwhich the parasite has been found are the central nerve mass(pi. 6, fig. 2, iVsp.), the nerve layer of the dorsal wall ofthe buccal shield (practically a continuation of the centraltract), the nerve layer of the ventral wall of the shield(pi. 6, fig. 1, Nsp.), and the lateral nerve tracts near thecollar pores. The parasite has not been found in the nervetracts of the plumes, nor in those of the stolon.

The occurrence of N e u r o s p q r i d i u m in the nerve tractsof Cepha lod i scus is not the first instance of a Sporozooninfecting nerve tissues, although infection of such tissues isnot common. Three instances are known of Neosporidia

NECTR0SP0RTD1UM. 8 5 .

infecting the nervous system of Teleostean fishes. Pfeiffer1

in 1892 discovered a Myxosporidian in the cranial and spinalnerves of the Grayling (Thymallus vulgaris), the cystalying between the medullary sheath of the nerves and thesheath of Schwann; Schuberg and Schroder3 in 1905described a similar, but not identical Myxosporidian, whichthey named Myxobolus neurobius, in the nerves of theBrook Trout (Salmo fario); and in 1898 Doflein3 describeda Microsporidian, Glugea lophii (now known as Nosemalophii), occurring in the ganglion cells of the central nervoussystem of the Angler (Lophius piscatorius).

The proportion of infected specimens of C. nigrescens isconsiderable. Out of a total of twenty polypides examined,Neurosporidium was found to occur in eight.

Spherical or slightly oval spores of Neurosporidiumappear to set up a local degeneration in the nervous, deeperlayer of the external epithelium of Cephalodiscns, and inthe oval spaces so formed the parasites grow and sporulate,and become surrounded by an ill-defined, somewhat irre-gularly alveolar capsule (pi. 7, figs. 9, 12, 15, cps.). Thiscapsule is not of the nature of an ordinary cyst, secreted bythe parasite peripherally, but would appear rather to be formedby the surrounding tissue of the host, much in the sameway as is the enveloping membrane of the Sarcosporidia.

The capsule is not in all cases clearly evident. Some ofthe parasites, appearing as groups of spherical bodies whenexamined under a Zeiss A or D objective, nearly fill thespaces in the host tissue in which they lie, while others aremore or less closely surrounded by a capsule (pi. 7, fig. 12,cps.), which itself is separated from the wall of the host tissueby a space. The space (fig. 12, cav.) is bounded externallyby a thin layer of the same material (fig. 12, cps1.) lining the

1 Pfeiffer, L., ' Untersuchungeu iiber den Krebs; die Zellerkrankungenund die Geschwulstbildungen durch Sporozoen,' 1893, p. 75.' a Schuberg, A., and Schroder, 0., 'Archiv fur Protistenkunde,' iv, 1, 1905,pp. 50 et seq.

3 Doflein, P., 'Zoo!. Jalirb.,' Anat., xi, 1898, p. 332.

86 W. G. BIDEWOOD AND H. B. FANTHAM.

host tissue. The capsule and this latter layer are refringent,and yellowish-brown in colour; they seem to be chitinous,but special tests were not applied.

These chitinoid layers are not always present, and evenwhen they are to be recognised, they are so indefinite andirregular that we pi-efer not to use the word cyst at all, butto employ a more indefinite term, such as capsule. Probably,the layers in question are produced by the host tissue in anattempt, unsuccessful as it appears, to isolate the parasitefrom organic connection with itself, or else they consist ofbroken-down nerve tissue which has taken a more or lessspherical shape in accommodating itself to the space betweenthe parasite and the still healthy part of the host tissue.

In some of the cavities a coagulum is to be noticed sur-rounding the parasite (fig. 7, coag., also figs. 5 and 6). Thecoagulum is probably directly connected in its origin with thedegeneration of the host cells, and the production of the

STRUCTURE AND LIFE-CYCLE OF NEDBOSPOEIDIOM.

