THE GREGABINES OP GLYCERA SIPHONOSTOMA. 2 0 5

The Gregarines of Glycera siphonostoma.

ByHelen Li. M. Fixell-Goodi-icli, D.Sc,

Beit Memorial Research Fellow.

With Plate 18.

THE large Polychaet, Glycera s iphonostoma D. Ch.(sometimes placed in a separate genus Rhynchobolus), is attimes infested with numerous gregarines, both intestinal andcoelomic. On this account it is somewhat difficult to differen-tiate the stages in the life-history of anyone of them. Out oEfifty-two specimens examined at Naples in March and April,1914, twelve (I to IX, XII, XIV, and XXVI) were infectedwith a species of Gonospora which presented some interestingpoints. The affinities of this form with previously establishedspecies will be discussed later. It does not agree in detailwith the published account of any, notwithstanding the factthat Leger (11) has already recorded Gouospora sparsa nsoccurring in some undetermined species of Glycera at BelleIsle.

A specimen of Gl. s iphonostoma infected with Gonosporacan generally be detected in the living, for, through its body-wall, both the largeattacbed trophozoites and free cysts canbe distinguished. The latter move backwards and forwardssuspended in the coelomic fluid, as the host expands andcontracts. The large trophozoites are attached to the thickmuscular pharynx from the region of the jaws to the intestine(PL 18, fig. 1). The numbers aud size of the individualparasites vary very much in different specimens. Sometimesonly one or two occur, but in the case illustrated (VI) they

VOL. 6 1 , PAKT 3 . NEW SERIES. 1 4

206 HEfiEN L. M. PIXBLL-GOODEICH.

were numerous towards the posterior half of the pharynx.In another (IX) a fringe of trophozoites was found just infront of the jaws. The length of the single trophozoitesvaries from about 1 mm. to 4 or 5 mm. in length. Narrowat the attached end, they gradually widen out, and then taperto a blunt point. The smallest specimen shown in PJ. 18,fig. 1, is at e. The small projections at / were thought to bepossibly young forms, but on cutting sections of this part theywere discovered to be only the remains of the attached endsoE associates which had become free. The nucleus is sphericaland near the widest part of each individual; it containsgenerally four or five caryosomes (Plate 18, fig. 4). Each ofthe parasites is covered with a layer of the host's ccelomicepithelium, thicker in some parts than in others. It seemsclear, as will be further explained later, that the parasites oi:a certain length, having reached the coslom, attach themselvesto the pharynx. Thereupon they penetrate a little into thehost's tissue, and the peritoneum, greatly increasing in theneighbourhood, rapidly grows l-ound the parasite, forming alayer in contact with, but not attached to, the exterior cuticleof the gregarine. This covering of host's cells does not keeppace with the growth of the trophozoite, which consequentlyhas to become bent on itself to some extent, especially at itsnarrow attached end (Plate 18, fig. 4).

The trophozoites evidently revolve to a certain extent abouttheir points of attachment, and in this way the free extremitiesof two forms may come together as at a and b (PI. 18, fig. 1),when association proceeds to take place. The layer of hosttissue is eliminated from between the* contiguous extremities,and the end of one associate projects intotheendof the other,which consequently becomes cup-shaped, and thus union ofthe two associates is made secure (PL 18, figs. 2 and 3). Thisdove-tail arrangement reminds one very forcibly of the cupand ball structure described by Huxley (10) in Ganymedes,and I would suggest that in this gregarine also the distinctiveends, which the author states he could not always find, wereprobably only temporary forms taken on by the parasites at

