LIV 3005 Xiverpool flDariuc Biolociy (Tontinittcc. L.M.B.C. MEMOIRS ox Typical British Makixe Plants c~ Animals EDITED BY W. A. HeRDMAN, D.Sc, F.K.S. VII. LINEUS R. C. PUNNETT, B.A. (With 4 Plates) Price Two Shillings. I-ONDOX Williams ,:v NoKdA-n- Al'KIL. h)()l HARVARD UNIVERSITY. LIBRARY MUSEUM OF COMPARATIVE ZOOLOGY. {Jo {y\A..ei)(/\K c30jULJ^wv>X;JLn_ Ck ^Q 10 L.M.B.C. MEMOIRS Vll. L I N E U S NOTICE. The Committee desire to intimate that no copies of these Memoirs will l)e ]iresentefl ot exehanfjed, as the prices have been fixed so low that most of the copies will have to be sold to meet the cost of ])rodnction. The Memoirs may he o1)tained, post free at tlie nett piices stated, from Messrs. AVilliams and Norg-ate, 14, Henrietta Street, Covent (larden, London. Memoir T. Ascidia — pnhlished in October, 1899, HO pp. and five plates, price 2s. ,, IT. Cardinm — published in ])ecember, 1899, 92 pp., six plates and a map, price 2s. 6d. ,, III. Echinus — published in February, 1900, • Ui pp. and five plates, price 2s. ,, lY. Codinm — published in April, 1900, 20 pp. and three plates, price Is. (Id. ,, Y. Alcyonium — published in January, 1901, • )0 pp. and three plates, price Is. Od. ,, YI. Lepeophtlieirus aiul Lernaea — pu])lislied in March, 1901, 02 pp. and five plates, price 2s. ,, Yll. Linens — now ready (April, 1901) 37 pp. aiul four plates, price 2s. Xivcrpool flDarinc BiolO(j\; dommittcc. L.M.B.C. MEMOIRS ON Typical British Marine Plants o- Animals EDITED BY IT. A. HERDMAN. D.Sc L\R.S. VII. LINEUS BY R. C. PUNNETT, B.A, (With 4 Plates) Price Two Shillings. '^ L(3N ,ONDOX Wn.l.lAMS Sc NORGATF, April, 1901 EDITOirS PREFACE. Tjie Livi'ipool Marine HioloQ-y ( 'onunittee was const itiitcd in 1(S(S5, with tli(> ohj(M't of investigating' the Eaiina and Flora of the Irish Sea. 'I'lie diedging, trawling, and other collecting expeditions orji-anised bv the Coniniittee have heen carried on inter- niittently since that time, and a considerable amonnt of material, both pnblisli(Ml and unpublished, has been accumulated. Fourteen Annual Reports of the Committee and Kve volumes dealing vrith the "Fauna and Flora" have been issued. At an early stage of the investigations it became evident that a Biological Station or Laboiatory on the sea-shore nearer the \isual collecting grounds than Liverpool would be a material assistance in the work. ('onse(piently the Committee, in ISST, established the Puffin Island liiological Station on the North Coast of Anfflesev, and later on, in 1S!)'J, moved to the more commodious and convenient v^tation at Port I'irin in thr centre of the ricdi collecting grounds of the south end of the Isle of Man. In these thirteen yeais" e\[)erience of a biological Station (five years at Puitin island and eiglit at Port J'jiin), where College students and young amateurs formed a large proportion of the workers, the want has been constantly felt of a series of detailed descriptions of the structure of certain common typical animals and plants, choscMi as representatives of their groups, and dealt with l)v specialists. The same want has probably been felt in other similar institutions and in many College laboratories. The objects of the Committee and of the workers at the Piological Station have hitherto been chiefly faunistic and speciograi)hic. The work must necessarily be so at hrst VI. when opening up a new district. Some of the workers have published papers on mori)hological points, or on embiyology and observations on life-historie.s and habits ; but the majority of the papers in the volumes on the " Fauna and Flora of Liverpool Bay ' have been, as was intended from the first, occupied with the names and characteristics and distribution of the many different kinds of marine plants and animals in our district. And this faunistic work will still go on. It is far from finished, and the Committee hope in the future to add greatly to the records of the Fauna and Flora. But the papers in the present series are (juite distinct from these previous publications in name, in treatment, and in purpose. They will be called the " L.M.B.C. Memoirs," each will treat of one type, and they will be issued separately as they are ready, and will be obtainable Memoir by Memoir as they appear, or later bound up in convenient volumes. It is hoped that such a series of special studies, written by those who are thoroughly familiar with the forms of which they treat, will be found of value by students of Biology in laboratories and in Marine Stations, and will be welcomed by many others working privately at Marine Natural History. It is proposed that the forms selected should, as far as possible, be common L.M.Ji.C. (Irish Sea) animals and plants of which no ade({uate account already exists in the text-books. Probably most of the specialists who have taken part in the L.M.B.C work in the past, will jn'epare accounts of one or more representatives of their groups. The following have already promised their services, and in many cases the Memoir is already far advanced. The first Memoir appeared in October and the second in December, 1899, the third in February, and the fourth in April, 191)0, the hfth in January, and the sixth in March, Vll. 1901, while this seventh one will be ready early in April ; others will follow, it is hoped, in rapid succession. Memoir I. Ascd)ia, W. A. Herdnian, 00 pp., 5 Pis., 2s. II. Caruium, J. .Johnstone, 92 pp., 7 Pis., 2s. 6d „ III. Echinus, H. C. Chadwick, .JO pp., 5 Pis., 2s. „ lY. CoDiUM, R. .). H. Gibson and Helen Auld, 26 pp., -J Pis., Is. (jd. Y. Alcyonium, S. J. Hicksoii, 80 pp., 8 Pis., Is. 6d. „ VI. Lkpkui'htheirus a.nd Lern^a, Andrew Scott, 02 pp., 5 Pis., 2s. ,, VII. LixEus, 11. C. Punnett, 37 pp., 1 Pis., 2s. GiGARTiNA, 0. V. Darbishire. Plaice, P. J. Cole and J. Johnstone. BuGULA, Laura P. Thornely. Oyster, W. A. Herdnian and J. T. Jenkins. Dexdronotl-s, J. A. Cliibb. Peridinians, G. Murray and P. G. Whitting. ZosTERA, E. J. Harvey Gibson. Himanthalia, C. E. Jones. Diatoms, F. E. Weiss. Fucrs, J. B. Parmer. BoTRYLLoiDEs, W. A. Herdnuui. Cl-ttle-Fisii (EledoxVe), W. E. Hoyle. UsTRAcou (Cytuere), Audrew Scott. Patella, J. 11. A. Davis and H. J. Fleure. Calanls, I. C. Thompson. Actinia, J. A. Clubb. HvDRoiu, ]-:. T. Browne. Myxine, G. B. Howes. BucciiXUM, M. F. AVoodward. Calcareous Sponge, 1{. Hanitsch. Arenicola, .1. H. Ashwoith. Antedon, H. C. Chadwick. Porpoise, A. M. Paterson. In addition to these, otlier Memoirs will lie ari'finged for, on suitable types, sucli as Sayitta (by Mi-. Cole), a Cestode and a Turbellarian (by Mr. Shipley), Carcinus, an Aniphipod, and a Pycnogonid (probably by Dr. A. 11. •lackson). As announced in the preface to Ascidia, a donation from Mr. F. H. Gossag'e of AVoolton met the expense of preparing' fjie plates in illustration of the first few Memoirs, and so enabled the Committee to commence the publication of the series sooner than would oiherwise have been possible. A second donation received since from Mr. Crossage, and another recently from Mrs. Holt, are regarded by ihe Committee as a welcome encouragi'- ment, and will be a great help in carrying on the work. W. A. Hekdmax rnivei^itv College, Lixerpcol, April, 1901. L.M.B.C. MEMOIRS. No. Vir. LTXEU8. .[{. C. rrXXETT, 13.A. Ixi'KoDrcriox. SixcK (lie liiue when it was first noticcMl by rallas, in I7(i(i, as " alia Limibiici species marini tota atva," the Nemertean now known as Lim-iis (/esscrens/M (which we choose as our type) has been the recipient of no less than 10 generic and 1-5 specific names associated together in various perniirtations and combinations. Our knoAvledge of the habits and anatomy of the worm are chiefly due to """M'Intosh, Barrois, Hubrecht, Uudemans, and ^[ont- o-omerv. Yaivine- in size from about (i-2U cm., it is one of the commonest Nemerteans of our shores, occurring abundantly, and frequently in tangled masses, under stones between tideniarks and in the laminarian region. * M'Intosh, W. C. — British Annelids, The Nenieiteans, London, ISTiJ. Barrois, J.