The pai'asites in their smallest phase occur as little roundor oval cells (free spores or amsebulse), about 3ju (2 to 4,/u) indiameter (see fig. 3), each with a single nucleus. These arefound lying in or among the nerve cells of the host, havingapparently entered this tissue by infiltration. It is notpossible, with the material at our disposal, to say definitelyif there is an intracellular phase of the parasite within anervous epithelial cell at the beginning of the life-history ofthe parasite. Nuclear division takes place in the amasbula,and young trophozoites, with two or three nuclei in a some-what irregular mass of protoplasm (fig. 3, b, c, d) are ofcommon occurrence, often lying in a space among thenervous tissue cells of the host. Figures of nuclear division(fig. 3, c) are very rarely and imperfectly seen in our material.

The young trophozoites are usually about 5 fi in diameter.The protoplasm of the trophozoite at this stage is granular

NEPROSPORtDIUM. 87

and easily stained, while in the one-cell or oval corpusclestage the protoplasm is hyaline and only slightly granular,and does not stain deeply. Several young trophozoites mayoccur in one space (fig. 6).

Further nuclear division rapidly takes place, and multi-nucleate trophozoites, spherical, ovoid, or cylindrical in shape,are found with their many nuclei embedded in a more or lessrounded or slightly lobulated mass of deeply-staining granularprotoplasm (figs. 8 aud 9). The multinucleate trophozoitescorrespond with the "plasmodium" stage of the Haplosporidiaof Oaullery and Mesnil,1 though they are not markedly irregu-lar- or amseboid in outline. These trophozoites are 30 /x to40 /x long, and 25 fi to 35 fx broad.

A " plasmodium " may be formed by the nuclear divisionand growth of the oval corpuscle or amaebula that constitutesthe earliest stage of the parasite. Tt may also be formed bythe fusion and growth of several young trophozoites (fig. 7),as can be seen by careful examination of a series of sectionspassing through a cavity containing several parasites at thisstage. Such a fusion or concrescence approximates veryclosely to the formation of a true plasmodium.

In the multinucleate trophozoite phase the protoplasm ofthe parasite is not surrounded by a cuticle, and N e u r o -s p o r i d i u m hereby differs from such a type as B e r t r a m i a ,which it otherwise much resembles at this stage. The proto-plasm is opaque, and crowded with closely distributed granules.Around each nucleus, however, is usually an area of clearprotoplasm (fig. 8). Some of the nuclei are rather large, 1 fxto 1'5 [X, or even 2fx in diameter, but no definite nucleolus orkaryosome is visible ; the chromatin of these nuclei seems tobe evenly distributed. No refringent granules such as occurin B e r t r a m i a a s p e r o s p o r a 8 were noticed in the deeply-staining protoplasm. I t is to be noted, however, that while

i Caullery, M., and Mesnil, ]?., " Recherches sur les Haplosporidies,"'Arch. Zool. Exp. et Gen.,' iv, 1905, pp. 101—181, three plates.

8 Minchin, E. A., "Spoiozoa," Lankester's 'Treatise on Zoology,' part 1,fasc. 2, 1903, p. 3J0.

88 W. G. RIDEWOOD AND H. B. FANTHAM.

our Neurospo r id ium consists of preserved material only,B e r t r a m i a was described from living specimens.1

In the larger and older rnultioucleate trophozoites thenuclei and their surrounding zones o£ clear cytoplasm aredistinctly seen (fig. 9), marking the beginning of the segrega-tion of the plasmodial mass into pansporoblasts.

The term pausporoblast was first employed by Grurley 2 inconnection with the Myxosporidia (in the wider sense, includ-ing Microsporidia). A pansporoblast ('sphere primitive' ofThelohan) may be described as a spore-inother-cell differenti-ated within the substance of the trophozoite, destined to giverise to sporoblasts, from which spores are developed. (Theterm sporoblast is applied to cells which develop each into asingle spore, and the term pausporoblast is applied to a cellwhich by division gives rise to sporoblasts.) In the Myxo-sporidia the differentiation takes place by the concentrationof cytoplasm around one of the nuclei of the endoplasm, andas a rule the protoplasm becomes more transparent. Thepansporoblasts of a trophozoite may be many or few, or evena single one. A greater or less amount of the protoplasm ofthe trophozoite usually remains over, together with a numberof residuary nuclei of the endoplasm, although in someMicrosporidia, e.g., The lohau ia and P le i s tophora , thewhole trophozoite becomes converted into a single pansporo-blast. In the typical Myxosporidia each pansporoblast givesrise ultimately to two spores; in the Microsporidia, on theother hand, each pansporoblast may give rise to four, eight,or a larger number of spores.3