THE UREGARINJES OF GLYOEBA SIPHONOSTOMA. 207

the beginning of association. That is at any rate the csisehere.' Before association all the parasites have regularlytapering ends. A somewhat similar mode of association oftwo or several individuals was described by Caullery andMesnil (3) in G o n o s p o r a l o n g i s s i m a , where they state thatthe extremity of one associate sometimes forces itself into theother, invaginating it " en doigt de gant." They comparethis with the similar phenomenon in the D i d y m o p h y e s ofStein. In this polycystid gregarine, however, it is thesatellite which attaches itself to the primite of a syzygy in thisway (17, Taf. IX, fig. 40), and therefore necessarily theattachment is by opposite ends. Iu the G-onospora from&]. s i p h o n o s t o m a it is union of similar ends that is affectedin this secure way. Following this association the paircontinue to rotate, and since the proximal ends are stillattached to the pharynx they become much conYoluted andshortened (PI. 18, figs. 2 and 3). It would have beeninteresting to see what would have happened during shorten-ing of the attached ends in the case of the associates in PL 18,tig. 1, marked a and b, which, it will be seen, have becomeintertwined. In PI. 18, fig. 2, is shown a case of attemptedtriple association. Evidently after the firm union of twospecimens a third has become attached, and is seeking tocome into closer relation, though as yet separated by ranchhost tissue. In normal cases the associated ends graduallyenlarge and become rouuded off, forming a spherical cyst, intowhich is drawn up most of the protoplasm from the attachedends. A little, however, may be left with its covering of hosttissue outside the cyst, as in K a l p i d o r h y n c h u s 1 (7). Thecuticle of the trophozoite may be seen to be closely ribbed byexamining the ball end of an associate. After formation ofthe cyst the cup and ball arrangement disappears and the thinpartition between the gametocytes becomes sti-aightened out,the remaining cuticle thickening slightly to form the cyst wall.Meanwhile the covering provided by the host forms a thickwall round the associates, and its proximal parts shorten and

1 The correct generic name of this parasite is discussed on p. 213.

208 HELEN L. M. PIXELL-GOODRI0H.

thicken, but continue to attach the spherical cyst to thepharynx for some time. There is no organic connectionbetween parasite and host cyst. The latter may easily beremoved at any time (PI. 18, fig. 3).

Daring the changes recorded above nuclear division hasbeen proceeding in each gametocyte. The stage showing thefirst nuclear division has not been seen. It would take placein a couple at a stage between those represented in PI. 18,fig. 1, by a and b on the one hand, and c on the other. Forin the pair of associates a and b there were only the tropho-zoite nuclei in each, while in c there were several nuclei whichwere still dividing. Here it may be mentioned that althoughall these nuclei appear similar, there is at this early stage amuch greater number in. one associate than in the other. Asimilar observation is recorded by Cunningham (7, p. 205)in the case of his K a l p i d o r h y nchus arenicolae, namely,that one gametocyte has fewer nuclei than the other. Thisdifference is here at any rate only transitory, for at the stagerepresented by d (PI. 18, fig. 1) there seems to be the samenumber of nuclei in each gametocyte; and there is still nodifference to be distinguished in the size or appearance ofthe nuclei of the two gametocytes. However, Brasil (1)stated that in Gr. var ia the difference in appearance of thenuclei of the two gametocytes only appeared distinctly on thepassage of the nuclei to the surface formed by their convo-lutions (1, p. 31). Unfortunately these later stages arowanting among my cysts. Both the formation and fusion ofthe gametes appear to be gone through rapidly at the timewhen the cyst breaks away from the pharnyx and becomesfree in the ccelom. The youngest of these cysts obtained freein the coelom still had processes at either side where thehost's cyst had broken off from the short stumps / (PI. 18,fig. 1) left attached to the pharynx. This cyst contained,besides some residual protoplasm, numerous young sporeseach with an undivided syncaryon (PI. 18, fig. 5). Twoolder cysts containing ripe spores had become completelyspherical, showing no signs of having been attached. Round

THE GREGABINBS OF GLYCERA SIPHONOSTOMA. 2 0 9

these the host cells formed a colourless layer of uniformthickness.

SPORES.

The spore is provided with a very characteristic wall com-posed of endospoi'e and exospore, and having a funnel at oneend (PI. 18, fig. 6). The exospore is thick and transparent,and is supported by processes running through it from theendospore. Similar, though longer, processes run up andsupport the sides of the funnel. The appearance of the latteris evidently what has given rise to the general statement thatin Gonospora the spore is terminated by a crown of spines—" couronne de fines pointes hyalines," Leger (11, p. 156)and Brasil (1, p. 20). When unstained, or washed out, theendospore and its processes tire refringent, but they can beshown up much more clearly by overstaining with iron-heema-toxylin. The endospore measures about 10 /u by 8 fx. Theexospore is very delicate and easily overlooked unless theillumination is very good. The processes of the endosporeenlarge slightly towards their outer ends, sometimes appearingto have globular extremities. They might almost be minutecanals, out of which a sticky fluid was oozing. At any rate,granules in the neighbourhood adhere easily to them. .Ripespores contain eight sporozoites, which escape through thefunnel. The thickness of the fully-formed spore-coat makesit difficult for reagents to penetrate it and stain the contents(PI. 18, fig. 6).