^Embryologie des Nemertes, Annales des Sciences Naturellcs, Paris, 1877. Kulireclit, A. A. W. — Contributions to the p]mbryology of the Ncmcrtea, Quart. Journ. of Mic. Sc. 1886. Oudcmans, A. C. — The Circulatory and Xephridial apparatus of the Nemertca. Quart. Journ. ^fie. Sc. Supplement, 1SS5. ^Ii)nl.uonicry, T. H. — On the Connective Tissues and Body Cavities of the Nemerteans. Zoolog. Jahr. 1897. ,, Studies on the elements of the' Central nervous system of the Heteronemertiiii. Journal of ?*[oi-p1io- logv.' ls<.)7. Its geogiaphical range is extensive, since it occurs on the shores of Inith sides of the Xorlh Atlantic, extending from Greenhmd in the Xorth to ^Jadeira on the one side and Florida on the other. It al>o occurs, though not com- monly, in the Mediterranean. Two very distinct colour varieties are to he met with, vi/., reddish blown and olive green — the latter being the more common. Specimens intermediate between these two are also fre([uent. The colour is more pronounced in front, and is darker on tlie dorsal surface than on the ventral. The snout and mouth are bordered by pale margins. (Jn the head a reddish patch marks the position of the brain, and a bright red colouration is also found in the head slits which is said to be due to the presence of hiemoglobin. On the snout in front of the brain occur the deeply pigmented e3^es usually varying from 3 to (J in number on each side, though moie may be present on one side than on the other. Of the many openings on the body two may be readily seen — the proboscis pore at the tip of the head, and the mouth on the ventral suiface just behind the brain. The anus is small and terminal. Of the other openings the lateral nephridial pores in the uesophageai region can only be made out in sections, whilst the numerous generative pores are at certain seasons visilde under a lens as a row of white dots on either side in the intestinal region. For what is known of the habits of the worm we are chiefly indebted to M'Intosh, who kept them in captivity. He states (in the work above cited) that "Linens gei^serensis " progresses in an easy, graceful manner, with slight " undulatory motions of the head, its body being marked *' by successive contractile waves, which proceed from " before backwards. The specimens frequently herd ■' together in the water, which they are prone to leave, and 8 " remain attached to the side of the o-lass a considerable " time. They are very easily kept in confinement for " years : but, as with many of their allies, great diminn- " tion of bulk occurs, from deprivation of the natural " supply of food. AVhen recently captured specimens are " placed in a jar containing injured Annelida, numerous " fa3cal masses, cjnsisting of the l)ristles of Xcirla jiclariica, " and other annelids and digested matter, are found lying " on the bottom of the vessel, sliowin"' how yieedilv tliev " have fed : a fact, indeed, very easily ascertained by "actual ol)servation. It is also frequently noticed that " specimens confined in vessels along with the deep green ''Kii/(i/i<( r/i-/(iis assume a similar hue, probably from " feeding on the rejected debris of those animals, if not " upon the latter themselves. In their native haunts the " stones under which they lie are often placed on dark, " muddy, and highly odoriferous *and or gravel, and the " water cannot be otherwise than brackish at the estuary " of a river."' BODY WALL AXD MUSCULAK SYSTEM. The outside of the animal is completely covered with cilia, which are borne by long slender cells (PI. II., tig. ■!), widest externally where they carry the cilia, antl narrow- ing to a ver}' tine process which is somewhat liranched. and is inserted into the basement membrane. They con- tain small elongated nuclei in the more external portion. The larger and more rounded nuclei found in the epidermis are for the most part connected with the large greenish unicellular gland cells occurring all over the skin. These last stain vividly Avith picric acid or with eosin, and it is probably their contents which give the skin its markedly acid reaction. These two forms of ceil constitute the main mass of the epithelium, though it is [)(issil)le that caii^ful liistdluti'iral work mav demonstraio tlip j)r('s(>ii((' of ciliated cells modified to form sense colls somewhat resembling- tlH> rods of the vertebrate eye. The basement membiane is structureless, and is com- })osed of the intercelhilar substance around connective tissue cells. It takes a deep stain with hiPmatoxylin or nig-rosin. though it is ajjiiarently unaffected by the carmine dyes. Beneath the basement membrane is found a thin layer of fine circular muscle fibrils, and underneath those again a ditfuse layer of small gdands, each composed of scxeral c(dls. The secretion of these cutis g-fands stains deeply with carmine or lueniatoxylin, and by it the minute ducts of the g'lands may l)e traced to their external oj^en- ing's. ^lixed up with the cutis gdands are numbers of connective tissue cells of a peculiar form containing pig- ment. These will be referred to later among the connec- tive tissues. Around the cutis glands are also found the most external fibres of the external longitudinal muscle layer. The fibres are separated into a numl)er of small bundles by the connective tissue, a structureless invest- ment exhibiting the same staining alftnities as the base- ment membrane, with which it is piobably identical as regards composition. Separating the outer longitudinal and the circular muscde layers (PI. IL, hgs. 1 and 'J. PI. III., fig. 8) is a tunic of nervous tissue, consisting of fibrils o-iyen off from the lateral nerve cords, wlii(di also lie between the same two muscle layers. The ciicular muscle layer is on the whole not so thick as the outer longitudinal layer, and of about the same thickness as the inner longitudinal layer which it immediately invests. Towards the posterior portion of the body the outer longitudinal muscle layer becomes greatly diminished, whilst the internal longi- tudinal layer here surpasses the circular layer in thickness. A thill layer of (loisn-vential luusclcs (PI. IIT., fios. (i and 8) is found limning' along- each side of an intestinal divertitniluni. In some Lineidip a horizontal nuiscle layer runs between the [)ro1)oscis sheath and (esophagus in the mouth region, l)ut this layer does not occur in the jnesent species. The muscle hhres in the head are continuorrs with the outer longitudinal layer. The inner longitudinal and circulai' layers of the trunk are not continued anteriorly beyond the l)rain. In the head, however, aie developed an inner layer of longitudinal and an outei- laver of circular tibies round the rhyiiclioda'uni and cephalic blood lacunie (PI. I., hg. '2). rUOHOSClS .\X1) PKOIiOSCIS siik.xth. The remaining portion of the muscular system is that in connection with the proboscis and its sheath. These two structures commence at the anterior level of the brain, and their relations to one another and to the rhyncho- dauim may l)est be gathered from a reference to PI. II., fig'. 7. The proboscis sheath consists of a layer of longi- tudinal muscle fibres, surrounded externally by a layer of circular ones, and internally invested by the flattened epltheliiim of its contained cavity, the rhynclioc(elom. In its retracted state the proboscis lies in the last-named cavity, and its outer surface is covered by the rhyncho- (•(elomic epithelium. The internal . surface of the piolxtscis (/.r., in tiie retracted state! is lined by the high glaiidulai- probosci- dial epithelium, which is of ectodermal origin, aiul is continuous with the epithelium lining the rhyncho- da'um. In other words, the ca\'ities of the retracted proboscis and the rhynchoda'um aic one and tihe .