The term pansporoblast may be extended in its applica-tion to the Haplosporidia, for although in B e r t r a m i a the

1 See also Warren, E., "On Bertramia kirkmani," 'Annals Natal Gov.Mus.,' vol. i, London, 1906, pp. 7—17.

3 Gurley, R. R., " On the Classification of the Myxosporidia," ' Bull. U. S.Fish. Comm. Rept. for 1891,' xi (1893), pp. 407—420: see his footnote onp. 40S.

3 Except perhaps in Nosema pulvis. See Perez, Ch.,. "Microsporidiesparasites des ciabes d'Aicachon," ' Soc. Sci. d'ArcacliOn, Stat. Biol./ann.viii, Paris, 1905, pp. 15—36.

NETJROSPORIDIOM. 89

trophozoite divides into a number of bodies, each of whichbecomes a single spore, yet in the case of H a p l o s p o r i -dium scolopl i and H. n iarchouxi * the bodies into whichthe trophozoite divides undergo each a division into fourspores.2 The " bodies " in question are thus pansporoblastsin the sense described above.

In Neurospor id ium a definite segregation of the cyto-plasm of the trophozoite around the nuclei marks the begin-ning of the formation of pansporoblasts (figs. 7, 10, and 1]).The general shape of the trophozoite at this stage is ovoid(fig. 11), or sometimes nearly sausage-like (fig. 10). The pan-sporoblasts are uniuucleate cells with hyaline protoplasm,lying embedded in the opaque, granular protoplasm of thetrophozoite. They are irregularly, but fairly closely packedin this granular mass, sometimes slightly more crowdedtogether at the periphery of the trophozoite than in thecentre (fig. 10). They are spherical or oval in shape, aud 3/xor 4 fi in diameter.

The pansporoblasts in some trophozoites are very quicklyformed, the phase of nuclear division being rapidly passedthrough, or even omitted, since this stage may be reached,in true " plasmodial" manner, by the fusion of severalyoung trophozoites, all lying within a common cavity in thehost-tissue, as already noted. By careful examination ofseries of sections, what appears to be the fusion of suchmasses of young trophozoites can be seen, and a stage in theprocess is shown in fig. 7.

The cavity in the host-tissue enclosing the parasite in-creases in size as the growth of the parasite proceeds; in allprobability the parasite causes progressive degeneration ofthe neighbouring host tissue.

Up to this point the development of N e u r o s p o r i d i u mhas been simple, in many respects closely following that of atypical Haplosporidian, such as Be r t r amia . The mode of

1 Caullery and Mesnil, loc. cit., p. 114 and p. 117.2 In Haplospoiidia there is no visible distinction between sporoblast and

spore; the former passes directly into the latter.

90 W. G. RIDEWOOD AND H. B. FANTHAM,

sporulation, however, is distinctive, and of considerable im-portance in determining the systematic position of the para-site.

All the pansporoblasts in a trophozoite of Neurospor i -dium are formed, by internal segmentation, at the sametime. They then commence to increase in size, growing atthe expense of the protoplasmic ground mass of the tropho-zoite. Each pansporoblast becomes less well-marked in out-line ; its protoplasm is still rather homogeneous (non-granu-lar), but deeply staining, and its centrally-placed nucleusdivides and discharges chromidia throughout the cytoplasmof the pansporoblast. Each pansporoblast thus becomes aspore-morula, for the division of its nucleus into groups ofchromidia, or small portions of chromatin (daughter-nuclei),results in the formation of many small unicellular sporoblasts,distributed not only peripherally, but throughout the sub-stance of the pansporoblast, as may be seen by carefulfocussing.

The division of the nucleus of the pansporoblast is not ofthe serial karyokinetic type, first into two, then into four, andso on, but is a kind of simultaneous " multiple fission " of thereproductive chromatin into groups of chromatin forming thenuclei of the many sporoblasts, each of which soon becomesa spore with little or no further differentiation. The vegeta-tive chromatin of the nucleus of the pansporoblast remainsin the centre of the spore-morula; it is a pale-staining body,and cannot be recognised until the spores have scattered(see fig. 14, r.n.).

