POLYZOA is the name applied by J. Vaughan Thompson in 1830 (l)1 to a group of minute polyp-like organisms which were subsequently (1834) termed "Bryozoa" by Ehrenberg (2). The forms included in this group were stated by Thompson to be "in a general way the whole of the Flustraceas, in many of which I have clearly ascer-tained the animals to be Polyzose," they having been pre-viously considered by zoologists to be allied to the Hydra-like polyps. These organisms had previously been known by the hard corneous " cells " or chambers which are formed by the animals on the surface of their bodies, and build up, in consequence of the formation of dense colonies by bud-ding, complex aggregates known as " sea mats " and " sea mosses." Thompson expressly stated the opinion that the organization of the animals detected by him led to the conclusion that " they must be considered as a new type of the Mollusca Acephala." 1 These numbers refer to the bibliography at the end of the article.

Subsequently (1844) Henri Milne-Edwards (3) pointed out the relationship of Thompson's Polyzoa to the Brachio-poda, and, adopting the latter's view as to their Molluscan affinities, proposed to unite these two classes with the Tunicata in a group to be called " Molluscoidea." Re-cent researches have entirely separated the Tunicata from this association, and have demonstrated that they belong to the great phylum of Vertebrata. On the other hand the association of the Polyzoa with the Brachiopoda ap-pears at present to be confirmed, though the relationship of these two classes to the Mollusca has been shown to rest on mistaken identification of parts; see, however, Harmer (18).
The Polyzoa appear to be related to the Sipunculoid Gephyraean worms (Gephyrasa inermia) more nearly than to any other class of the animal kingdom. The study and interpretation of the facts of their ontogeny (growth from the egg) presents such extreme difficulty that in the pre-sent state of our knowledge it is necessary to regard them

ad interim as forming with the Brachiopoda and Sipuncu-loids an isolated group, to which the name " Podaxonia" may be applied, pending the decision of their affinities by the increase of our knowledge of the embryology of import-ant members of the group.
The forms included at the present day in Thompson's class of " Polyzoa " may then be thus classified :—
PHYLUM PODAXONIA. CLASS l.—SIPVNCULOIDEA. CLASS 11.—BRACHIOPODA. CLASS III.— POLYZOA.
Section 1.—VERMIFORMIA.
Sole genus : Phoronis (figs, 4 and 5). Section 2. —PTEROBRANCHIA. Genus 1: Rhabdoplenrct (fig. 7). Genus 2: Cephalodiscus (figs. 8, 9, 10). Section 3.—EUPOLYZOA. Sub-class 1.—Ectoprocta.
Order 1.—PHYLACTOL/EMA.
Examples: Lophopus, Plumatella (fig. 2, B), Cristatclla (fig. 3), Fredericella. Order 2.—GYMNOL^EMA. Sub-order 1.—Cyclostoma.
Examples: Crisia (fig. 13, A), Hornera, Tubulipora, Discoporella. Sub-order 2.—Ctenostoma.
Examples: A Icyonidium, Vesicularia, Serialaria, Bower-bankia (fig. 1, A), Paludicella (fig. 1, E and fig. 2, A). Sub-order 3.—Chilostoma.
Examples : Cellutaria, Scrupiocellaria, Kinetoskias (fig. 14), Bugula, Biccllaria, Fhistra (fig. 1, G), Mucro-nella (fig. 1, C, D, F), Membranipora, Lepralia, Fschara, Cellepora, Retcpora. Sub-class 2.— Entoprocta.
Genera: Pedicelliua (fig. 15), Loxosoma (fig. 16), Urna-tella, Ascopodaria. We shall most readily arrive at a conception of the •essential structure of a Polyzoon, and of the variations to which that essential structure is subject within the class, by first examining one member of the group in detail and subsequently reviewing the characters presented by the divergent sub-classes, orders, &c, above indicated.
The most convenient form for our purpose is Paludicella Ehrenbergii (fig. 2, A), belonging to the typical section of the class (the Eupolyzoa) and to the order Gymnolaema, The organism occurs as minute tree-like growths (figs. 2, A and 1, E) attached to stones in freshwater streams and canals. The branches of the little tree are rarely more than an inch in length, and are regularly swollen and jointed at intervals. Each of the very numerous joints is about one-fifth of an inch long, and is in reality a tubular horny box attached above and below to the preceding and succeeding joints, and having on one side of it a spout-like aperture from which a crown of tentacles can be protruded. Each joint is thus inhabited by a distinct animal which is more or less completely shut off from the one in front of it and the one behind it, although it originated from the hinder and has given rise to the fore-lying individual by a process of budding, and retains a continuity of substance with both. A single cell or joint with its contained animal is repre-sented in fig. 2, A.
Paludicella produces an arboriform colony, the main trunk or stolon being adherent to some stone or piece of wood. The substance of the wall of the cells is formed by a chemical body allied to chitin. Other Polyzoa may form mat-like expansions—the cells being placed in one plane, side by side (fig. 1, C, D, F, G), as well as in linear series; others again form solid masses, whilst many agree with Paludicella in the simple linear arrangement of their units. Phoronis and Loxosoma, on the other hand, do not form colonies at all—the former because it does not bud, the latter because the buds become detached from their parent as soon as formed, as do the buds of the Hydrozoon Hydra.
On the whole Paludicella presents us with a very simple form of Polyzoon-colony (technically termed a "zoarium"), in which the aggregate of budded persons, each of which

FIG. 1.—Various forms of zoaria of Eupolyzoa.
A. Bowerbankia pustulosa, one of the Ctenostoma; natural size.
B. A cluster of polypides of Bowerbankia pustulosa, some with expanded
tentacles ; more highly magnified.
C. Zocecia ot Mucronellapavonella (Chilostoma); highly magnified.
D. Zoarium of Mucronella pavonella, forming a disk-like encrustation on a
piece of stone; natural size.
E. Zoarium of Paludicella Ehrenbergii (Ctenostoma), natural size.
F. Zocecia of Mucronella Peachii ; highly magnified. Compare with C in order
to note specific characters.
G. Zoarium of Flvstra securifrons; natural size.
is called a "polypide," does not exhibit any marked indi-viduation, but is irregular and tree-like. But, just as in the Hydrozoa we find the Siphonophora presenting us with a very definite shape and individuality of the aggregate or colonv. so in the Polyzoa we find instances of high indi-

viduation of the zoarium of a similar kind. The most remarkable example is afforded by the locomotive zoarium or colony of Cristatella (fig. 3); and another very striking instance is that of the stalked zoaria of Kinetoskias (fig. 14) and Adeona.
The horny consistence of the cells which are produced by Paludicella is very usual in other Polyzoa; but we find frequently that the substance which forms the cells is gelatinous and soft instead of being horny, or again may be strongly calcareous. The term ccencecium is applied to the mass of cells belonging to a colony or zoarium when considered apart from the living polypides which form it. Often such ccencecia are found retaining form and structure when the soft living polypides have decomposed and dis-appeared. A single cell of the ccencecium, corresponding to a single polypide, is called by the special students of the Polyzoa a zoceciwn.