-ame, whilst that of the rhyiichoc(eloni is entiiel\- closed and separated from the rhynchoda'uni by the attaclimeut of the proboscis. Wlien, as sometimes happens, the proboscis is violently extruded and broken off at its attach- ment the rhynchocoelom and rhynchodseiim form a con- tinuous cavity until the proboscis is regenerated. The rhynchopcplom contains a colourless corpusculated tlniil. At its commencement between the rhynchocielom and rhvnchodteum the proboscis consists of a layer of longi- tudinal mnscle fibres covered externally l)y rhvnchocd^- lomic and internally by pr()])oscidial epithelinm (PI. I., fig. 4, p.). Just beneath the last layer are the two prolios- cidial nerves. Further back the proboscis becomes con- siderably thicker, and a layer of circular muscles makes its appearance l)etween the longitudinal muscles and the proboscidial epithelium. Fiom this circular layer at two opposite poles, as seen in transverse section, fibres pass throuffh the longitudinal lavei' to the basement membrane just beneath the rhynchocoelomic epithelium, crossing one another in two directions (PI. II., fig. (J, nirr.). In this way are formed the so-called muscle crosses characteristic of the Lineid proboscis. The crossing is more apparent on one side than on the other. In this region also the proboscidial epithelium is thrown up into papillae, and is highly glandular, whilst just beneath it there is a com- plete investing nervous layer which has been formed fiom the two probo.!vo the purpose of respiration, and the vascnlaT network wliicli surrounds this reg-ion of the alimentary canal (PI. TI., tig. '2, oe.-:/.), lends some support to such a view. Histohigically the (esophagus, like the rest of the alimentary tract, is lined hy ciliated epithelium. S([uee/.ed in among the ciliated cells (PI. IV., Hg. •)) is a number of laige unicellular gland cells, among which two, or jiossihh- three, types may be distinguished. After staining with borax carmine, followed by picro-nigrosin, many of these gland cells (PI. lY., tig. o, /9) take a yellow stain. They appeal- to be full of laige coarse granules. Others similarly gianulated, though much less numerous, shew an intense crimson colouration (PI lY., fig. 3, }f). P^xcept for the difference in staining reaction, these two tyj)es are indistinguishable. The greater number of the gland cells, however, present a ditferent a])pearance. Their contents consist of a coarse spongy network, which presents a slaty-piirple hue (PI. I\ ., fig. •!, a). Whether these various types are in reality d.istinct, or whether they re})resent difierent stages in a single type, is a question which must be left for future his- tological investigation to decide. Besides being found l)etween the ciliated cells, these unicellular gland cells also occur massed beneath the ridges of the (esophageal epithelium. li seems feasible that these large glandular (esopha- geal cells sup])ly the active juices of digestion, whilst tlie intestinal region is more concerned with al)S()r])tion. The moje granulai- of them bear a close resemblance to the large unicellular glands of the integument which, as has already been seen, are probably concerned with the markedly acid skin reaction. Siu-h a fact lends some supj)ort to the view that the (eso])hagus is derived from <) tlie ectodcnii, and, sluiuld futuic invc^stio-atioiis })r()vo siicli to 1)0 the case, the Xemeiteans wcnihl present the interesting feature of possessing- an endoderm which pro- hahly does not contrihnte to the digestive jnices, hnt is only concerned Avith absoipiioii. Jkit in the absence of decisive embryological data, and of the histological appearance of the varions parts of the lining of the alimentary canal after injection of various sul)stances, nutritious or otherwise, the qnestiini must be left open. The oeso23hagus is constricted at its jiosteiior end, and behind this constriction starts the intestine, with its lateral pouches, where the character of the lining of the alimentary canal becomes entirely changed. The intes- tinal cells are laige, though long and nariow (PI. HI., rig. 2). ]*]acli contains a somewhat (dongated nucleus near its base, whilst between it and the ciliated surface are a number of small round bodies which shew neither nucleus nor anv definite stiiu'luit'. These little bodies have been considered to be stored food material absorbed in this region, and in some* Xemerteans several types have been distinguished. The region of the regularly airanged intestinal diver- ticula (PI. TIL, tigs, (i, and 7, /.d.) continues almost to the anus where the alimentary canal opens to the exterior bv a veiv short rectum. It is worthy of note that traces of food are rarely found in the digestive canal of a Xemertean. Yet various anecdotes illustrating their vcnacity have been given, among which may be mentioned an observation of *l{iches, who, writing about Mirrura pui'ijurea, states that " A specimen of about ■! or 4 cm. was placed in a dish with * Rifhos. T. ir. A list of the Nemertines of Plymouth Sound. Journ. >hu-ine Biol. Assoc. Vol. III. 10 "a Ki//i( )in r/i s net si of (luito 'JO cm. leuotli. Soim^ little ■■ Hinc aftoi T was astonished to find the Min-uid busily " engaged in swallowing the Kunemertes. The posterior " one-fifth of the latter had already disappeared into the "mouth of the former when I noticed them, and still the "assailant was struggling to gulp down more of its prey. "In the meantime the victim glided round th(> dish, "apparently not suffering the slightest inconvenience "from the attack upon its posterior extremity. " T'ltimaiely Loth attacker and attacked became (juiescent, " the formei having become moic than twice its previous " girth. The portion of the Eitnemo-fes in the gut of the Mirnifd still remained in continuity with the rest of " the body, though apparently undergoing digestion." Possibly the food is digested and absorbed and the excreta expelled with a rapidity whicdi prechules its presence within the alimentary canal for any length of time. THE VASCT'L.\E SYSTEM. According to the histological structure of the walls, the vascular spaces in Lifuiis are spoken of partly as lacuna^ and })artly as vessels. The lacuna^ are found in the head and in the (esophageal region. They are surroimded by connective tissue, and their only wall consists of a delicate membrane on which occur small oval nuclei ai intervals. The vessels, which occur only in the intestinal legion, jiossess a rather more elaborate structure. They are lined by an endothelium (PI. IV., fig. 2), closely packed with spherical nuclei, and in which cell outlines are not readily lo be distinguished. This endothelium rests iipon a well- marked structurek'ss basement membrane, l)ut no circular muscle fibres are j)resent. External to the basement mem- 1)rane is a layer of large parenchyma cells, highly vacuolated and with definite c(dl outlines. 11 In the pieeerebial reo-jon the vascular system consists of two laciinjie (PI. I., tig. 2), which join at the tip of the \\vad over the rhynchoiheuni. Tliese two cephalic laciinse, together with the proboscis sheath, are surrounded by the nervous ring. About the level of the ventral com- missure of the l)rain they unite in the mid-ventral line lielow the proboscis sheath, and from the commissure so formed is given oft' the median dorsal vessel (or more l)ro])erly lacuna). Further back the lateral lacuna' form a second ventral communication, whence spring the two small l)uccal vessels which join the vasculai' network in the (esophageal region. Soon after this the lateral lacuniF widen out greatly, and surround the hinder portion of the cerebral ors-ans. -lust behind the mouth theie is a con- tinuous network of (esophageal lacunae surrounding the ventral surface of the (eso})hagus (PI. II., fig. 2), though the lateral lacunie can still be recognised as the largest and most dorsal of the spaces seen in transverse section. Thev are in close proximity to the })roboscis sheath, and the nephiidial tububvs come into contact with them (PI. II., fig. 2, c.rf.). The median dorsal vessel pierces the wall of the probi)scis sheath directly after its formation at the level of the l)rain, and lies in the median ventral line. Directly over it the i hynchoccelomic epithelium, which alone sepaiates the vessel from this cavity, assumes a cohimnar appearance. The (esophageal lacunar network extends nearly to the end of the (esophageal region. It then terminates, and the lateral lacuna:^ become much smaller and trausfoinied into the lateral vessels which at first lie just above the level of the lateral nerves, but soon take up a position ventral to the intestine, and not far removed from the mid-vential line. At about the same level the median dorsal vessel leaves the proboscis sheath. V2 ;ni(l hoconiino- a tiiie vessel, runs for the rest of its extent between the ])7'ol)ospis sheath and the intestine. In the intestinal reg'ion the lateral and median dorsal vessels coinnumicate liy a series of commissures (PI. IV., fio-. 