Each spore is about 1 /j. or 2 n in diameter, and stainsdeeply; its small nucleus is surrounded by hyaline proto-plasm. From a full-grown trophozoite a very large numberof spores is formed, since the trophozoite gives rise to manypansporoblasts, and each pansporoblast grows and dividesinto many spores. A full-grown trophozoite containingspore-morulEe in sporulation is more or less spherical, andmeasures about 40 ju to 70^ in diameter (figs. 12 and 13).The spore-morulse are spherical, and 10 fi to 15 ju in diameter.

NBUR0SP0K1DIUM. 91

The uuinucleate spores, which are small gymnospores orameebulEe, gradually pass out of the cavity around the parentorganism by more or less irregular apertures in the capsule,and so invade the surrounding tissue of the host, and startfresh infections in other parts of the nervous system of theCephalodiscus.

Each spore on becoming free grows to abont 2, 3, or 4 fiin diameter, and division of its nucleus begins, while itscytoplasm becomes granular and a little more opaque. Inthis way young trophozoites are formed, such as those de-scribed in the first paragraph of this section.

All the cytoplasm of the spore-morula is not used up in theformation of spores, and there is also nuclear material leftover. These residuary masses are clearly shown in our pre-parations, and their fate may here be considered in somedetail. When the spores have left the parent spore-morula,the latter has a fairly definite contour, rather moi'e distinctthan during sporulation. The protoplasm is granular, andthere is usually a pale, centrally-placed nucleus. The proto-plasm stains feebly, and the granules in it are distinct. Theresidual nucleus is pale in colour, and reddish or brownish-red after staining with hsematoxylin, and it thus stands inmarked contrast with the deeply purple nuclei of the spores inthe same prepai-ation. In a few spore-morulse examined littleor no residual nuclear matter was found, even after theexamination of all the sections into which they had beencut.

One or two rather more deeply staining dots may be seenin the pale residual nucleus in some, rather rare, cases(fig. 15). The residual nuclear matter is, as already stated,the remains of the vegetative chromatin of the spore-mornla,the reproductive or generative chromatin being distributedin the nuclei of the spores. Deeply staining granules ofchromatin are seen in capsules at this stage, sometimes intwo or three groups (fig. 15, chr.), but usually aggregatedtowards the centre of the mass (fig. 17, chr.). These aregroups of chromidia, or small masses of chromatin, of about

92 w. a. BID*;WOOD AND H. B. PANTHAM.

half the diametei' of the nuclei of the spores. Some of themare surrounded by a clear area of cytoplasm, and may formspores, bat many would appear to be residual, and notdestined to give rise to spores.

The actual transition between the stage of sporulation iua spore-morula and that of residual matter is depicted infig. 14. The capsule, of which a small part only is shown, isfull of ripe spore morulee, and these are in process of divi-sion into uninucleate spores; one of them, however, isslightly in advance of the rest, and has already divided intospores, which have scattered, leaving the remains of the spore-morula, pale in colour, with granular protoplasm, and remainsof nucleus. The outline of this relic (fig. 14, r.sp. m.) in thecapsule is more clearly seen than that of its neighbours; itsprotoplasm is of a light brownish tint, and its nuclear residueof a brownish pink, after staining with haematoxylin, whichimparts a characteristic purple coloration to the neighbour-ing sporulating masses, forming a clear and striking con-trast.

The remains of spore-morulse, after sporulation, undergodegeneration in old capsules (fig. 16). Their outlines becomeless distinct, and vacuoles appear in the now coalescing massof protoplasm (fig. 17). The residual nuclei become lessclear, but chromidia may still be seen near the centre of thedegenerating mass (fig. 17). Some of the spores, formedfrom the spore-morulae before they degenerated, may for atime be seen in the neighbouring tissue and in the cavity inwhich the parent capsule lies (fig. 16).

While the above account of the life-history of Neuro-spor id inm suggests how the infection may spread from onepart of the host to another, by the amsebulas and youngtrophozoites migrating along the course of the nerve tracts,we have no evidence to offer which can explain the spreadof the parasite to new hosts. In B e r t r a m i a aspero-spora , of Rotifers, the mode of spread has been watched byBertram; the spores have no power of independent move-ment, but are disseminated passively on the death and disin-

NEUBOSPOETDIDM. 93

tegrat ion of the body of the host.1 Possibly the same may

happen in the present instance.