FIG. 2,—A. Polypideof PaludicellaEhrenbergii, seen as a transparent object in optical section and highly magnified (from Gegenbaur, after Allman). For natural size see fig. 1, E. a, anus; br, peristomial circlet of ciliated ten-tacles; i, thickened cuticle of the body-wall, forming the horny cell or zocecium; m, median retractor muscle of the introversible part of the body; r', anterior retractor of the same; mr, great retractor muscle of the same;
0, ovary, passing from which to the stomach is the anterior mesentery or funiculus ; t, testis ;
1, oesophagus; v, stomach; x, posterior mes-entery or funiculus; x', anterior mesentery or funiculus. Observe at the right upper corner of the figure the base of a second polypide and the "rosette-plate" of separation.
B. Diagram of a polypide of Plumatella. Letters as above.
If we examine a single cell or zocecium of Paludicella more carefully whilst its polypide is alive, we discover that the horny cell is nothing more than the cuticle of the polypide itself, to which it is absol-utely adherent. At the so-called " mouth " or spout of the cell the euticle suddenly changes its character and becomes a very delicate and soft pel-licle instead of being thickandhorny. There is no real discontinuity of the cuticle at this region, but merely a change in its qualities. This gives to that por-tion of the body of the polypide which lies beyond the spout a mobility and capacity for folding and pleat-ing which is entirely denied to that part where the cuticle is more dense (fig. 2, A). Accordingly we find that the anterior por-tion of the body of the polypide can be pulled into the hinder part as the finger of a glove may be tucked into the hand. It is, in fact, an " introvert" (for the use of this term see MOLLTJSCA, vol. xvi. p. 6 5 2). This arrange-ment is universal in the Ectoproctous Eupolyzoa, but does not obtain either in the Entoprocta, the Pterobranchia, or the Vermiformia. In Phoronis, Rhabdopleura, and Cepha-lodiscus the anterior part of the body can not be tucked or telescoped into the hinder part as it can in typical Eu-polyzoa. On the other hand it is very important to note that the Sipunculoid Gephyraaans are all pre-eminently characterized by possessing identically this arrangement. The introversion is effected in Paludicella (as in other Eu-polyzoa) by a series of long detached retractor muscles of considerable power (fig. 2, A, mr, r', m); the same is true of Sipunculus.
The view has been advanced by Allman (4) that the re-tractile part of the polypide is to be considered as a distinct individual budded from the basal portion, which is regarded as an equivalent individual. It does not appear to the present writer that such a theoretical conception tends to facilitate the understanding of the structure and relations of these animals.
An "ectocyst" and "endocyst" have also been distin-guished in former treatises, and these terms form part of a special " polyzoarial " nomenclature, but do not appear to be any longer needful. Equally undesirable is the misap-plied term "endosarc" lately introduced by Jolliet (5) to denote a certain portion of the Polyzoon structure which will not be referred to here by that name.
The retractile or introversible portion of the body of the polypide of Paludicella is terminated by a crown of sixteen stiff non-contractile tentacles (fig. 2, A, br) which form a circle around a central aperture—the animal's mouth. These tentacles are hollow and beset with vibratile cilia. The beating of the cilia causes a powerful current in the water by which food is brought to the animal's mouth. Each tentacle is also muscular, and can be bent and straightened at will. The tentacles not only serve to bring food into the mouth, but they are efficient as gill-filaments, being possibly homologous with (as well as func-tionally similar to) the gill-filaments of Lamellibranch Molluscs. They also serve as delicate tactile organs, and are the only sense organs possessed by the Eupolyzoa.
In Paludicella the platform around the mouth from which the tentacles arise, or lophophore, as it is termed, is circular. This is the case in all members of the large group of Gymnotema and in the Entoprocta. But in the Phylactolasma the lophophore is drawn out on each side, right and left, so as to present a horse-shoe shape (fig. 2, B), and in some forms, notably Lophopus and Alcyonella, the two arms or diverging rami of the horse-shoe are very strongly developed.
In the Pterobranchia the tentacles are confined in one genus (Rhabdopleura) to the two arm-like outgrowths of the lophophore, and are not simply hollow but contain a well-developed cartilaginoid skeleton (fig. 7). In the allied genus Cephalodiscus there are not merely a single pair of such arm-like processes, each bearing two rows of tentacles, but the lophophore is developed into twelve arm-like pro-cesses (fig. 9), which form a dense tuft of filaments around the anterior extremity of the animal.
In the Vermiformia (Phoronis) we again meet with a very perfect horse-shoe-shaped lophophore (fig. 4). The tentacles upon the crescentic or otherwise lobed circumoral region of the Sipunculoids are the representatives of the tentacles of the Polyzoa; whilst the tentaculiferous " arms " of the Brachiopoda appear to be the equivalents of the Polyzoon's lophophore much drawn out and in most cases spirally rolled.
Just below the circular crown of tentacles in Paludicella we find an aperture which the study of internal anatomy proves to be the anus. In all Polyzoa the anus has this position near the mouth; and in this respect we again note an agreement with Sipunculus and the other so-called Gephyrsea inermia. In one division of the Polyzoa alone is there any noteworthy variation in the position of the anus, namely, in the Entoprocta (sub-class of the section Eupolyzoa). In these forms the anus, instead of lying just below the lophophore or platform from which the tentacles spring, is included like the mouth within its area (fig. 15, C).
Passing now to the deeper structure of Paludicella, we find that it is a Coelomate animal; that is to say, there

exists between the body-wall and the wall of the aliment-ary tract a distinct space termed "perigastric space," "body-cavity," or "ccelom." This is true of all Polyzoa, though it has been erroneously stated by G. O. Sars that Rhabdopleura does not possess such a ccelom. In Eu-polyzoa (excepting the Entoprocta) the ccelom is very capacious; it is occupied by a coagulable hsemolymph in which float cellular corpuscles, and also the generative products, detached, as is usual in Ccelomata, from definite "gonads" developed on its lining membrane (fig. 2, A, o, t). This lining membrane or " ccelomic epithelium " is ciliated in the Phylactolaema, but its characters appear not to have been definitely determined in other Eupolyzoa. The ccelomic space and the tissues bounding it are continuous throughout the colony or zoarium of a Polyzoon—either directly without any constriction marking off one polypide from another, or through perforate septum-like structures as in Paludicella (see right-hand upper process of fig. 2, A), which form incomplete barriers between juxtaposed zooecia, and are termed " rosette-plates " or " communication-plates." The ccelomic cavity is continued in Paludicella and probably in all Polyzoa into the tentacles, so that these organs expose the haemolymph fluid to a respiratory action, and hence may be called branchial.
The body-wall of Paludicella consists, alike in the anterior introversible region and in the posterior region, of an outer cuticle which has already been spoken of as ! thickened around the base of the polypide so as to become ' there the hard tube-like zooecium. Beneath this is the delicate layer of living epidermic cells which are the j mother-cells or matrix of that cuticle. Beneath this again are a few scattered annuli of muscular fibre-cells arranged ring - wise around the cylindrical body; more deeply placed than these are five large bundles of longitudinally placed muscular fibre-cells which are attached at three different levels to the soft introversible portion of the body, and by their retraction pull it in three folds or tele-scopic joints into the capacious hinder part of the body. In some Polyzoa the muscular fibre-cells present trans-verse striations. These folds are shown in fig. 2, A;