1, chr) passinof round the diverticula, and which are of lacunar natuic, hein.o' devoid of the investing laver of ])ai(>nchyinatous cells. As the laleiai vessels do not coni- municale vt'ntially, tlie last coiiiuiissuie (d' the vascuhii- system lies dorsal to the alimentary canal. The 1)1 ood is colon iless, and contains some corj)Uscles. Ill some ]N'emeiteans it is icd from the presence of hsemoglohin. As to its course, it has heen stated to How backwards in the median dorsal and forwards in the lateral vessels. It is ])ossil)le, hoM-ever, that, in the absence of contractile fibres in the vascular system, there is no definite circulation, but that the blood is inter- mittently kej)i in motiini by contraction of the mus(des of the body wall. I'o what extent the vascular Huid acts as a respiratory medium is open to (pU'stion. Idie vascular network round the oeso])hag-us of many Xemerteans and the occasional presence of luvmoglobin suggest that it mav have siicli a function, but against this must be set the fact lliai the (esophageal lacume are not present iii some gr(nips, and also that the presence of luemoglobin is of rare occurrence. A more likelv view is that respiration is mainly, if not entirely, carried out by the integument of the smaller forms: and this may al>o be the case even in the larger ones, since the body is usually capable of e.xtreme attenuation. i:\(iti;i'()i!'i svsri'.M. The so-called excretorv or nepliiidial systejii in Liiniis (/(■ssrri'iis/s consists of a iiumbei- of small tubuh\s lying in close proxiiiiitv to the lateral hicuiue in the (esophageal IB leo-ion (PL II., fig. 2), and roinniunicatiiig with tlio exterior by a immher nf fine duets wliicli pierce the liodv wall and open laterally above the level of the lateral nerves. The number of ducts varies both in different specimens and on the two sides of the same specimen (PI. IV., fig". 1, t'.rd.). There are usually from 0 to 12 on each side. It otttMi happens that some of tlu> diicts are incomplete, the portion which would ])ierce the circular muscle layer being missing, though whether this is due to such ducts being ntMv formations in cours(> of inward growtli fron) the ectoderm, or whether they are commencing to atrophv is an undecided point. It has beeii stated that the number of ducts increases with the growth of the animal, from which it would ai)peai- that such incomplete ducts belong to the former of the above categoiies. On the other hand, the writer's own experience is that a large specimen may have but half as many ducts as one considerably smaller, though in such a case certain of the ducts in the large sjjecimen may possess a very much wider lumen than the rest. Consequently it is quite likely that there is a period in which the numl)er of dm^ts increases, and then later a period in which certain of the ducts enlarge, with the result that others atrophy through disuse. Put the ([uestion is one that requires more fully working out. The excretory tubules commence not far behind the mouth, and extend almost to the end of the (esophageal region (PI. IV., fig. 1). ])irectly they cease the lateral blood lacunie become the lateral vessels and the median dorsal vessel leaves the proboscis sheath. Histologically th(> tubules consist of what would })robably be styled cubical epithelium, were it possible to distinguish the cell outlines. Its piotoplasiu, which stains icadilv, is somewhat granular and contains 14 splierical nuclei (PL III., fig-. 5). It is spaiing-ly provided with long cilia. Xear the blind end of a tnhule these cilia are often more ahnndani, and are directed awav from hlind end (PL III., fig. 5). Xo flame cells, however, have l»een detected in this species. The excietory system lies almost wholly above the level of the lateral nerve cords, never extending into the lacunar network ventrally. Though it is in close con- tact with the (esophageal lacunae, there is no communica- lion between the vascular and excretory systems. Although certain observers claim to have demonstiated such a communication for certain of the more primitive forms, such as Cor/iieJIa and Carinoma, it is exceedingly doubtful whether it exists. Concerning the nature of the fluid contained in the excretory system no observations ha VI' l)een made. NERVOtTS SYSTEM. The nervous system consists essentially of two longitu- dinal cords extending throughout almost the entire length of the body and dilating anteriorly into the brain. Pos- teriorly they unite by a fine commissure ventral to the rectum. For their Avhole course they lie immediately outside the circular muscle layer, and are conspicuous objects in a transverse section of the worm (PI. III., fig. inner 1.^ iicurileiniua i.s hrokeu ut intervals by small biuullcs of axis cylinders from the ganglion cells, which can be traced into the filjrons core. Scattered nuclei are found inside the fibrous core, for the most part just inside the inner neurilemma, though a few may be seen about the centre of it. These are the nuclei of the neuroglial cells, whose branched processes form the supporting groundwork of the nervous system. The hbrous core itself is composed of: — (1) Fibres of the neuroglial cells, which stain with many reagents (r.c/., eosin), and compose the bulk of the core. (2) Xerve tubules, consisting of homogeneous axis cylinders which are usually unstained by leagents, and [)robablv in life consist of a semi-fluid substance bounded by a fine spongio-plasmic sheath. These tubules are very small, and are for the most part scattered about inside the fibrous core, though neai- the centre a large space is seen, more or less circular in outline in transverse section, which is found throughout the whole length of the core, and probably represents a bundle of nerve tubules. {'■]) Irregular spaces containing fluid.* Eeturning now to the general arrangement of the neryous system, it has already been mentioned that the lateral nerve cords dilate anteriorly to form the biain. This structure is composed of a dorsal and a ventral ceie- bral ganglion on eithei- side. The ventral ganglion is merely the expanded end of the lateral cords, and it is -It should be mentioned that the above view of the iiatui-e of the elements of the nervous coic is that advoeated by Montgoniery (loc. eit. p. 428). Biirger on the other hand (Die Nemcrtinen, Fauna und Floi-a dew Golfcs von Neapel Bd. XIX., 1895) supposes the densely staining elements, considered to be neuroglial processes on the above view, are the nervous fibrils, and that the so-called nerve tubules are clefts filled witli Ihiid. difficult to sa}- exactly wlieie the one ceases and the other starts. The dorsal gang-lia arc closely united "vvith the ventral (PI. I., tig. 5, and IM. 11., tig. 4). The two dorsal g'anglia are connected by a doisal commissure (IM. L, fig. 4) and the ventral ganglia l)y a much stouter ventral com- missure. The nervous ring thus formed siirrounds the proboscis sheath, and not, as in most worms, the alimentary canal. The posterior ends of the dorsal ganglia, now no longer in contact with the vcMitral ganglia, are continued into the so-called ceiebral organ, which will be referred to under the sense organs later. Histologically the general structure of the brain is similar to that of the lateral coids, with the difference that