SYSTEMATIC POSITION OF N E U K O S P O K I D I D M .

I n determining the systematic position oE N e u r o s p o -

r id ium cepha lod i sc i the following features in the life-histoi-y, so far as we have been able to trace it, should bekept iu view:

1. From a capsule lying in a cavity in the nervous systemo£ Cephalodiscus gymnospores or amEebulse are liberated,as rounded masses o£ clear, naked protoplasm, 2 to 4 ju indiameter, each with a centrally placed nucleus.

2. The next stage is one in which a small mulbinucleatetrophozoite lies in a small cavity in the nervous system.This stage is reached, either by the rapid growth and nucleardivision of a single amsebula, or by the coalescence of severalamssbulae (plasmodium).

3. Nuclear division of the trophozoite continues, and thesize increases, uutil the body is about 30 to 50 fi in diameter,and ovoid in shape. The general protoplasm is granular andopaque, but there is a clear zone of protoplasm around manyof the large nuclei.

4. The large ovoid trophozoite segments into pansporo-blasts, each a single cell, 3 to 5 /j. in diameter, consisting of alarge nucleus surrounded by clear cytoplasm.

5. The nucleus of each panspoi'oblast divides into manydaughter-nuclei, and the pansporoblast enlarges and becomesa spore-morula, 10 to 15 ju in diameter. Each daughter-nucleus, with its small mass of clear cytoplasm, becomes asporoblastj and then a spore. The spores are generally dis-tributed throughout the spore-morulse.

6. The spores ultimately separate, and pass out into theadjacent parts of the nervous system of the host. The remainsof the spore-morulas, consisting of granular protoplasm, pale-staining remains of.nuclei, and some free chromidia, undergogradual degeneration.

1 Caullery and Mesnil, loc. cifc., p. 136.

94 W. G. KIDEWOOD AND H. B. FANTHAM.

7. Around the parasite in its various stages (except theamEebula stage) is an ill-defined capsule, partly alveolar, andprobably secreted by the cells of the host.

The systematic position of Nenrosporidiu m may bereadily determined by reference to the characters summarisedabove. The somewhat irregular form of the trophozoite andits possession of several nuclei; the appearance of pansporo-blasts, denoting the commencement of spore-formation, at anearly stage in the growth of the parasite; the division of thepansporoblast to form many spores; and the intercellular(histozoic) habitat of the parasite; all these are characteristicof the Neosporidia.

Further, on account of the small, simple uninucleate spores,without polar capsules, the increase in the number of nucleiduring the trophic stage, and the segmentation of the full-grown trophozoite into ovoid pansporoblasts, we place theparasite among the Haplosporidia.

The division of the pansporoblast into many spores is afeature which does not occur in the life-history of the typicalHaplosporidian. In Haplospoi'idium scolopli and H.marchouxi the number of spores formed from each pan-sporoblast is four; in Bertramia it is.one. In Rhino-sporidium kinealyi, however, a parasite from the nasalmucous membrane of man, recently described by one of us(Fantharn)1 in collaboration with Prof. Minchin, and also byBeattie,2 the portion of the life-history in question finds afairly close parallel.

In Rhinosporidium there is a multiuucleate trophozoitephase, followed by segmentation into many closely packedpausporoblasts lying within a thin peripheral layer of un-

1 Mincliin, E. A., and Fantham, H. B., "Rhinosporidium kinealyi»n. g., n. sp.: a Sporozoon from 1 he Mucous Membrane of the Septum Nasi ofMan," 'Quart. Journ. Micr. Sci.,' n. s., xlix, 3, 1905, pp. 521—532, twoplates.

3 Beattie, J. M., "Rhinosporidium kinealyi: a Sporozoon from theNasal Mucous Membrane," ' Journ. Path, and Bact.,' xi, 3, 1906, pp. 270—275, two plates.