Only on the one occasion, cited above, were cysts with ripespores found free in the cavity, and those had evidently onlyjust become detached from the pharynx. In all other casessuch cysts were embedded in brown masses composed of hostphagocytes. Their brown colour was due to granules andother waste products removed from the coelom. Thesemasses were sometimes very large, one, in specimen VI, was9 mm. long by 1 mm. wide, and contained embedded in itabout a dozen cysts with ripe spores. More usually theymeasured less than 2 or 3 mm. in any direction ; sometimes

210 HELEN L. M. PIXELL-GOODEI0H.

tliey contained small tropliozoites in addition to spores, atother times necrotic Nematodes and their eggs, and veryoften old broken setse. They give, in fact, evidence of thevery effective way in which the host is destroying its parasitestogether with other useless matter. The ultimate destructionof these masses in Glycera s iphonostoma has beendescribed by Goodrich (9, p. 450), as taking place in thonephridial sacs. Much has been written also on generalphagocytic action of the leucocytes of aunelids, and destruc-tion of parasites by them has been described by Siedlicki(15). Cuenot (5 and 6), Caullery and Mesuil (2), and others.I am unable, however, to find any recorded instance ofphagocytes having to attack a layer of their own host's tissuein order to reach their quarry. In Gl. s iphonostoma theyappear to attack first any free foreign bodies, find it is onlyafter dealing effectively with these that an onslaught is madeon attached forms. The attacking phagocytes can easily bedistinguished from the covering cells of ordinary ccelomicepithelium by their larger size, branching character, andbrown colour due to enclosed granules. In one specimen(IX) a brown mass free in the ccelom was found to containfour tropliozoites and four cysts, all in a necrotic condition.The only other parasites were three or four small tropho-zoites attached quite close to the jaws. Two of these werebeing held together laterally by a mass of phagocytes withthe usual brown granules. These two trophozoites, in spiteof their covering of colourless host cells, were evidently inprocess of being detached from the pharynx and destined toultimate destruction by phagocytes. This is the only sign ofanything like lateral association that I have been able toobserve in this form, and of course here it is really nothing ofthe kind, but. merely a case of two small trophozoites beingconveniently destroyed at the same time by the host. Leger,however, described lateral association in G. sparsa, a formfound by him in Glycera sp. This makes me hesitate toconnect this parasite from Gl. s iphonos toma with thatfrom other Glycerids until such time as it may be possible to

THE GBEGARINES OF GLYCEBA SIPHONOSTOMA. 2 1 1

compare the living forms. To prevent confusion, therefore,it is perhaps well to refer to the form under consideration asG-. glyceras.

SYSTEMATIC POSITION OP GONOSFORA OLYCEEJE N.SP.

The genus Gonospora was established by Schueider(14, p. 597) in 1875 to include a species found by him in"Audou in i a l amark i i at Roscoff and also in Terebellids."The trophozoite of this form is described as elongated, broader atone end than the otherand. the spores as oval, without pi'ocesses.This would appear to be not unlike Leger's G. var ia fromAudouinia (see below). Unfortunately Schneider called thisspecies G. terebel l te Koll., although he stated immediatelyafter that he dared not affirm that his species was the sameas Kolliker's G r e g a r i n a (Monocysfcis) terebellas. If,however, the description and figure of the latter given byKolliker (10a, fig. G) be studied it will be clear thatthis form should be included in the genus Selenidium. Thisfact has already been pointed out by Dogiel (8) in hisinteresting work on the gregarines.