FIG. 3.—The locomotive zoarium of the freshwater Phylactolffimous Polyzoon Cristatelta mucedo; magnified six times linear (after Allman). a, individual polypides with their horse-shoe-shaped crown of tentacles exserted; o, stato-blasts seen through the transparent tissues; c, the muscular foot or base of the colony by means of which it crawls; d, portion of water-weed upon which the Cristatella is crawling.
but when the longitudinal muscles are completely con-tracted the tentacular crown would be pulled down far out of sight into the midst of the body by the great longitu-dinal muscle mr. Deeper than the longitudinal muscles, and clothing them and everything else which projects into the ccelom, is the ccelomic epithelium, not easily observed, and sufficiently known only in the Phylactolaema. Part of it gives rise to the generative products (fig. 2 A, o, t). Other Eupolyzoa have a similar but not identical arrange-ment of the longitudinal muscles—acting essentially as retractors of the " introvert " or soft anterior region of the body—and a similar structure of the body-wall which is in essential features identical with that of the Sipunculoid
worms, the Chaetopod worms, and other typical Ccelomate
animals. *
The alimentary canal of Paludicella forms a closely com-pressed U-shaped loop depending from the closely approxi-mated mouth and anus into the capacious ccelom. It is clothed on its ccelomic surface (in Phylactolaema at any rate) with ccelomic epithelium, and beneath this are extremely delicate muscular layers. Within it is lined, except in the immediate region of the mouth (which is lined by the in-pushed outer cell-layer), by the enteric cell-layer—the digestive cells derived from the archenteron of the embryo. We can distinguish in Paludicella a contrac-tile pharyngo-cesophagus (fig. 2, A, as), a digestive stomach v (the lining cells of which have a yellow colour), and an intestine which forms that arm of the loop connected with the anus. This simple form of alimentary canal is uni-formly present in Polyzoa. In Boiverbankia and its allies a muscular gizzard with horny teeth is interposed between oesophagus and digestive stomach.
The alimentary canal of Paludicella does not hang quite freely in the ccelomic cavity, but, as is usually the case in other classes where the ccelom is large, mesenteries are present in the form of fibrous (muscular 1) bands clothed with ccelomic epithelium and suspending the gut to the body-wall. In Paludicella there are two of these mesen-teries, an anterior (x) and a posterior (x). The presence of two mesenteric bands is exceptional. Usually in the Eupolyzoa we find one such mesentery only, corresponding to the hinder of the two in Paludicella. The special name funiculus (Huxley) is applied to this mesenteric band, and it is noteworthy that the cells of the ccelomic epithelium, either upon its surface or at its point of insertion into the body-wall, are modified as reproductive elements, forming either the testis or ovary; in the Phylactolaema they form here also special asexual reproductive bodies, the stato-blasts. The nervous tissue and organs of Paludicella have not been specially investigated, but in many Eupolyzoa an oval mass of nerve-ganglion cells is found lying between the mouth and anus, and there is no doubt that it is present in this case. In Plumatella nerve-fibres have been traced from this ganglion to the tentacles and other parts around the mouth (fig. 11, w, x, y). A "colonial nervous system" was described some years ago by Fr. Miiller in Serialaria; but modern histologists do not admit that the tissue so named by Miiller is nerve-tissue. The ganglion above mentioned is the only nervous tissue at present known in Polyzoa (but see fig. 17, x).
No heart or blood-vessels of any kind exist in Paludi-cella nor in any of the Eupolyzoa or Pterobranchia. On the other hand the isolated vermiform genus Phoronis presents a closed contractile system of longitudinal vessels (dorsal and ventral) which contain nucleated corpuscles coloured red by haemoglobin (figs. 4, 5).
No excretory organs (nephridia) or genital ducts have been observed in Paludicella, nor have such organs been detected in the majority of the Polyzoa which have been studied. In the Entoprocta, however, a pair of minute ciliated canals are found in the nearly obliterated body-cavit}' opening to the exterior near the tentacular crown in both Pedicellina and Loxosoma, which represent the cephalic-nephridia of worms. A definite pair of nephridia occur in Phoronis. A similar significance is perhaps to be attributed to the " intertentacular organ " of Farre—a ciliated pas-sage opening between two tentacles of the lophophore in Membranipora, Alcyonidium, and other forms—through which Hincks has observed the spermatozoa to escape in largo numbers. This organ occurs equally in female speci-mens of Membranipora, and is not therefore simply a sper-matic duct.

Paludicella, as we have seen, develops both ova and spermatozoa in one and the same polypide. The details of impregnation and development have not been followed in this instance, but in some of the marine Eupolyzoa (Gymnolaema) remarkable bud-like structures termed ocecia are developed for the special reception of the ova, and in these organs fertilization takes place. In the Entoprocta there is a peculiar brood-pouch. The spermatozoa of one polypide probably in all cases fertilize the ova of another, but we have not yet in many cases a knowledge of how the spermatozoa get to the eggs, or how the eggs escape from the body-cavity of the parent. In the hippocrepian freshwater Polyzoa (Phylactolaema) the ova appear to be fertilized and undergo the early stages of development within the body-cavity of the parent or in a hernia-like protrusion of it. Probably in such cases the embryos escape by the death of the parent and rupture of the parental tissues, as do also the peculiar asexual internal buds or statoblasts of these forms.
The embryo Polyzoon or " larva" swims freely in its early condition by means of cilia, and is in this condition a single polypide or " person." The forms assumed by these ciliated larvae in different Polyzoa are very various and exceedingly difficult of interpretation. We shall have more to say with regard to them below (see figs. 19, 20, 21). The ciliated larva then fixes itself and commences to produce polypides by a process of budding, the buds remaining not merely in contact but in organic continuity, and increasing continually in number so as to form a large colony or zoarium. In Paludicella we have seen that this colony has a simple tree-like form. The new buds form as wart-like growths, usually one, sometimes two in number, at the free end of a cell or zooecium near the spout-like process from which the tentacular crown is everted. In Paludicella all the polypides of a colony are alike; there is no differentiation of form or distribution of function amongst the members of the colony. In many Eupolyzoa this simplicity is by no means maintained, but a great variety of form and function is assumed by various members of the aggregate. The only approach to a differentiation of the polypides in Paludicella is in the arrest of growth of some of the buds of a colony in autumn, which, instead of advancing to maturity, become conical and invested with a dark-coloured cuticle. They are termed hybernacula. Should the rest of the poly-pides die down in winter, these arrested buds survive and go on to complete development on the return of spring.
In Paludicella we have thus seen a fairly simple and central example of Polyzoon structure and life-history. The variations upon this theme presented in different groups of Polyzoa have been to some small extent noted in the preceding account, but we shall now be able to indicate them more precisely by considering the various groups of Polyzoa in succession. The limit assigned to this article necessitates very large omissions. The reader who wishes to have the fullest information on the many difficult and uncertain matters connected with this subject is referred to Allman, Freshwater Polyzoa (Ray Society, 1856); Hincks, British Marine Polyzoa (Van Voorst, 1880); Haddon, "Budding in Polyzoa," Quart. Journ. Micr. Sci., 1883 ; Balfour, Embryology, vol. i. p. 242 ; and the original memoirs cited by these writers.
THE VERMIEORMIA.
The first section of the Polyzoa comprises but a single genus, Phoronis. It differs from all other Polyzoa first in its greater size (species 2 inches long are known) and elaboration of organization, and correlatively with that in the fact that it does not produce buds. Further,

it does not produce a closely adherent cuticular zooecium as do Paludicella and the Eupolyzoa generally, but a-leathery tube in which the animal freely moves, resem-bling that of some Cháetopods (Sabel-la). Like some Sabellae, Phoronis forms closely packed aggregates of indi-viduals not brought together by any process of budding, but each separately developed from an egg. Phoronis has an elongate, worm-like, unsegmented body, with a conical posterior termina-tion (like Sipuncu-lus), and anteriorly-provided with a horse - shoe - shaped crown of tentacles surrounding the mouth (figs. 4, 5). There is an inter-tentacular " web " between the bases of the tentacles as in the Phylactolae-ma. Caldwell (6) has recently shown that the tentacles are supported by a
ventral vessel; g, g, two anterior vessels which unite to form f) i, longitudinal muscular coat of the body-wall ; k, interrentacular membrane.
mesoblastic skele- ^IG' ^—Ptioronis hippocrepia, Wright; magnified six times linear (from Allman). a, horse-shoe-ton, as IS also the shaped lophophore with tentacles; c, epistome case in Rhabdo- (P™-0™1 lol)<; or prostomium); d, oesophagus;/,
pleura, but appar-ently not the case in any other Polyzoa. Close to the mouth, as in all Polyzoa, is placed the anus, outside the horse-shoe-shaped lophophore or tenta-cular platform (fig. 11, i). The tenta-cular crown is not introversible; in this point Phoronis differs from Paludicella and the Ectoproctous Eu-polyzoa, and agrees a^ with the Entoprocta and the Pterobranchia. Overhanging the mouth is a small prae-oral lobe or " epi-stome " (figs. 4, 5, c). This organ is aborted in Paludicella, and in-deed in all the Gym-nolaema, but is present in the other Polyzoa,
i • „'oll lo ,v FlG* —L^1"111 view °f tne anterior region of"
ana IS especially large phor0nis. The tentacles of the right arm of the
and Well developed in lophophore are cut short in order to expose clearly
-P,, , , , n the mouth b and the overhanging "epistome"
Jinabaopleura and Ce- or prai-oral lobe c. e, intestine ; h, dorsal vessel.
phalodiscus. It has other letters as m fig. 4. been compared to the Molluscan foot, but undoubtedly in Phoronis it is the persistent representative of the prse-oral