NEUROSPORIDIUM. 95

differentiated protoplasm; each pansporoblast becomes aspore-morula, dividing into about a dozen small, round, uni-nucleate spores. In these respects R h i n o s p o r i d i u mresembles Neurospo r id ium. In Rhinospor id ium, how-ever, there is a well-defined cyst secreted by the parasite,and the outline of the spore-morula is rather more distinct,there being a definite membrane round each, and the outlineof each spore in a full-grown spore-morula is also well marked.This slight difference in the definiteness of contours is oflittle importance as compared with the segmentation of pau-sporoblasts into many spores, which is in marked contrastwith what occurs in the typical Haplosporidia.

N e u r o s p o r i d i u m differs also from Rh inospor id iumin that the spore-formation from pansporoblasts is not pro-gi'essive, as it is in the latter. In Rh inospo r id ium thepansporoblasts, within a large cyst, are disposed in threezones, merging the one into the other, namely, a peripheralzone of uninucleate pansporoblasts just within a thin layer ofundifferentiated protoplasm, an intermediate zone of pan-sporoblasts with two or three nuclei each, in which zonespore-formation is proceeding, and a central mass iu whicheach pansporoblast, now a spore-morula, contains about adozen uninucleate spores. In other words, spore-formationproceeds in a centrifugal direction within the cyst, the centralpansporoblasts completing their sporulation before the peri-pheral ones begin to divide.

In Neurospor id ium, on the other hand, there is aninterval between the simultaneous formation of the uainucleatepansporoblasts and their segmentation, and the segmentationof the spore-morulEe occurs simultaneously throughout thesame capsule.

As pointed out by Mitichin and Fantham,1 Rh inospor i -dium, in the successive formation of its pansporoblasts,resembles SchewiakofPs parasite of the Oyclopidse,3 to which

1 Loc. cit., p. 529.2 Schewiakoff, W.,"Ueber einige ekto- und ento-parasitische Protozoen der

Cyclopiden," 'Bull. Soc. Imp. Nat. Moscou,' n. s., vii, 1893, pp. 1—29, pi. 1.

96 \V. G. RIDE WOOD AND H. B. FANTHAM.

Caullery and Mesnil1 have recently given the generic nameScheviakovella.

The degree of importance to be attached to the mode ofsporulation, whether simultaneous or successive within thesame cyst, is for future investigations to decide; for thepresent, our opinion is that the production of many spores byeach pansporoblast instead of a few spores or a single spore,a feature possessed in common by Rhinosporidium andNeurosporidium, is of more importance than the formerfeature, in which the two genera differ.

After due consideration of the various poiuts above setforth, we definitely place Neurosporidium and Rhino-sporidium in the order Haplosporidia, extending the orderto include these forms. Further, we divide the extendedorder Haplosporidia into two sections; (1) the Oligosporulea(norn. nov.) for forms like Haplo sporidium, Bertramiaand Coclosporidium, in which each pansporoblast pro-duces only a small number of spores or a single spore, and(2), the Polysporulea (nom. nov.) for forms like Rhino-sporidium and Neurosporidium, in which the pansporo-blast gives rise to many spores, either successively or almostsimultaneously.

Some protozoologists express doubt as to the homogeneityof the Haplosporidia as a group, and not altogether withoutreason. But Ccelosporidium, Bertramia, and Haplo-sporidium form a well-defined section, and we have shownabove the relation of Neurosporidium and Rhinospori-dium to this section. The precise limits of the order Haplo-sporidia are difficult to define because of the elementarystructure and simple developmental cycle of the organismswhich that order includes.

The Haplosporidia show points of resemblance with theMycetozoa and with the Rhizopoda. The possibility, how-ever, must not be overlooked that while some of the formsare truly primitive, others may be spuriously so, and mayowe their simplicity to degradation from a higher stock, a

1 Loc. cit., p; 156.

NEUROSPOEIDHJM. 97

process which in so many groups of animals is known toresult from indulgence in a parasitic mode of life.

For the present thelSTeosporidia may be divided as follows :1. Cnidosporidia (Doflein), with polar capsnles.

Myxosporidia, e.g. M y x o b o l u s , S p h a s r o s p o r a .Microsporidia, e.g. G-lugea, P l e i s t o p h o r a .Actinomyxidia, e.g. H e x a c t i n o i n y x o n .? Sarcosporidia, e.g. S a r c o c y s t i s .