the ganglion cells are not all alike. In the brain three varieties of ganglion cells may be distinguished: — (1) Small cells of shortened pyriform shape, the deeply staining nuclei of whicli almost till the cell bodies. They occur on the dorsal and ventral aspects of the dorsal ganglia (PL II., tig. 4), and also in the cerebral organs, and are probably sensory in function. {2) Medium sized cells, more or less elongated and pear-shaped. These occur in the ventral l)rain lol)es and in the lateial cords, forming the greater part of the gang- lion cell layer of the latter. They vary somewhat in size, but may be distinguished fiom the next type by the shape of their nucleus, whicli is oval and not spherical, as in tht> (•">) Large cells. These are also of elongated jjyriform shap(\ and are found in the dorsal and ventral ganglia, as well as in the lateral cords. The larger ganglion cells of the last two types are })robably motor in function. In some Lineidte a yet larger ty])e of cell may be j)r(\scnt in the vential ganglia, and sometimes also in the lateral cords. They possess 17 vol V large axis cylinders, which have been termed neuro- rhords. They are not ])ri'sent in 7J/i layers of the I)ody wall. (4) The proboscis sheath nerve — an exceedingly fine nerve situated just beneath the circular muscle layer in the median dorsal line. It probably innervates the stiucture from wliicii it i-eceiv(>s its name. 18 (5) The proboscis nerves which are given off one on cacli side of the vontial g'ang'lia and pass thence inio Uio piohoscis. Insidi> this stnictuie they soon spread out and fuse to form a ncivous sheath investing- the prohos- cidial epithelium in the retracted state of the organ. On the nerves of the ])eripheral system are found some nuclei, hiit these prolnihly helong to neuroglial, not ganglion cells. S]:XSE ORGANS. The ciliated cells of the epidermis douhtless function as sensory cells, though whether the sensory elements can be distinguished a])art from the ordinary ciliated cells has not been determined in the case of Linens f/esserens/'s. Some observers, however, have been able to distinguish such cells in other species. Apart from these, three forms of sensory organs are found in the present species. (1) The cerebral organ. It has already been noticed that on either side of the head there is a groove bounded l)y mobile lips, reaching from the tip of the head nearly to the mouth region, and deepening as it passes back- Avaids. At the posterior extremity of each of these head slits (PI. IV., fig. 1, hs.) is a small aperture marking the opening of a fine blind canal which, taking first a back- ward and then a forward coui-se (PI. III., fig. •!, re), lies for its whole ext(Mit in clo.-e pioximity to the hinder portion of the dorsal ganglion. Into it open two sets of glands. The lirst set (PI. I., fig. -•;, and PI. III., fig. ••), "(■e of which are niiiuit(> dei^ply-stained o'ranules. On the outer side of the canal the cells are even iuor(^ higdih- specialised, and are ti\e in nuniher (as ^eell in transverse sec'tion), viz., a median one, two smaller ones on either side, and two very larg-e ones again on either side of these. More than one nucliMis is present in all of them. The cells of these five rows are separate at their inner ends both from one another and from the cells of the internal half of the canal. Like the latter they possess an inner hyaline poition, consisting of fused cilia projecting- into the lumen of the canal. The basal portions of all these cells are without a well-niaiked limiting memlnane, and come into close contact with the tibrous core of the posterior extremity of the dorsal ganglion. A'unibers of small sensory ganglion cells of the tirst type ( page 1() ) are massed round the canal (PI. I., fig. (i, and PI. II., tig. 4), and the projection of the tibrous core from the dorsal ganglion. The fuiK-tion of this elaborate organ is still j)i'o- blematical. liy some writers it has been supposed to contribute to the respiration of the brain lobes, though the specialised character of its epithelium and the numlier of ganglion cells in it Avould seem to lend more counte- nance to the view that its function is rather concerned with the elaboration of some special sensory impulses. (2) The frontal organ consists of three small projecting patches of high columnar glandless epithelium bearing cilia (PL lY., fig. 1, fr.). The median patch is situated just above the proboscis pore with a lateral patch on either side of it. These patcdies are retractile, and aftei' 20 ])i('seivatioii api)pai- in sortion as small pits (PI. I., JAg. 1). ()j)(>iiin<>' ii(>a)' tliom aio the so-calliMl lioad glands, wliicli in J. //If IIS f/rssn-('iisis foini a small mass of <2;'land ccdls lying" in the anterior ])oition of the snout jnsi above ilie iliyncliodaMim. (•!) The eyes vary in nuiul)er, the adult animal usually liaYinc" a dorso-lateial row of ahout live on eacdi side. 'I'hev lie inil)edded in the tissue of the snout well below the epidermis and dorsal to the head slits. Each eye consists of a deep layei- of cells containing a dark hrowii l)igment, ovei- which is a layer of j)yriform cells (PI. III., fig 4), whose more pointed ends are drawn out into long processes which are inserted into a hne nucleated mem- brane. On the long processes of these ganglion (r') cells may often l)e seen minute deeply-staining bodies, whilst between them is a clear fluid kept in by the fine limiting membiane and forming a lens. The eyes are supplied by some of the cephalic nerves which enter them from the pigmented side. Instances may frequently be observed in which Iavo eves are incompletely separated, Avhence it may pi'obably be inferred that their numl)er is augmented by division of those already existing. The young JJiin/s when hatched has but a single eye on either side. THK COXXECTJVE TISSFES. These have l)een studied in tJie jiresent species by ^lontgomery (loc. cit. p. 1), Avho distinguishes the follow- ing kinds : — (1) ]}ran(died connective tissue cells with inter-cellular substance, composing the basement membrane of the external epithelium, the outei' and inner neurilemma, the slu>aths around the muscular hbres, tlu^ layer immediately sui rduiid iiiti' the intestine, the laver outside the endothe- 21 liuiii of the blond vossols, and probably also the onveb)])- iiig iiieiubrane of the g-onads. Tlie inteicelliilar substauco formed by these cells is struetuieless and of g-elatinons appearance, and takes a deep colour with many staining- reagents (especially hannatoxylin or nigrosin). (2) Pigmented connective tissue without intercellular substance. This occurs in the cutis, and consists of meuLbraneless cells Avith tiue branching fibrils containing greenish yellow [)igment granules. It is more plentiful on the dorsal surface where the colour is darkest. 'J'h(» amount and distribution of this pig'ment probably deter- mines the colour variety {i.e., whether red or green), since a greater amount of pigment usually occurs in the red variety. On this view the red colour must be looked up(ui as due to the refraction of light rays coming from the greenish pigment. (3) Mesenchyme tis.sue composed of bi- or multi-polar cells without intercellular substance. This tissue is much reduced in the present sjjecies, being only found in the anteiior region of the l)ody between the proboscis sheath and the (esophagus. (4) Parenchyme tissue consisting of large, much-vacuo- lated cells with an outer membrane. 'I'his occurs round the dorsal and lateral blood vessels in the intestinal it'gicm (PL I\'., fig. '-vM, though it is not pi-esent on the conimissurul vessels. BODY C.WITV .\XD ca).