lobe of the larva (fig. 6), and therefore cannot be compared to the Molluscan foot. If we are right in associating Phoronis with the Polyzoa, this fact is sufficient to show that the epistome of the Phylactokema (fig. 11, e) and the buccal shield of Ehabdopleura (fig. 7, d) and of Cephalodis-cus (fig. 9, b) are also cephalic in nature, and cannot rightly be identified with the post-oral and ventral muscular lobe known as the foot in Mollusca. A circum-oral nerve ring •occurs at the base of the tentacles and sends off a cord which runs along the left side of the body. The alimen-tary canal presents the same general form and regions as in Paludicella. It hangs in the body-cavity, to the walls of which it is suspended by definite mesenteries.
Phoronis presents a closed contractile vascular system 'containing red-coloured blood-corpuscles (figs. 4, 5, /, g, h). A pair of ciliated canals acting as genital pores is found near the anus; these have been shown by Caldwell to be typical nephridia.
The development of Phoronis is remarkable. The egg gives rise (after the usual phases of cleavage and gastrula-tion) to the larval form known as Actinotrocha (fig. "%). This larva possesses a hood-like region overhanging
A A.

FIG. 6.—Development of Phoronis and typical ciliate larvae. (1), (2), (3), (8), (9), (10), stages in the development of Phoronis—(1), earliest larva; (2), lateral view of the Actinotrocha; (3), ventral view of the same; (8), the ventral in-vagination iv is formed; (9), the ventral invagination is everted, carrying with it a loop of intestine; (io), the permanent relations of mouth, anus, and body (Podaxonia) are attained. (4), (5), Echinoderm larva with arehitroch, as in Actinotrocha, but band-like, not digitate. (6), Echinoderm larva, with the arehitroch divided into apraj-oral cephalotroch (Molluscan and Rotifer's velum), and a post-oral branchiotroch. (7), Chaetopod trochosphere larva with cephalo-troch only, and elongation and segmentation of the oro-anal axis, a, anus; o, mouth; pr, prostomium ; iv, ventral invagination of Phoronis larva. A B, oro-anal axis; VD, dorso-ventral axis.
the mouth and a number of ciliated post-oral processes or tentacles. The anus is placed at the extremity of the elongate body opposite to that bearing the mouth and prse-oral hood. The prae-oral hood becomes the epistome, and the tentacles, by further development (new tentacles replacing the larval ones), become the horse-shoe-shaped group of tentacles of the adult. A very curious process of growth changes the long axis of the body and results in the anus assuming its permanent position near the mouth. An invagination appears on the ventral face of the larva between the anus and mouth, and attains con-siderable size. At a definite moment in the course of growth this invagination is suddenly everted, carrying with it in its cavity the intestine in the form of a loop. Thus a new long axis is suddenly established at right angles to the original oro-anal axis, and continues to de-velop as the main portion of the body. The short area extending from the pras-oral hood to the anus is thus the true dorsal surface of Phoronis, whilst the elongated body is an outgrowth of the ventral surface perpendicular to the primary oro-anal axis, as conversely in many Mollusca we find a short ventralr area (the foot) between mouth and anus, and an outgrowth of the dorsal surface (the visceral hump) perpendicular to the primary oro-anal axis, forming the chief body of the animal. In these relations Phoronis (and with it the other Polyzoa) agrees with Sipunculus. On the other hand Echiurus, the Chaetopods, Nemertine worms, and some other groups which start from a simple larval form not unlike that of Phoronis, present a continual elongation of the original oro-anal axis, and no transference of the long axis by the perpendicular or angular growth of either the ventral or the dorsal surface of the larva.
Phoronis was discovered originally in the Firth of Forth by Dr Strethill Wright. It occurs in the Mediterranean and in Australian seas (Port Jackson).
THE PTEROBRANCHIA.
This section of the Polyzoa also comprises forms which differ very widely from Paludicella. Inasmuch as their development from the egg is at present quite unknown, it may possibly prove that they have other affinities. Only two genera are known, Ehabdopleura (Allman) and Cephalodiscus (M'Intosh), the. former dredged by Dr Norman in deep water off the Shetlands (and subse-quently in Norway), the latter taken by the " Challenger " expedition in 250 fathoms off the coast of Patagonia.
The Pterobranchia have the mouth and anus closely approximated, and immediately below the mouth are given off a series of ciliated tentacles, but these do not form a complete circle as in Paludicella, nor is the lophophore (the platform of their origin) horse-shoe-shaped as in Phoronis. The lophophore is drawn out into a right and a left arm in Ehabdopleura (fig. 7), upon each of which are two rows of ciliated tentacles; no tentacles are developed centrally in the region between the two arms, so that the mouth is not completely surrounded by these processes. The horse-shoe-shaped lophophore of Phoronis could be modified so as to represent the tentaculiferous arms of Ehabdopleura by suppressing both rows of tentacles at the curve of the horse-shoe, and leaving only those which occur on the arms or rami of the horse shoe (see fig. 4). The lopho-phore of Cephalodiscus presents us with twelve processes, each carrying two rows of ciliated tentacles; in fact we have six pairs of tantaculiferous arms instead of a single pair, and each of these arms is precisely similar to one of the arms of Ehabdopleura (fig. 9), excepting that it terminates in a knob instead of tapering. There is no arrangement for introverting the anterior portion of the body into the hinder portion in the Pterobranchia.
The little epistome or prae-oral lobe of Phoronis is repre-sented in the Pterobranchia by a large muscular shield or disk-like structure (fig. 7, d and fig. 9, b) which over-hangs the mouth and has an actively secreting glandular

surface by which the tube or case (tubarium) in which the polypide is enclosed is secreted.
Both Rhabdopleura and Cephalodiscus produce colonies by budding; but the colonies of the former are large, definite, and arborescent, whilst those of Cephalodiscus are remarkable for the fact that the buds do not remain long in organic continuity with their parent, but become detached and nevertheless continue to be enclosed by the same common envelope or secretion. The bud-formation of Bhabdopleura recalls that of Paludicella in the fact that it leads to the formation of continuous arboriform com-munities. That of Cephalodiscus resembles the budding of Loxosoma, since no two fully-formed individuals remain