2. Haplosporidia (Oanllery and Mesnil), simpler forms thanthe above, without polar capsules.

Oligosporulea, e.g. Bertramia, Haplosporidium.Polysporulea, e.g. Rhinosporidium, Neurospori-

dium.

SUMMARY.

The parasite begins its life-cycle as around or oval gymno-spore or arnasbula in the nervous system of the host (figs. 3and 4).

The amsebulse cause a degeneration of the nerve-tissueimmediately around them, and come to lie within cavities,one amasbula in a cavity, or several (fig. 5).

The amsebula becomes a multinucleate trophozoite, eitherby enlarging and undergoing nuclear division, or by coales-cence with other amasbulas or young trophozoites, or by acombination of both processes (figs. 6, 7, 8, 9).

The capsule surrounding the parasite is ill-defined, and isprobably formed by the host (fig. 12, etc., cps.).

The trophozoite segments into uninucleate pansporoblasts(figs. 10, 11), each of which enlarges and becomes a spore-morula.

The spore-morula gives rise to many small spores (figs.12, 13), and after the liberation of these, there remains amassof granular protoplasm, with residual nuclei (figs. 15, 16, 17).

The infection probably spreads through the nervous systemof the host by the migration of the amsebulas and trophozoites.The mode of cross-infection from host to host is not known.

It is proposed to divide the Haplosporidia into the Poly-VOL. 5 3 , PART 1. NEW SEMES. 7

98 \V. G. HIDE WOOD AND H. B. FANTHAM.

sporulea (pansporoblasts giving rise to a number of spores)and the Oligosporulea (pansporoblasts giving rise to a fewspores or to a single spore each).

The systematic position of Neurospor id ium is as follows:Phylum.—Protozoa.

Class.—Sporozoa.

Sub-class.—NTCOSPOEIDIA.Order.—Haplosporidia.

Section.—Polysporulea (nov. sec.).Genus.—Neurosporidium (nov. gen.).

Species.—cephalodisci (nov. sp.).

EXPLANATION OF PLATES 6 AND 1,

Illustrating Dr. Ridewood and Mr. Fanthain's paper onNeurosporidium cephalodisci, n.g., n. sp., a Sporo-zoon from the Nervous System of Cephalodiscusnigrescens."

PLATE 6.l?or the two photomicrographs from which figs. 1 and 2 were drawn we are

indebted to Dr.N.H. Alcock, Lecturer on Physiology in St. Mary's HospitalMedical School, and to him we hereby tender our best thanks. The photo-graphs weie taken by monochromatic light obtained by a Thorpe's gratins,and Zeiss's apochromatic objectives were used.

EIG. 1.—Section of a polypide of Cephalodiscus nigrescens takentransversely to the length of the body, and passing behind the pedicle of thebuccal shield and behind the mouth, x 120. Nap. Neurosporidium inthe thick ventral wall of the buccal shield, ph. Wall of pharynx, i.e. Trunkcoelom. m. Muscles of the ventral body-wall.

FIG. 2.—Section of a polypide of Cephalodiscus nigrescens takentransversely to the length of the body, and passing through the central nervemass and the notochord. x 100. Nsp. Neurosporidium in the centralnerve mass. s. Septum between the right and left collar cavities, no. Noto-chord. p.e. Proboscis cavity, or cavity of the buccal shield, b.s. Thickventral wall of the buccal shield. Iph. Base of the lophophore, cut obliquely.a. Ectoderm around the anus, cut tangentially.

NEUROSPORIDIUM. 99

PLATE 7.

All the figures on Plate 7 were outlined wilh camera lucida (Abb6), usiugapocliromatio objective 3 mm. homogeneous immersion (Zeiss) and compen-sating oculars 4> and 8.

All tlie figures of this plate, except fig. 16, are enlarged 1000 diameters.