\Al)S. While some t)bservers hold that no body cavity is present in the Xemerteans, otheis consider that it is repre- sented 1)y spaces sometimes found lound the alimentary canal, and in which occur mesenchyme cells. Such spaces ai'c in some species w(dl marked with the mesenchyme cells so arranged as to foiia a more or less (lefinitc lining 22 ]U(Miil)iaue. Tt is possible, however, that the\- luav he due to shrinkage in the j)iocess of })ieservatiou. ^\s has alieady been seen, the only space of this kind which occurs in Linens f/esserensis is a small one between the proboscis sheath and the oesophagus. The space between the intes- tine and the inner longitudinal muscle la^'er is small, and is occupied by connective tissue cells and their inter- cellular substance. In this s[)ace occur sacs alternating with the intestinal diveiticula (IM. III., figs. (I and 8), and with the intestinal diverticula (PI. III., tigs, (i and 8), and lined ])y connective tissue cells. These are the gonads whose cavity, ap})arent iu the young animal, becouics obliterated in the matui-e worm by the sexual cells which fill it, and whitdi aie probably derived from the connective tissue cells which form its lining. Each gonad jiossesses a duct which opens dorso-laterally (PL III., fig. 8, fj(f.), and which is foimed partly bv a prolongation of the con- nective tissue lining of tln^ gonad, and partly from an ectodermal depression . The sexes are separate, and in the breeding season, which lasts fiom about Febiuary till -lune, the female dejiosits her ova under stones in a long tubular gelatinous cord. In the walls of this coid are the ova contained in small Hask-shaped transparent capsules (PI. III., tig. 7). One of the gelatinous cords produced by a single female usually contains a hundred or more of these little capsules, and each capsule contains the con- tents of a gonad, i.e., from one to seven ova, according to the size of the female. The spermatozoa of Linens (/cssei-eiisis possess a long pointed head (PL IV., fig. 4). The o-elatinous cord containing the ova is said to be the joint production of the male and the female. Into it the male then proceeds to discharge spermatozoa. Soon 23 afterwards the female deposits her ova, already surrounded by the characteristic capsule, which are then fertilized. The ova are deposited in the capsules, which are probahly secreted hv the lininc of the o>onad. DEVELOPilKXT.* The ova before fertilisation measure aboiit ■•'! mm. in diameter, and are opacjue owing to the numerous oily yolk g-ranules which they contain. The germinal vesicle is well marked, and in it is a large nucleolus or germinal sj)ot. After fertilisation segmentation is complete and regular, resulting in a blastula. A segmentation cavity is already present in the S cell stage. The blastula is covered with cilia by whose action the young embryo is kept in constant rotation. Invagination of the blastula then takes ])lace, and results in the formation of a typical gastrula. The differentiation between ectoderm and endoderm cells is now apjmrent, the latter being con- siderably- larger. The endodermal invagination is directed somewhat obli([uely (PI. IV., fig. 9), so that the future alim.entary canal lies entirely behind the blastopore, enabling one to distinguish already at this stage the anterior and posterior ends of the animal. The ectoderm of this stage does not directly l)ecom.' the ectoderm of the larva, but the latter is established bv a series of remarkable changes. In two small areas on either side the cells of the primary ectoderm of the gas- ti'ula divide lengthways forming palisade cells. These areas of secondaiy ectoderm, the cephalic and vential •'•The development of Linens gcsscrensis h;is been studied more especially by Desor, Barrois, M'Intosli, and Hubi-eeht. The account given by the last-named is the only one leased on modern methods, and has been followed in this paper. '24 plates, aic ili(Mi ovoi'p'i own liv ilic cells of the jiriiiiaiv iH'todeiiu (PI. I\'.. fig-. 7) sunounding- them, so that at these four areas the ectoderm becomes two-layered, viz., a layer of secondary ectoderm covered externally hy the layer of primaiv ectoderm, which has again become con- tinuous {vf. n. IT., fig. 9). At the anterior end, a fifth aiea of secondary ectoderm, the ])rol)oscidial plate, arises, though it differs from the others in being formed by (lelamiiiaiion, and not by sinking in (PI. IV., fig. 9, /')■/'.)■ The five areas of the secondary ectoderm then spread out and fuse with each other, forming a continuous coat which lies directly beneath, and subse(|uently becomes entirely separated from the 2:)rimary ectoderm. This secondary ectoderm eventually forms the ectoderm of the adult. The primary ectoderm is cast off later, degenerates, and is \itili>ed as food material by the embivo. IJefore the fusion of the five secondary ectoderm plates, however, two invaginations of the primary ectoderm are formed on either side of the blasto})ore (PI. IT., fig. S, forf/.). These later sink beneath the secondary ectoderm between the cephalic and ventral plates of the latter, and eventmxlly give lise to the ciliated canals of the cerebral organs. In the process these invaginations lose their communication with the exterior (PI. J\ ., fig. 10, coi-f/.), but late]' a secondary opening is formed in each case at the surface of the secondary ectoderm. Ai fhe time when the five plates of secondary ectoderm are commencing to appear the fiist traces of the future mesoderm are seen as cells l)udded off from both the primaiy ectodeini and the endoderm (PI. IV., figs. 7, (S, !)). AfttM- the estal)lishment of the secondary ectoderm as a continuous layei-, these mesodeiin cells come to be entirely enclosed within it. Meanwliile changes have been taking [)lace within the 25 endodenu. Whilst the secondaiy octodeini is making its appearance the hinder portion of the archenteron becomes shut off from the more anterior part by a coalescence of some of its cells (PI. IV., fig. 9), resulting in the forma- tion of a posterior intestinal portion with a closed cavitv, and an anterior (esopliageal i)ortion whose lumen opens to the exterior hx the blastopoie. The last-named eventually becomes the mouth of the adult. Later the cavities of the «soj)hagus and intestine become secondarily continuous, but before this occurs a small lateral evagina- tion is formed on either side of the inner portion of the cesophagus. These evaginations eventually lose their connection with the oesophagus, and acquire openings to the exterior through the body wall. In this way are established the nephridia. Thus the oesophagus must be looked upon as endodermal, and consequently the nephridia as diverticula of the archenteron. The anus is formed later at the posterior end of the intestine. During these changes the ectoderm of the proboscidial plate has formed an invagination, which will become the lining of the rhynchodieum and of the proboscis (PI. 1\'., hg. 10, /y.) The embryo now presents the appearance shewn in PI. IV"., fig. 5, and is known as Desor's larva. The fate of the ectoderm and of the endoderm has now been traced. With the exception of the gonads, the remainder of the body is derived from the mesoderm^ whose cells by this time have come together to form a continuous layer round the stru.ctures w^hose formation has already been described. The mesoderm gives rise to the connective tissues, muscles, nervous system, blood vessels, and proboscis sheath. The cavities of the two last are remains of the segmentation cavity. That the nervous system should be of mesodermal origin is a some- what remarkable fact. Still it has already been seen that 26 tlie mesoderm is derived in part from the primary ecto- derm, and it is possil)le that these cells are the ones con- cerned in the formation of the nervous system ; on which view its origin would be but ectodermal in disguise. An exception to the mesodermal origin of most of the organs was noticed above. This is the case of the gonads, which are stated to arise at a later stage as ectodermal ingrowths ventral to the level of the lateral nerve cords. The con- nection with the ectoderm is then lost, and the ducts are developed later above the level of the lateral nerves. The origin of the various organs has now been traced. During the later part of its stay in the egg capsule the larva lengthens