On the whole, then, it seems reasonably clear that the typespecies on which Schneider founded the genus Gonosporawas the one subsequently named G. var ia by Leger in 1892.The chief known characteristics of this and the other twospecies hitherto described (all apparently forms free in thecoelom) may be summed up as follows :

(1) G. var ia , Leger, from the coelom of Audou in ia(11, p. 157), further described by Brasil (1, p. 21). Tropho-zoites may be 2 mm. long; association terminal; gametesanisogamous; spores oval, 18 to 21 ju long.

(2) G. sparsa , Leger, from the coelom of Phyl lodoceand Glycera sp. (11). Trophozoites elongated, attaining alength of 1 mm. • association lateral; spores 10 ju long,nearly spherical with " couroune de pointes hyalines."

(3) G. longiss ima C. and M. (2 and 3) from coelom ofDodecacer i a. Trophozoites very large; association terminaland intimate; spore apparently oval (not clearly figured).

212 HELEN L. M. PIXELL-GOODBICH.

Ill none of the above cases, I venture to think, has thespore been studied at a sufficiently high magnification toreveal its true characteristics. Dogiel (8) has recorded theexistence of a funnel, and it seems possible that there mnyalso be found a more complicated coat than has been described.In the meanwhile it may be pointed out that the spore of thespecies from Grl. s iphonostoma in size and shape muchresembles G. s pa r sa ; but for the present its distinguishingcharacteristics may be summarised as follows :

(4) G. glyceras n. sp. from the coelom of Grlyceras iphonos toma D. Ch. Trophozoites attached to pharynxduring the greater part of their life and covered with a layerof the host's coelomic epithelium ; association terminal andmade secure by a dovetail arrangement; spores with arefriugent endospore which gives off processes supportingthe thick transparent exospore with its funnel.

In addition to Gronospora several other gregarines havebeen found in G-l. s iphonostoma which have not beenshown to have any connection with it. A few observations,however, on these may be of assistance to anyone who shouldhave a chance of obtaining this polycliEete at Naples and wishto continue the study of the parasites. I have not been ableto obtain any living specimen since leaviug Naples, andequally unsuccessful have been efforts to obtain fromthe British coasts GM. g igan tea , Quati'efages, whichMclntosh (12) maintains is the same species.

(1) Cys tob ia i n t e s t i na l i s Ssok. was obtained in threespecimens (namely, IV, XXII, and LII). This parasite wasdescribed by Ssokoloff (16) from specimens occurring in apreserved intestine of a G-l. siphouostorna (Rhynchobolus)which he received from Naples. In' this case the wholeof the sporogony is described as taking place in the intestinalwall. In my specimen of Grl. s iphonostoma with the bestinfection, while there are numerous trophozoites single orassociated in the intestinal wall (PI. 18, fig. 7), there are alsovery many free. Among these the ouly pair of associatesfound in which nuclear division has started is free in the

THE GREGABINES OP GLYCERA SIPHONOSTOMA. 2 1 3

lumen. Unfortunately no later stages are represented at all,so that I am unable to confirm the account given by Ssokolofffrom his somewhat inadequate material. This author doesnot refer to the peculiar dense meshwork of the ectoplasmwhich contains small deeply staining granules, although it canbe distinguished in one of his micro-photographs.

In his'classification (p. 227) SsokolofE has omitted 0.minchini i , established by Woodcock (18) as a speciesoccurriug in Cucumaria, but has included the badly-defined0 . schneider i Ming., which was stated by Ouenot (4, p. 4),so early as 1892, to be identical wioh 0 . holothurias. Hehas also iucluded Cunningham's K a l p i d o r h y n c h u s areui-colas among Cystobia, as advocated by Dogiel (8).1

(2) Another gregarioe occurring abundantly in at leastthree specimens (II, VIII, and XXIV) of Gl. s iphono-

1 That this last parasite is a rnonocystid Gregarine is quite clear froman examination of specimens which can easily be found in Areni -cola ecaudata at Plymouth. Since it does not undergo neogamy(i.e. precocious association, Woodcock (18)), it should be included inthe genus Gonospora rather than in Cystobia. The same is probablytrue of the parasite under consideration called Cystobia i n t e s t i n a l i sby Ssokoloff, but its life history requires confirmation.