SIGNIFICANCE OP THE LETTERING.

cav. The cavity in the host tissue in which the parasite lies; it often con-tains a coagulum [coag). chr. Chromidia or daughter-nuclei resultiug fromthe division of the nucleus of the spore-morula in the formation of spores,principally those daughter-nuclei remaining over in old and degenerating cap-sules, cps. Capsule, an ill-defined, sometimes alveolated membrane aroundthe parasite; it is not a true cyst, and is apparently formed by the tissue ofHie host. cps'. A similar layer lining the healthy tissue of the host. nu.Nucleus (figs. 4 to 9), or daughter-nucleus (figs. 12 to 14). p'spbl. Pan-sporoblast. r.n. Residual nucleus of Ihe spore-morula. r.sp.tn. Residualcytoplasm of the spore-morula. vac. Vacuole.

FIG. 3.—A number of free spores, a. An amtebula, with one nucleus.b and d. Later stage, young trophozoites with two and three nuclei respec-tively, c shows nuclear division, rather indistinctly, but as well as could beexpected from the material to hand.

FIG. 4.—Deeply staining spore (amsebula) lying in a definite cavity in thenervous tissue of the host.

FIG. 5.—Several such spores in a cavity, surrounded by a thin envelope{cps'.).

FIG. 6.—Several young trophozoites lying in a cavity.FIG. 7-—Young trophozoites coalesced into a plasmodium, lying in a cavity,

with another young trophozoite, which examination of the series of sectionsshows t.o be continuous with it. A coagulum is also seen, probably resultingfrom the remains of the host cells. This specimen is in the stage of pansporo-blast formation. Although smaller than those shown in figs. 8 and 9, it is ina later stage of the life cycle.

FIG. 8.—Multinucleate trophozoite or "plasmodium."FIG. 9.—Multinucleate trophozoite, surrounded by a capsule, lying in a

cavity in the host tissue.FIG. 10.—Elongate trophozoite, with oval pansporoblasts lying in deeply

staining granular protoplasm. There is also a small trophozoite by the side,lying in the same cavity.

FIG. 11.—Ovoid trophozoite, full of pansporoblasts.

FIG. 12.—Several spore-morulee (full-grown, segmenting pansporoblasts),which are dividing into spores. The parasite is seen surrounded by an ill-

100 W. G. TUDUWOOD AND H. B. TTANTHAM.

defined capsule (cps.), somewhat alveolar in character, and apparently formedby degenerating host tissue; it lies in a cavity in the host tissue, whichlatter is lined by a membrane (cps1.).

FIG. 13.—Typical capsule, full of spore-morulse, each dividing into manyspores.

PIG. 14.—Part of a capsule containing spore-morulse, one of which, havingshed its spores, shows a residual granular mass of cytoplasm, and a pale, ill-defined nucleus. The difference in the staining reaction of this remnant of thespore-morula from that of its sporulating neighbours is very striking.

FIG. 15.—Capsule containing residua of spore-morulse. Groups of nucleargranules (chromidia), some of which may perhaps form spores, while otherswill degenerate, are scattered about, chiefly in the centre. These chromaticdots are smaller than the daughter-nuclei shown in fig. 13, each of whichlatter becomes the nucleus of a spore. Some of these smaller nuclei {chr.)are surrounded by a layer of clear, refractive protoplasm.

PIG. 16.—Further stage in degeneration of residua of spore-morulse.These consist, at this stage, chiefly of residual granular protoplasm, which isbecoming vacuolated. At the periphery of the cavity, now ill-defined, inwhich the large mass lies, are seen some of the spores and young trophozoitesformed from these spore-morulse. x 500.

PIG. 17.—Still further degeneration of residual matter of the spore,morulee, showing vacuolated protoplasm and remains of nuclei (r.n.) of someof the spore-morula!. Some chromidia are grouped in the centre, but thesewere probably not destined to form spores.

a s d. &cu/n i .jt/fixyr-.Sfil Vbi.51, A S. <ft. t'.

VA.\.

Xkp

pc. fi.s.

N.liA.photo. fiuth iith.

NEUROSPORIDIUM CEPHALODISCI

in. .the nerve wacta of Cephalodiscus nigrescens.

UtSiM ft 7.

cps.

• Spore

Fig 5.

* . *

Fig 8

X

Sport mcnda

g 17

Fig.14.

/'a" I / . -v - r"f

Fig 15

.. cps.

Spun- nwr-uia-

Fit* L6

NEUROSPORIDIUM CEPHALODISCI

in the nerve tracts of CepkaMscus nigrescens.