considerably, until the little worm, now about 1"5 mm. long, forsakes the protection of its embryonic shelter to become an independent though microscopic unit in the teeming life around its birthplace. The development through the larva of Desor as sketched above is not the only form which occurs in the family of the Lineidse. In some other species of Nemerteans a free swimming pelagic larva, known as the Pilidium, is formed, and a slight knowledge of its develop- mental history throws some light upon the peculiar forma- tion of the ectoderm in Desor's larva. A typical gastrula is formed, M'hich then acquires a dorsal tuft of long, fused cilia and two lappets produced by ectodermal folds hang- ing down laterally on either side of the mouth. From its fancied resemblance to a helmet at this stage the larva derives its name. The young worm is then developed inside the Pilidium, in whose ectoderm five invaginations now make their appearance round the mouth, viz., two 2)aired and one anterior median unpaired. These invagina- tions lose their connection with the outer ectoderm 27 of the rilidiuni, and grow tog'ether to surround the endoderm. oi the hirva. When this process is complete the bod}' wall of the animal, exclusive of the mesoderm, now consists of four layers. Externalh' is the ciliated ectoderm of the larva, and internally the endoderm, whilst between these are two layers of ectoderm formed by the fused invaginations. Of these two layers the inner becomes the ectoderm of the adult worm, and corresponds to the secondary ectoderm of Desor's larva. The young ^"emertean continues to develop, and, when full grown, parts company Avith the remains of the Pilidium, which then consist of the original outer ectoderm and the outer layer of the fused ectodermal invaginations. Hence the discarded layer of primary ectoderm in Desor's larva corresponds to the ectodermal shell of the Pilidium which is cast off when the young A emertean escapes.* In several important respects the process of development by Pilidium is said to diifer from that by the larva of Desor. Among these may be more especially mentioned the origin of the nervous system and of the nephridia. The former is said to arise directly from the secondary ectoderm as local thickenings of this layer, whilst in the larva of Desor it has already been seen to take its origin from the mesoderm. Again in Pilidium development there is said to be an ectodermal cesophageal invagination when the nephridia arise, and these are consequently not of endodermal origin as in Desor's larva. There are also other differences, but the above are sufficient to show that Hubrecht's account (which has been followed above), though in its original form full and * An excellent account of Pilidium development, illustrated by numerous coloured diagrams, is given by L. Joubin in Les Nemertiens, Traitc do Zoologie de K. Blanchard, fascicule XI. Paris, 1897. •28 circumstantial, should be accej)ted with caution until confirmation has been received from other sources. REGENERATION. Though observations on the regeneration of lost parts do not exist in the case oi Lineus gesserensis, yet in a closelv allied species, Lineus sanguineus, interesting facts in this connection were brought to light by M'Intosh. When kept in captivity, examples of this species shew a tendency to rupture into man}- pieces. Each of these fragments may develop into a comjilete worm, both anterior and posterior ends being formed anew. PARASITES. Like most K^emerteans, Linens gesserensis is frequently infested with Sporozoan parasites. These occur chiefly in the intestinal region attached to the epithelium of the alimentary canal and hanging freely into its lumen. A curious large Mesozoan parasite has also been recorded in this species [Rhoijulura). It is found burrowing in the body wall, and its presence may Ije recognised, according to MTntosh, " by the perforated and honey- " combed appearance of the dorsum of the affected animal, " whose textures seem to be the seat of the workings of a " microscopic Tumicus typographicusy SYSTEMAIIC POSITION. The Nemerteans are divided by Burger into four orders based mainly upon the number of muscle layers in the body wall, and the position of the lateral nerve cords with respect to these layers. Briefly these orders, with the liritish families and genera l)elonging to each, are as follows : — 29 I. PROTONEMERTim. Two muscle layers in body wall, i.e., external circular and internal long-itiidinal. The lateral nerve cords lie outside the circular layer. Proboscis without stylets. Mouth behind brain. Fam. CarinelliDtE. Genus. Carinella. II. MESONEMEETINI. Two muscle layers in body wall, i.e., external circular and internal longitudinal. The lateral nerve cords lie in the midst of the longitudinal layer. Mouth behind brain. Proboscis without stjdets. Fam. Cephalothricid.I':. Grenera. Cephalothri.t Carinoma. III. METANEMERTINI. Two muscle layers in bod}' wall, i.e., external circular and internal longitudinal . The lateral nerve cords lie beneath the longitudinal layer. Mouth in front of brain. Proboscis armed with stylets. A. Proriiynchoccelomia. Body long and thin. Proboscis and pro- boscis sheath much shorter than body. Fam. EUNEMERTID^. Eyes j)resent. No otocysts. Genus. Eunemertes. Fam. OtotyphlonemertidtE. No eyes, but one or more pairs of otocysts venti'al to brain. Genus. Ototypldonemertes, 80 B. HoLORHYNCHOrOELOMIA. Body iisiially short. Prol)osois at least as long as body. Proboscis sheath i-eaches into hinder third of body. I'ain. Tetrastemmid.b. Four e^'es. Cere]>ral organs in front of lirain. Dioecious or hermaphrodite. Genera. Prosoroclimus. Tetrastemma. Fam. AMPTiTroKTD.E. Numerous eyes. C'erebral organs generally behind brain. As a rule members of this family are con- siderably larger than those of the preceding. Genera. Amphiporus. Drepan ophorus. Fam. Malac()]!Dkllid.e. Parasitic in Lamellibranchs. Sucker at posterior end. Genus. MalacohdeUa. rv. HETFRONEMERTINI. Three muscle layers in body wall, i.e., outer longitudinal, middle circulai, and internal longi- tudinal. Proboscis withoui stylets. ^foiith behind brain. Fam. Imi'oliidj-:. Without head slits. Genera. Enpolia Valenciniri . 0.ri/pol/ii. 31 Fam. LiNEiD-E. With kead slits. Genera. Linens. EuborJasia. Micrura. Cerelrratiil IIS. MicreUa. The genera of the Lineidte set down above are exceed- ingly difficult to define. The three genera Micrura, Cerehrntuln.'f and }f Icidia (together with the exotic genus Langin) agree with one another in the possession of a slender tail filament at the posterior end of the body. For this reason the}^ have been grouped together as Micrurse in opposition to the rest of the family, which are known as the Amicrurae. It is very doubtful, however, whether this caudal appendage is homologous in all the instances in which it is found, for in some cases the anus opens at its tip, whilst in others it opens at its base either just above or just below it. It also presents other anatomical differences in different species. The body form is regarded by some as aft'ording a character upon which to base generic distinctions. Especially is such the case in Cerehrahdus, of which genus the species are often characterised by their breadth, due chieflv to the sides of the animals being flattened out to form a kind of fin known as the side folds. Gradations between such a state and a more or less circular outline in section are found, so that the absence of well-marked side folds does not necessarily preclude a species from being relegated to this genus. Cerehrafulu.