The characteristics of the spores of Gregarines are much more im-portant, from a systematic point of view, than such details in their lifehistory as the manner of association (whether terminal or lateral), orthe time at which it takes place (whether neogamous or normal). Icannot, therefore, agree with Woodcock as to the necessity of the generaDiplodina (Woodcock, 1906), and Oystobia (Ming., 1891), in which, withone exception mentioned below, the spores are practically identicalwith those of Gonospora. The occurrence of neogamy, though of muchinterest, is not of systematic value. Woodcock himself points out thatthis condition is developed to a varying extent even in a single species,viz. Diplodina i r r e g u l a r i s (18, pp. 60 and 63), and also that thecondition occurs to some extent in other genera such as Diplocystis,Zygocystis (pp. 62 and 63). This being the case it seems preferable toplace the two species of Diplodina, viz. D. (Cystobia) i r r egu la r i sand D. (Cystobia) minchin i i , in the genus Gonospora, and Cys-tobia holothuriee (the spore of which is provided with a short flat-tened tail as well as a funnel) in the genus Lithocystis (13a). Wood-cock (18, p. 60) fully realised that such were their relationships.

214 HELEN L. M. PIXELL-GOODKIOH.

s torn a is a ccelomic form always seen in pairs (PI. 18,fig. 8). These are generally very much smaller than theGonospora trophozoites attached to the pharynx, and haveno apparent relationship with them. Specimens measuredvaried between "2 and l-6 mm. in length and were nevermore than 1 mm. wide. They were attached to the body-wall, retractor muscles, or intestinal wall, never to thephnrnyx. The two individuals of a pair are held togetherand attached to the host by a homogeneous secretion (PI.18, fig. 8, m), which extends between them for about onethird of their length from the attached extremities. Thecuticle is thin and the protoplasm finely granular. The large,oval nucleus varies in position, but is generally nearer thedistal end; it is limited by a well-marked niembraue andsometimes contains a couple oE caryosomes.

Presumably this form has no more connection with C.i n t e s t i n a l i s than with Gronospora glycera3;andl merelymake these few observations in case they may be of use insubsequent investigations.

(3) In two specimens (II and TV) free or attached mono-cystid gregarines were found in the intestine, especiallytowards the posterior end. There was nothing especiallycharacteristic about these, and it is unlikely that they shouldhave auy connection with the previously described forms..

(4) In one specimen (III) small active gregarines werefound in the anterior region of the intestine. These wereprobably the sporozoites of one of the above forms, althoughno spores were seen.

SUMMARY.

(1) There appear to be four different gregariues parasiticin Grl. s iphouostoma, including at least one species ofGronospora.

(2) Gonospora glycerse n. sp. is surrounded throughoutthe greater part of its existence by a layer of host epithelium.Association is made secure by means of a dovetail arrange-ment. The spores, under a high magnification, reveal a more

THE GREGARINES OF GLYOERA SIPHONOSTOMA. 2 1 5

complicated structure than has previously been described inthe genus.

THE MUSEUMS, OXFORD;J u l y 16th, 1915.

LITERATURE.

1. Brasil, L.—" Recherches sur la Reproduction des Gregarines Mono-cystidees," ' Avch. Zool. Exper.,' 4ine a., iii, 1905.

2. Caullery M., and Mesnil, F.—" Les formes epitoqnes et revolutiondes .Cirratuliens, G. longiss ima of Dodecaceria," 'Annales deUniversitie de Lyon,' xxxix, p. 86, 1898.

3. " Sur une Gregarine ccelomique nouvelle," ' 0. R.Ac. Sci.,' cxxvi, p. 262, 1S9S:

4. Cuenot, L.—" Commensaux et Parasites des Echinodermes," 'RevueBiol. du Nord de la Prance,' v, pp. 1-23, 1892.

5. " Etude physiologique sur les Orthoptcres," ' Arch, de Biol,xiv, pp. 293-343, 1895.

6. " Les globules sangnins et les organea lymphoides des In-vertebres," ' Arch. d'Anat. niicrosc.,' i, 1897.