FIG. 7.—Rhabdopleura Normani, Allman (original drawings, Lankester). A. A single polypide removed from its tube and greatly magnified, a, mouth; b, anus ; c, polypide-stalk or gymnocaulus, the " contractile cord " of Sars ; d, the praa-orai lobe (buccal shield or disk of Allman); e, intestine; /, thoracic region of the polypide; g, one of the ciliated tentacles. B. Lateral view to show the form of the buccal shield and its pigment spot, g, ciliated tentacle (in outline); h, basal ridge of the right arm of the lophophore. C. Lateral view of a polypide. z, ciliated patch (Sars's organ) at the base of the lophophore-arm. Other letters as above. D. Part of a lophophore-arm, with soft tissues rubbed off to show the cartilaginoid skeleton, a, epithelium and soft tissues still adherent at the tip of a tentacle; i», skeleton of tentacle; c, skeleton of axis. E. Portion of a colony of Rhabdopleura Normani, showing the branched tube-like cases formed by the polypides. The black line within the tubes represents the retracted polypides connected together by their common stalk, the pecto-caulus. Magnified to three times the size of nature.
in organic continuity. Both Bhabdopleura and Cephalo-discus (like Phoronis) produce cases or investments in which they dwell. These are free secretions of the organ-ism, and are not, like the ccencecia of Eupolyzoa, cuticular structures adherent to and part of the polypide's integu-ment. The dwelling of Bhabdopleura is a branched system of annulated tubes of a delicate membranous con-sistency, each tube corresponding to a single polypide, the rings of which it is built being successively produced at the termination of the tube by the secreting activity of the prae-oral disk (fig. 7, E). The polypides freely ascend and descend in these tubes owing to the contractility of their stalks. On the other hand the dwelling of Cephalodiscus
is a gelatinous, irregularly branched, and fimbriated mass (fig. 8), excavated bynumerous cavities which communicate with the exterior. In these cavities are found the nu-merous detached small colonies of Cephalodiscus (fig. 9),or we should rather say the isolated budding polypides. The remaining important feature in the organization of the Ptero-branchia, namely, the parts connected with the forma-tion of buds, are best un-derstood by first examining Cephalodiscus. The body of Cephalodiscus is seen (fig. 9) to be an oval sac; in this is suspended the U-shaped alimentary canal, and from the walls of its cavity (ccelom) the ova and the spermatozoa are de-veloped. Projecting from the ventral face of this oval sac is a muscular cy-lindrical stalk, into which the viscera do not pass, though the ccelom is con-tinued into it (fig. 9, c). This stalk is merely the outdrawn termination of the body.
Turning to Bhabdopleura, we find that each polypide has a body of similar shape and character to that described for Cephalodiscus, and a similar ventrally developed " stalk " (fig. 7, A, c). But, inasmuch as the buds deve-

loped on the stalk of a Bhabdopleura polypide do not detach themselves, we find that we can trace the stalk of each polypide of a colony into connexion with the stalk of the polypide from which it was originally budded, which may now be considered as a "branch" bearing many-stalked polypides upon its greatly extended length, and such a " branch-stalk " may be further traced to its junc-tion with the " stem-stalk" of the whole colony. The stem-stalk was at one time the simple terminal stalk of a single polypide, but by lateral budding it gave rise to other polypides, and so became a gemmiferous " branch "; and further, when some of these in their turn budded and became branches, it became the main " stem " of a copious colony.

A serious error has been made in comparing the contrac-tile stalk of the Pterobranchiate polypide to the "funi-culus " or cord-like mesentery of Eupo-lyzoa. With this it has morphologi-cally nothing in common, since it is not an internal organ, but simply the elongated termination or stalk of the body, comparable to the stalk of Pedi-cellina (fig. 15) and Loxosoma (fig. 16), or to the hydrocaulus of such a Hydro-zoon colony as Cordylophora. The stalk where it bears only very young buds, or none at all, as is always its condition in Cephalodiscus and in many polypides of a Rhabdopleura colony, may be called a "gymnocaulus"; when once its buds have devel-oped into full grown poly-pides, and it has elongated proportionally with their growth, it becomes a " pec-tocaulus"; that is to say, it is to that part of it which bears such polypides that this term may be conveni-ently applied. The peeto-caulus of Rhabdopleura, both in the form of branch and stem, undergoes remarkable change of appearance as com-pared with the gymnocaulus. It loses its contractility, shrinks, and develops on its surface a hard, dark, horny cuticle (whence its name), comparable precisely in its nature to the hardened cuticle which forms the zocecia of Eupo-lyzoa. It now has the appearance of a black cord or rod-like body lying within and adherent to the inner face of the much wider tubular stem, and branches formed by the gradual building up and arborescent extension of the annulated tubarium secreted by the individual polypides. It has been regarded both by Allman and by Sars as a special structure, and called by the former " the chitinous rod" or "blastophore," by the latter "the axial cord."
In reality it is the black-coloured pectocaulus of Rhabdopleura which corresponds to the ccencecium of an ordinary Polyzoon; whilst the term " ccencecium" is totally inapplicable morphologically to the annulated branched tube in which the Rhabdopleura colony lives, this having absolutely no parallel in the Eupolyzoa.
A sac-like testis has been discovered in Rhabdopleura opening by the side of the anus (Lankester, 7); but the ova have not yet been seen, nor is anything known of its development. Similarly the eggs of Cephalodiscus are observed within the body of the parent in the " Chal-lenger" specimens, but nothing further is known of its life-history.
A body-cavity is present (Lankester), though its exist-ence has been denied by Sars and by M'Intosh. Neph-ridia and nerve ganglia are not described. Cephalodiscus has two remarkable eye spots dorsal to the cephalic disk (fig- 10, g).
THE EUPOLYZOA.
Whilst it is necessary to include in the group Polyzoa the forms we have already noticed as Vermiformia and Pterobranchia, there can be no doubt that those organisms to which we assign the name Eupolyzoa are primarily those upon which naturalists have framed their concep-tion of the group, and that they constitute a very con-sistent assemblage, held together by well-defined characters, and yet presenting an immense number of varied forms showing a wide range of modifications.
All the Eupolyzoa have closely approximated mouth and anus, and, like Paludicella, a complete range of hollow ciliate tentacles, describing either a circle or a horse shoe, surrounding the mouth. The anus as well as the mouth is included in this area in a few exceptional forms (the Entoprocta); it lies near but outside the lophophore (as the area is termed) in the vast majority (the Ectoprocta). Except in the Entoprocta, where the movement is limited, the whole anterior portion of the body bearing the lophophore can be invaginated into the hinder part (as described above for the typical Eupolyzoon Paludicella). This character distinguishes the Eupolyzoa from both Vermiformia and Pterobranchia. The polypides of all the Eupolyzoa are minute, but all produce buds which remain in organic continuity with their parent (except in Loxo-soma) and build up very considerable and sometimes massive colonies.
In all Eupolyzoa the cuticle of the hinder part of each polypide is thick and dense, thus forming a hard-walled sac, the zocecium. This is peculiar to and universal in the Eupolyzoa (except Loxosoma), and is not to be confounded with the non-adherent tubes of Phoronis and Rhabdopleura or the jelly-house of Cephalodiscus. The connected zocecia of a colony of Eupolyzoa constitute a ccencecium. A simple nerve ganglion between mouth and anus, a large body-cavity (except in Entoprocta), simple gonads without accessory glands or ducts, usually testis and ovary in the same polypide, absence of a blood-vascular system, of any but the most rudimentary nephridia, and of eyes, otocysts, or other special sense-organs, are features characterizing all adult Eupolyzoa.
The section Eupolyzoa, with its vast number of species and genera, requires a somewhat elaborate classification. The forms in which the anus is enclosed within the tentacular circle are very few, and are peculiar in other respects. We follow Nitsche (8) in separating them as the sub-class Entoprocta from the majority of Eupolyzoa forming the sub-class Ectoprocta.
Sub-class 1. Ectoprocta, Nitsche.
Eupolyzoa with the anus not included within the area of the lophophore. Anterior portion of the body of the normal polypide introversible. Tentacles not individually capable of being coiled or flexed.
Order 1. PHYLACTOLJEMA, Allman.
Ectoproctous Eupolyzoa in which the polypide possesses a prae-oral lobe or epistome, similar to that of Phoronis, and comparable to the more highly developed buccal shield or disk of the Pterobranchia. Lophophore (except in Fredericella, where it is nearly circular) horse-shoe-shaped (hippocrepian). Polypides of a colony equi-formal, that is, not differentiated in structure and function. Neighbouring zocecia are in free and open communication, the bud never becoming shut off by a perforated cuticular plate from its parent. Cuticle of the zocecia either gelatinous or horny, forming massive or else arborescent ccencecia, in one genus