^ is also sup- posed to be characterised by a fine layer of diagonal muscles just outside the circular layer. This, however, is often absent. In fact at present the three genera. Linens, Cerehratulm and Micrura are exceedingly ill- defined. Many anatomical differences are found in the family, amongst which may more particularly be men- tioned the following:- — (a) A diagonal muscle layer, neurochord cells, eyes, and frontal organ may be either present or absent. (J)) A Avoll-marked cephalic vascular head loop may be present, or the cephalic vessels may form an anastomosing network. (e) The excretory system shews great variations in its backward extent ; the position of the tubules may be dorsal or ventral, or both ; they ma}' reach forward to the cerebral organ, or may com- mence some way behind it. Also some species possess a number of ducts whilst only one pair is present in others. In the majority of the Lineida^, and indeed in nian\ Jh'itish forms, we are as yet in ignorance with regard \(> many of these j^oints, and until they have been deter- mined it is useless to attempt to place the classification of the family upon a more satisfactory basis. 33 Explanation of Plates. Reference Letters. acg. anterior gLantl of cerebral organ. be. buccal vascular commis- sure. chv. commissural vessel between dv and Ihv. cc. ciliated canal of cerebral organ. cvic. circular cephalic nuiscles. evil, longitudinal cephalic muscles. en. cephalic nerve. coni. cerebral organ. cmil. glandular cutis. cvl. cephalic vascular loop. dc. dorsal commissure of brain. dg. dorsal ganglion. dv. median dorsal blood vessel. ep. epithelium. exd. excretory duct. ext. excretory tubules. fr. frontal organ. (jc. ganglion cells. gd. gonidial duct. liti. head slit. 7V/. intestinal diverticulum. iep. intestinal epithelium. Ibl. lateral l:)lood lacima. Ibc. lateral blood vessel. m. mouth. me. circular muscle layer. mcp. circular muscles of pro- boscis. mer. muscle cross. mdv. dorso- ventral muscles. ynli. internal longitudinal muscle layer. iiilu. external ditto. inlp. longitudinal muscles of proboscis. n. nerve to eye. nd. median dorsal nerve. }d. nervous layer. }dl>. nervous layer of pro- boscis. nuc. nuclei. ocg. ganglion cell (?j layer of eye. oes. oesophagus. ocsc. oesophageal nerve com- ixiissure. oed. oesophageal vascular lacunae. oesn. oesophageal nerve. oep. oesophageal epithelium. ogl. gland cells round oeso- phagus. or. ovary p. proboscis. par. parencliymatous cells. peg. posterior gland of cere- bral organ. ^4 pep. proboscis epithelium. Pil. pigment layer of eye. p}i. proboscis nerve. ps. proboscis sheath. /•(/. rhynchodaeum. rhc. rhynchocoelom. rhce. rhynchocoelomic epith- elium. adij. superior lobe of dorsal ganglion. ss. lateral nerve. vc. ventral commissure. X'ep. epithelium of blood vesse 1 vg. ventral ganglion. Fig-. Fig. Fig. Fig. Fig. Fig. Plate L 1. Transverse seotinn through the lip of the liead. X 60. 2. Transverse section taken between brain and tip of snout. X 45. 3. Transverse section through hinder part of brain, where the anterior gland of the cerebral organ opens near the end of the head slits. x 60. 4. Transverse section through dorsal commissure before the two limbs of the cephalic vascular loops have fused ventral to the proboscis sheath. x 45. 5. Transverse section through brain at a level between 4 and 8. x 45. 6. Transverse section through level of cerebral organ, buccal vascular commissuje and oesophageal nervous commissure. Plate II. I'ig. 1. Transverse section through mouth region. The (esophageal vascular lacunae are just com- mencing. X 45. Fig. 2. Transverse section through about the middle of the oesophageal region. x 45, 35 Fig. -1 Transverse section of epithelinm from anterior intestinal region. x .'JOO. Fig. 4. Longitudinal vertical section through brain taken rather to one side of the median line. X 45. Fig. 5. Transverse section through hinder region of proboscis. The circular muscdes and nervous layer have both disappeared. x 120. Fig. 6. Transverse section thioTigh ])roboscis at its widest — about the middle. x 80. Fig. 7. Longitudinal median section through anterior end, shewing the relations of the proboscis to. the rhynchodiBum and rliynchocceloiu when retracted. x 45. Somewhat schematic. Plate IIL Fij;. 1. Transveise section through so-called ciliated canal of cerebral or^'an, sliewinef the seven large external cells and the internal homo- geneous cell layer all with crystalline ends formed from fused cilia projecting into the lumen. x ."JOO. Fig. 2. Portion of intestinal epithelium, shewing the circular refractive l)odies enclosed in the elongated ciliated cells. x 108. Fig. ^. Schematic longitudinal horizontal section through the cerebral organ of a Heteronemer- tean. (After Biirger). Fig. 4. Section through eye just anterior to the entry of the nerve into the pigmentary layer, x 240. Fig. 5. Section through blind end of an excretory tubule (left poition), shewing elongated cilia, x 300. 36 Fig-, fi. Lonoitiulinal hoiizoiital section throug-h intes- tinal region, shewing the intestinal divertirnla alternating with the gonads. x 45. Fig. 7. Flask-shaped egg capsule, containing a single embrvo in the inornla stage. (After M'Intosh.) X 25. F'ig. 8. Transverse section through intestinal region passing between two diverticula. x 45. Plate IV. Fig. 1. Schem'atic figure, shewing ihe relations of the various systems in the anterior end of the animal as vieAved from above. The proboscis and its sheath, the oesophageal nerves and the buccal vessels have been omitted. Fig. 2. Transverse section through lateral blood vessel. On the side of the alimentary canal the paren- chyma cells are smaller and complete. On the outer side no cell wall is to be distinguished away from the vessel. x KiS. Fig. '■). Portion of oesophageal epithelium fiom a trans- verse section. Three kinds of gland cells are seen among the ciliated epithelium and below it: — (a), (S), and (h. (for explanation vide text). X 168. F'ig'. 4. Two s])ermatozoa (aftei' MTntosh). x 700. Fig. 5. Larva of ])esor as seen from the ventral surface. The outer ciliated coat is not yet slied. (After Barrois). Fig. n. Young Linens just hatciied. x 40. (After M'Intosh). F^igs. 7-lL Diagrammatic sections through larvjB of Lineus at different stages. (After Hubrecht). 37 Fig. 7. 'rransverse section, shewing' the secoudaiy epiblast of the cephalic plates [r/j.) g'ladiialh- overgrown by the piiniaiy [prep.). The pro- boscidial plate {pTp.) arises antero-dorsally by delaniination. Fig. 8. Transverse section of slightly later stage, shew- ing the two invaginations from the primary epiblast on either side of the blastopore, which will eventiially give rise to the cerebral organs [corg.). Fig. y. Longitudinal section through a stage slighlly younger than 7. The archenteron is sub- divided into intestine (int.) and oesophagus, which do not communicate. Fi<2:. 10. Horizontal section of older embryo. Proboscis now invaghiated and mesoblast accumulating. The secondary epiblast, consisting of proboscis, cephalic [pp-) and ventral plates {vp.), now forms the external surface of the worm, having sunk in, and become separated from the primary epiblast {p^'ep-}- The section corre- sponds to the stage shewn m figure 5. I'lg. 11. Median longitudinal section through somewhat later stage. The hinder portion of the pro- boscidian mesoblast is now attached. The oesophagus and intestine communicate. Rh}n- chocoelom now apparent [rhc). C. Tinliug & Co., Printers, 03, Victoria Street, Livcriiool, L.M B.C. Memoir YII. Plate I. fr... 'M rmi ep r>, ea '0:,., ^•> ca^l cml t '■'•'■ '-/x ^cugZ CTl ■ - ■'■■ ■ ^,s '-■'■;■,<>■- \ >m}- '■:-:;"':^^-;'H:'^^^-''':'\,;§^^ JMAii-i'' ,_ i *# /"/^V cvi ^*- St* dy... dc 7. 1. ci^qL. a.c^. |P Fig *'*^"?s^;:;;i'"'-'''v' ■■''".■,:a>"' F/g.5. \ ■' f/G. 6. R.CP.del vy. S.B.Iith LINEUS. L.M.B.C. Memoir YII. Plate 11. PS \\ ^tr. MA^AM^^*^ 1 (i W ) V r" w v'f^. J, F/gJ. TTlCr. Fj6 4 rhce 59" j«^; - V''v\t ^. :^''^":> 77z2/D. tTk. rhce ^s ^ep ti:. Fig 7 '■«M^ <^ fr H I jii'iTJ^.jaiaiBcrrr': R.CPdel. S.B.hth. LINEUS. L.M.B.C. Memoir VII. Plate HI. Fig/ ^4 '«'.'$ F,6 2 X^ %^|l^^ ,<, /^wr ^f ^y- <. fell ^/ n ^u ?!S. f/G.3. Fig. 4-. Fg6 PS- Fig .r<^^f'» - --■ — 7 .^yi-ofiVV \N » ' ; ;^5 r/^f^^" J€/7 FjgS ^\ Ibv a: mlt. S.B/ifh. LINEUS. L.M.B.C. Memoir VIT. Plate IV. (V i-' '-'< r — V; c*; ■s vT . /S^' .'a < ^^; /"/e.Z prep. prp. ■M 0 -'-^.-< tTu:. FigS R.C.Pdel '^^liS^^i^^^ ^e^ /y6' // SB/ith LINEUS. ^