7. Cunningham, J. T.—" Ka lp ido rhynchus arenicolse," 'Arch.Protistenk.,' p. 199, 1907.

8. Dogiel, V.—" Beitriige zur Kenntniss der Gregarinen. iii, Ueberdie Sporocysten der Colom-MonocystideiB," 'Arch. Protistenk.,'xvi, pp. 194-208,1909.

9. Goodrich, E. S.—'• On the Nephridia of Polychajta. i, Glyceraand Goniada," ' Quart. Journ. Micr. Sci.,' 43, 189S.

10. Huxley, J. S.—" On Ganymedes anaspid is , " ibid., 55, pp. 155-177,1910.

10a. Kolliker, A.—" Gregarinen," ' Zeits. wiss. Zool.,' i, 1S48.11. Leger, L.—" Recherches sur les Gregarines," 'Tablettes zoologiqnes,'

iii, 1892.12. Mclntosh. W. O.—' Monogr. Brit. Annelids,' II, ii, p. 482.13. Mingazzini.—" Le Gregarine delle Oloturie," ' Rendic. R. Accad. dei

Lincei,' vii, p. 313,1891.13a. Pixell-Goodrich, H.—" The Sporozoa of Spatangolds," ' Quart.

Journ. Micr. Sci.,' 61, 1915.14. Schneider, A.—" Contributions a Thistoire des Gregarines," 'Arch.

Zool. Exper.,' iv, p. 597, 1875.15. Siedlecki, M.—" Quelques observations sur le role des amibocytes

dans le coelonie d'un annelide," 'Ann. de l'lnstitute Pasteur,'xvii, pp. 449-462,1903.

216 HELEN L. M. PiXELL-GOODRICH.

16. Ssokoloff, B.—" Cystobia in tes t ina l i s , nov. sp.," Arch. Protis-tenk., xxxii, pp. 221-228,1914.

17. Stein, F.—" Ueber die Natur der Gregarinen," ' Arch. Anat. Physiol.Med., Berlin,' p. 184,1848.

18. Woodcock, H. M.—"The Life-Cycle of Cystobia i r r egu la r i sMinch.,' ' Quart. Journ. Micr. Sci.,' SO, p. 1,1906.

EXPLANATION OF PLATE 18,

Illustrating Mrs. Helen L. M. Pixell-Goodricli's paper on"The Gregarines of Glyoera siphonosfcoma."

(All preparations, unless otherwise stated, were stained with iron-hsematoxylin and drawn with the aid of a camera lucida.)

Kg. 1.—Posterior region of pharynx of Glycera s iphonostomashowing attachment of single and associated Gonospora. a, b, c, d,pairs of associates in order of development; e, young trophozoite; / ,remains of attached ends left by liberated cysts. Drawn after fixationin corrosive acetic mixture. X 8.

Tig. 2.—Optical section of distal ends of two associates, a and b,showing intimate union between them, and a third specimen, c, attempt-ing to effect multiple association. Stained iron-hffiniatoxylin. x 23.

Fig. 3.—A pair of associates slightly compressed, after removal ofenvelope of host cells. Drawn from the living, x 23.

Pig. 4.—Optical section of a single trophozoite, after being detachedfrom pharynx, showing small nucle\is with several caryosomes. h,Covering of host's cells ; p, proximal end bent on itself, x 75.

Pig. 5.—Optical section of young spore, with characteristic coat justformed. The nucleus (syncaryon) starting to divide, btit much over-stained. X 2000.

Pig. 6.—Ripe spore showing characteristic endospore with processes,faintly staining exospore, and funnel supported by slightly longer pro-cesses, x 2000.

Pig. 7.—Cystobia i n t e s t i na l i s Ssok. Section through trophozoiteat base of intestinal cells, n, Remains of hypertrophied nucleus ofhost cell, ec, Dense ectoplasm with deeply staining granules, x 500.

Pig. 8.—A pair of cceloniic gregarines held together by a structure-less membrane (m), extending to nearly half their length from theattached ends, x 250.

Jcwrn,, fticmScl. VoL. 61, U.S. c?l.18.

6.w 8 .

H.M.Pixell Gootlrioh del Huth,Lit.hrL ,;H

P I X E L L - G O O D R I C H — G R E G A R I N E S OF G L Y C E R A .