(Cristatella) having the form of a plano-convex ellipse and locomotive (fig. 3). In addition to the multiplication of polypides in a colony by budding, and to the annual production of new individuals from fertilized eggs which initiate new colonies, a reproduction by internal buds called "statoblasts," comparable to the gemmae of Spon-gilla, has been observed in all the genera (fig. 3, b). The statoblasts are developed from the funiculus (mesentery), and are enclosed in ornate lenticular capsules of chitinous substance, characteristic in form in each species.
The fertilized egg of the Phylactolaema does not give rise to a zonociliate larva, but to a uniformly ciliate cyst-like diblastula, which develops directly and produces polypides by budding. The Phylactolaema are all inhabit-ants of fresh water (lacustrine).

f IG. 11.—Semi-ideal view of part of the lophophore of Lophopus and its tentacles, —intended to show the nerve-ganglion, nerves, and parts around the mouth. The tentacles have been cut away all along the right arm of the lophophore and from the inner margin of the left arm. c, foramen,placing the cavity of the epistome in communication with the body-cavity ; c', body-wall; d, mouth ; e, the epistome or pras-oral lobe; /, wall of the pharynx ; h, wall of the intestine; i, anus; k, lophophore ; I, a ciliated tentacle; r, elevator muscle of the epistome; tc. the nerve-ganglion ; x, x', nerves to lophophore and tentacles • y, nerve to pharynx.
The Phylactolaema include the genera Lophopus, Cristatella, Alcyonella, Plumatella, and Fredericella, which have been beauti-fully figured and described in Airman's classical Freshwater Polyzoa, Ray Society, 1856. The colonies of Lophopus are small, consist-ing of half a dozen polypides embedded in a massive glass-like coenoecium. Cristatella (fig. 3) is remarkable amongst all Polyzoa for its locomotive zoarium. Alcyonella forms massive ccencecia of many hundred polypides, as large as a man's fist. Plumatella and Fredericella are delicate arborescent forms commonly encrusting stones and the leaves of water-plants. All the genera known are British.
The Phylactolaema furnish a remarkable instance of a well-marked zoological group being confined to fresh water. Their reproduction by statoblasts (not known in the marine Polyzoa) appears to be related to the special conditions of lacustrine life, since it is also observed under the same exceptional conditions in the single freshwater genus of another great group of animals, viz., Spongilla. Also related to their non-marine conditions of life is the development of the fertilized egg, which, as in so many similar cases, does not produce the remarkable banded forms of locomotive larvae which are characteristic of their marine congeners.
Order 2. GYMNOMIMA, Allman.
Ectoproctous Eupolyzoa in which the polypide is devoid of any trace of the prae-oral lobe or epistome, whilst the

lophophore is perfectly circular. The polypides of a colony are frequently highly differentiated as avicularia, vibracu-laria, ocecia (egg-receptacles), and even as root and stem segments. The neighbouring polypides of a colony communicate (?) with one another by " rosette-plates" or " communication-plates "_—perforated areas in the walls of contiguous zocecia. The greatest variety in the character of the cuticle forming the zocecia (gelatinous, horny, calcareous) and in the grouping of the polypides, as well as in the shape of their zocecia, is observed in different sub-orders and families. In addition to the ordinary sexual reproduction, there are various modifications of the process of budding, the full exposition of which would necessitate more space than is here allotted, and is not yet indeed within the possibilities of present knowledge. The fertilized egg of the Gymnolaema gives rise to remarkable ciliate larvae of various forms (figs. 19, 20, 21), from which the first polypide of a colony is developed by an extraordinary and unexplained series of changes. The Gymnolaema are, with the single exception of the genus Palu-dicella, inhabitants of the sea.
The Gymnolscma are divided, accord-ing to the system of Busk, into three sub-orders characterized by the shape of their zocecia, and the nature of the ; mouth-like margin wdiich it presents when the exsertile portion of the poly-pide is withdrawn within it. The Cyclostoma have long tubular zooecia, often of large size and often calcified, placed side by side in cylindrical bun-dles, or in other definite grouping ; the mouth of the zocecium is circular and devoid of processes. There is little or no differentiation of the polypides con-stituting a colony. Most of this group are fossil, and the living genera belong mostly to southern seas. The genera Crisia (fig. 13, A), Diastopora, Tubuli-pora, and Hornera are typical. The Ctenostoma have usually a soft zoce-cium ; its orifice is closed by the folds of the retracted polypide or by a circlet of bristles which surround it. Aleyonidium gelatinosum is the com-monest representative of this group on the British coasts. Bowerbankia (fig. 1, A) and Paludicella (fig. 1, E) also belong here. The Chilostoma form the largest and most varied sub-order of Gymnoleema. The zocecia are „
marchis (Bugula) avicularia, Lrax. (Chilostoma), of which the anterior contains a living polypide, whilst the posterior is empty. To each is attached one of the characteristical-ly modified polypides known as an " avieularium" o ; the hinder of these has grasped and holds in its beak a small worm, a, anus; i, intestine ; v, stomach; r, body-cavity (ccelom); t, tentacular crown surrouuding the mouth; to, testis cells developed on the surface of the terminal mesen-tery or '' funiculus "; o,o, avi-cularia.
horny or calcified; their orifices can be ^J^Z^JSTlSid^ closed by a projecting lip in the form of an operculum. The operculum is a separable plate developed on the cuticle of the retractile part of the polvpide, and has muscles attached to it (fig. 13, B, C, D). The surface of the zooecia is frequently sculptured, and its orifice provided with processes and spines (fig. 1, C, F). Very usually some of the polypides of a colony are modified asavicularia, vibracularia, radi-cal fibres, and ocecia. The avieularium is a polypide reduced to a simple muscu-lar apparatus working upon the modified operculum and zocecium so as to cause these hard parts to act as a snapping apparatus comparable to a bird's head (fig. 12, o). They are frequently found regularly distributed among the normal cells of a colony, and probably have a cleansing function similar to that attributed to the Pedicellariae of the Echinoderms. '' Vibracularia " are even more simplified polypides, being little more than motile filaments, probably tactile in function. The opercula of zocecia, ocecia, and avicularia have recently been used by Busk in character-izing genera and species, in a systematic way. Stem-building and root-forming polypides are frequently found, being closed polypides which subserve anchoring or supporting functions for the benefit of the whole colony. The stem of Kinetoskias (fig. 14) is produced

in this way. The Chilostoma include a large series of genera arranged in the sections Cellularina, Flustrina, Escharina, and

FIG. 13.—A. Csencecium of Crisia eburnea, Lin., one of the Cyclostoma; g, g, tubular zocecia with circular terminal mouths; x, ocecium, being a zooecium modified to serve as a brood-chamber.
B. Diagram of a single polypide of one of the Chilostoma in a state of expansion,
in order to show the position and action of the operculum, a, operculum,
a plate of thickened cuticle hinged or jointed to 6, the main area of dense
cuticle of the antitentacular region known as the zocecium ; c, the soft-walled
portion of the polypide in expansion.
C. The same zooecium with the polypide invaginated (telescoped) and the
operculum a shut down over the month of the zooecium.
D. Operculum detached, and seen from its inner face, to show the occlusor
muscles d d.

Eupolyzoa in which the anal aperture lies close to the mouth within the tentacular area or lophophore. Lopho-phore sunk within a shallow basin formed by the inversion of the broad truncated extremity of the cup-shaped body. Tentacular crown not further introversible, the individual tentacles (as in the Pterobranchia and unlike the Ecto-procta) capable of being flexed and partially rolled up so as to overhang the mouth (see fig. 15, B and C). Body-cavity (ccelom) almost completely obliterated. The anti-tentacular region of the polypide's body is drawn out to form a stalk similar to the gymnocaulus of the Pterobran-chia. The extremity of this stalk is provided with a cement gland in the young condition which persists in the adult of some species (Loxosoma neapolitanum, fig. 16, shs). Cuticular investment (zooecium) of the polypides feebly developed. A pair of small nephridia are present.
The Entoprocta consist of the marine genera Pedi-cellina (fig. 15), Loxosoma (fig. 16), and probably the
the reader is referred to the works of Busk (9), Hincks (10), Smitt (11), and Heller (12). See also Ehlers (13) on Hypophorella.
insufficiently known freshwater American genus Urnatella of Leidy. To these must be added Busk's new genus Ascopodaria, as yet undescribed, based on a specimen dredged by the " Challenger," showing a number of Pedi-cellina-like polypides, carried as an umbel on a common stalk of very peculiar structure. Pedicellina is found at-tached to algae, shells, zoophytes, &c, and to the integu-ment of some Gephyraean worms (Sipunculus punctatus)-and Annelids (Aphrodite) ; Loxosoma occurs on various worms, &c. Whilst the buds of Pedicellina remain connected so as to constitute a colony, those produced by Loxosoma are continually detached, so that the polypide is solitary. Further, the cup-like body of Pedicellina is deciduous, and frequently falls from the stalk and is replaced by new growth. There is less distinction between body and stalk in Loxosoma, and the former does not become detached. Apparently a very important feature in the structure of the Entoprocta is the absence of a body-cavity. This is, however, more apparent than real. The Entoprocta are true Ccelomata, but the ccelom is partially obliterated by the growth of mesoblastic tissue. The nephridia presum-ably lie in a space which, small as it is, represents the ccelom. See Harmer (18) for details.

Genealogical Relationships of the Groups of Polyzoa.
It is necessary that we should try to form some opinion as to which of the various groups of Polyzoa are most like the ancestral form from which they have all sprung, and what are the probable lines of descent within the group. Any attempt of the kind is speculative, but it is absolutely needful since zoology has become a science—that is to say, an investigation of causes and not merely a record of unexplained observations—to enter upon such questions. Colonial organisms have necessarily descended from soli-tary ancestors, and it is probable that the ancestral form of Polyzoa was not only solitary, as are Phoronis and Loxosoma at the present day, but of relatively large size and more elaborately organized than the majority of living Polyzoa. Whilst the polypides have dwindled in size and

FIG. 16.—Diagram of Loxosoma Neapolitanum (after Kowalewsky). A single polypide devoid of buds, m, mouth; st, stomach ; shs, basal gland of the polypide-stalk.
lost some of their internal organs, the modern Polyzoa have developed pari passu with this degeneration an elaborate system of bud-production and colony-formation. The new individuality (the tertiary aggregate) attains a high degree of development (Cristatella, Kinetoskias) in proportion as the constituent units merged in this new individuality have suffered a degeneration. The prae-oral lobe (epistome, buccal disk) present in all Polyzoa except the most minute and most elaborately colonial forms— namely, the Gymnolaema—is to be regarded as an ancestral structure which has been lost by the Gymnolaema. The horse-shoe-shaped lophophore, such as we see it in Phoronis and in Lophopus, is probably the ancestral form, and has given rise to the two other extreme forms of lophophore,— namely, the " pterobranchiate," -associated with a great development of the epistome, and the " circular," associated with a complete suppression of the epistome. The ento-proctous lophophore is a special modification of the horse-shoe-shaped, as shown in the diagram fig. 15, C. The formation of zocecia, and so of an elaborate colonial skeleton, was not a primary feature of the Polyzoa. Even, after budding and colony-formation had been established zocecia were not at once produced, but possibly dwellings of another kind (Pterobranchia). We are thus led to look upon the Gymnolaema as the extreme modification of the Polyzoon type. Starting with an organism similar to Phoronis, we may suppose the following branchings in the pedigree to have occurred.
VERMIFORMIA
I
B. The complete hippocrepian-lophophore retains its form, but acquires a gradually increasing power of being telescoped into, the hinder part of the body. = The Pro-Eupolyzoon.
I
A. The complete hippocrepian lophophore becomes specialized in the form of ctenidia or gill-plumes ; the epistome enlarged.
= PTBROBKANOHIA. a. The anti-tentacular region of the body elongated as a stalk gives rise to one or two rapidly detached buds (Ce-phalodiscus). $. The stalk gives rise to buds which do not detach them-selves, but remain in con-tinuity so as to form a colony of a hundred or more individuals (Rhabdo-pleura).

B. The complete hippocrepian i lophophore remains in its origi-nal form, and also the pra>oral epistome, but the telescopic in-troversibility of the anterior region of the body is greatly de-veloped at the same time that the cuticle of the hinder part of the body is increased in thickness and toughness. Bud production, not from a stalk-like pedicle, but from all parts of the body, now becomes characteristic, the buds, which were at first deciduous, now remaining in permanent continuity so as to form colonies.. = The Pro-Ectoproeton.
A. The polypides acquire the property of carrying their young so as to avoid the disastrous influences of fluviatile currents, and also the property of produc-ing résistent statoblasts, and thus are enabled to become isolated and to persist in the peculiar conditions of fresh waters.
= The 1st order (of Ectoprocta) Phylactolaema.


B. The polypides forming relatively larger colonies, and themselves becoming relatively more minute, lose by atrophy the prfe-oral epistome ; and simul-taneously the arms of the hippo-crepian lophophore dwindle, and' a simple circum-oral circlet of tentacles is the result. The cuticle of the hinder part of the polypide becomes more and more specialized as the cell or zoce-cium, and in different polypides in various parts of the colony acquires special forms—as egg-cases, snappers (avicularia), ten-tacles, stalk and root segments. = The 2d order (of Ectoprocta). Gymnolaema.
Distinctive Characters of the Polyzoa.
From all that has preceded it appears that the really distinctive characters common to all the Polyzoa may be summed up as follows :—
Ccelomata with closely approximated mouth and anus,, the bulk of the body forming a more or less elongate growth at right angles to the original (ancestral) oro-anal axis, and starting from the original ventral (i.e., oral) sur-face. A variously modified group of ciliated tentacles is disposed around the mouth, being essentially the develop-ment by digitiform upgrowth of a post-oral ciliated band..

As negative characters it is important to note the absence of all trace of metameric segmentation, of setae, and of paired lateral (parapodia of Appendiculata) or median \ventral (podium of Mollusca) outgrowths of the body-wall.
Larval Forms of Polyzoa.
In the consideration of the probable pedigree and affinities of the "Polyzoa, we are not at present able to make use of the facts of .development from the egg, on account of the extreme difficulty which the study of the young stages of these organisms presents. In the case of Phoronis we have the only readily intelligible his-tory. The larva, to start with, is of that form known as an archi-troch (see Lankester, '' Notes on Embryology and Classification," Quart. Joum. Micr. Sci., 1876), having a prae-oral ciliated area (velum or cephalotroch) continuous with a post-oral ciliated band (the branehiotroch), which latter becomes developed into the ten-tacular crown of the adult.
The actinotrocha (Phoronis) larva is readily comparable with the trochosphere larvae of Echinoderms, Chaetopods, Gephyraeans, and 'Molluscs. Its special character consists in the strong develop-•ment of the post-oral ciliated band, whereas the prae-oral ciliated band is in most other classes (the Sipunculoids excepted) the .predominant one. The Phoronis larva exhibits first of all an oro-.anal long axis, and this is suddenly abandoned for a new long axis by the growth of the ventral surface of the larva at right angles to the primary axis (hence the term Podaxonia).
In the other Polyzoa we do not at present know of any larva which retains even in its earliest phases the original oro-anal long axis. They all appear to start at once with the peculiar and secondary long axis of the adult Phoronis, so that Balfour has diagrammaticaily represented the Polyzoon larva by the sketch given in fig. 19. This diagram applies, however, more especially to the Entoprocta, since the anus is represented as included in the area of the post-oral ciliated ring. The development of Pedicellina has :been very carefully followed by Hatschek, and may be said to be

Fig. 17. Fig. 18.
(FIG. 17,—Larva of Pedicellina (from Balfour, after Hatschek). «, vestibule (the cup-like depression of the tentaculiferous end of the body); m, mouth ; I, digestive gland; an.i, anal invagination ; fg, the ciliated disk (corresponding to the cement gland of Loxosoma (fig. 16, shs); x, so-called "dorsal organ," supposed by Balfour to be a bud, by Harmer (18) regarded as the cephalic ganglion.
FIG. 18.—Later stage of the same larva as fig. 17. Letters as before, with the addition of nph, duct of the right nephridium ; a, anus ; hg, hind-gut.
the only instance among the Eupolyzoa in which the growth of the different organs and the consequent relation of the form of the larva to the form of the adult is understood (see figs. 17 and 18).
In the other Polyzoa, in spite of the painstaking and minute studies of Barrois (14), the fact is that we do not know what face of the larva corresponds to the tentacular area, what to the stalk or anti-tentacular extremity, what to the anterior and what to the posterior surface. The conversion of the larva into the first polypide has not st. been observed in the ease of these free-swim-ming forms, and it is even probable that no such conversion ever takes place, but that the first polypide forms as a bud upon the body-wall of the larva.
Two of the most remarkable forms of free-swimming larvas of Gymnolaema are repre-sented in figs. 20 and 21. In both, in addition to the chief post-oral ciliated band, a smaller ciliated ring is observed, which is identified by Balfour with that which is found at the anti-tentacular extremity (base of the stalk) in the Pedicellina larva.

It does not seem justifiable, in the face of the existing uncertain-
ties as to identification of parts, and in view of the high probability
that the Gymno-
laema are extremely SC
modified and degen-erate forms (a con-sideration which applies in some re-spects even mor strongly to the En-toprocta), to assume
that the larval form s
schematized in fig. FlG 20.—Larva of Alcyonidium mytili (from Balfour 19 represents an an- after Barrois). m?, problematic structure; si, oral cestral condition of invagination (?) = Harmer's cephalic ganglion ; s, cili-the Polyzoa Pro- ated dislt (corresponding to fg in figs. 17, 18, and 21). fessor Balfour (15) was, however, led to entertain such a view ; and, assuming that the chief .ciliated band (drawn as a broad black line) corresponds to the single prse-oral ciliated band of the trochosphere larva of Echiurus, Polygordius, | Chaetopods, and Mollus-ca, he pointed out that in both cases the ciliated girdle divides the larva into a hemisphere in which mouth and anus lie and a hemisphere which is the complement of this ; in most classes the first hemisphere elongates and forms the bulk of the body,. whilst the second hemisphere forms the prostomium or prae-oral lobe. But, ac-cording to Balfour's theory, in Polyzoa it is the second hemisphere which enlarges and becomes the stalk-like body of the adult, whilst the first hemisphere remains small and insignificant. Thus the Polyzoa would fix themselves in later growth by wdiat corresponds to the head or prostomium of other animals, as do the Bar-nacles and the Ascidians. In-genious as this speculation is, we must remember that it takes no account of the facts known as to Phoronis, nor of the Ptero-branchia, and that it is con-fessedly based upon the assump-tion that the larvae of extremely degenerate and peculiar members of the group are not adaptive and modified, but retain primary and archaic characters. Further, it is to be distinctly borne in mind that the interpretation of parts upon which this speculation rests is, except in the case of Pedicellina, altogether hypo-thetical.
Relations of the Polyzoa to the
Prachiopoda. The Polyzoa were first asso-ciated with the Brachiopoda by H. Milne-Edwards. The inves-tigation of the development of Terebratulina by Morse (16) led to a further perception of the points of agreement in struc-ture between these two groups. Lastly, Caldwell (6) has shown that the mesenteries of Phoronis have precisely similar relations to the lophophore, the nephridia, and the termination of the intes-tine as have the gastro-parietal and ilio-parietal bands or mes-enteries of the Terebratulidse. The young Terebratulina (fig. 22) may be readily compared with Loxosoma (fig. 16),—the peduncle with its cement glands in the former being identical with the stalk and basal gland of the latter. The form of the alimentary canal

and the disposition of the tentacular arms (fig. 23) is the same in Brachiopoda and Polyzoa. The nephridia (oviducts) of Terebratula have a position and relations similar to those of the nephridia (geni-tal ducts) of Phoronis. The chief difference between Polyzoa and Brachiopoda consists in the special development of the margin of the cupped end of the body, into which the lophophore is sunk, as in Pedi-cellina (see fig. 15, B, c). This up-standing margin is enormously



Fig. 23. Fig. 24.
Flo. 23.—-Lophophore and epistome of young Terebratulina, showing the horse-shoe shape; the aims are turned in the direction the reverse of that taken by the hiphophore-arms in Polyzoa (sec fig. 4). In later growth they will become spirally coiled. (After Morse.)
FIG. 24,—Larva of the Brachiopod Argiope (from Gegcnbaur, after Kowalewsky). m, setigerous lobe; b, seta;; d, enteron.
increased in the Brachiopoda, so as to form a voluminous hood or collar, which surrounds the large tentacular arms, and forms a pro-tective chamber for them. It is notched right and left so as to be divided into two lobes, and on its surface is developed a horny or a calcareous shell in two corresponding moieties. Until recently it was held (see Lankester, 17) that both Brachiopoda and Polyzoa were modifications of the Molluscan type, and the Brachiopods' collar was identified with the pallial fold of Mollusca. The resem-blance of the two structures must now be considered as purely homoplastic, and not as having any real morphological (homo-genetic) significance.
The larvae of the Brachiopoda (figs. 24, 25) are as exceptional and difficult of interpretation as those of Polyzoa, but no attempt has been yet made to show that the one can be reduced to a common form with the other. The three segments presented by some Brachiopod larvaj (fig. 25) have been compared to the segments of Chretopod worms by some writers ; and these, together with the presence of setse, have been regarded as indicative of affinity between the Brachiopoda and Chastopoda (Morse). But it is sufficient, in order to dispose of this suggestion, to point out that the segments of the Chsetopoda follow one another along the primary oro-anal axis, whilst those of Brachiopoda are developed along an axis at right angles to this (Caldwell). The Brachiopoda must be classified together with the Polyzoa

FIG. 25.—Surface views of ten stages in the development of Terebratulina, showing . the free-swimming larva and its mode of fixation (after Morse), c, lophophoral segment; th, thoracic segment; p, peduncular segment; d$, deciduous seta;.
and Sipunculoidea in a phylum (Podaxonia) characterized by the development of this secondary axis.

Bibliography.—(1) J. Vaughan Thompson, Zoological Researches, Memoir v.,-
"On Polyzoa, a new animal, an inhabitant of some Zoophytes," &c, 1830; (2)
Ehrenberg, Abhandt. d. k. Akad. d. Naturwiss. zu Berlin, 1834; (3) Henri Milne-
Edwards, Recherches anatoniiques, physiologiques, et zoologiques sur les Polypiers
de France, Svo, 1841-44; (4) Allman, The British Freshwater Polyzoa, Ray
Society, 18-16; (5) Jolliet, " Bryozoairea des cötes de France," Arch. d. Zoo'l.
experim., vol. vi., 1877; (6) Caldwell, Proceedings of the Royal Society, 1883;
(7) Lankester, "Rhabdopleura," Quart. Jour. Micr. Sei., 1884 ; (8) Nitsche,
Zeilschr. für wiss. Zoologie, 1869, and supplement volume, 1876 ; (9) Busk,
Catalogue of the Marine Polyzoa in the British Museum (1852), and Voyage
of the " Challenger," "Report on the Polyzoa," vol. x.; (10) Hincks, British
Marine Polyzoa, London, 18S0; (11) Smitt, Kritisk Förteckning öfver Skandi-
naviens Hafs Bryozoa, 1S64-68; (12) Heller, Die Bryozoen d. Adriatischen
Meeres, 1867; (13) Eiders, " Hypophorellaexpansa," AbJiandl. d. kö'nig. Oeselisch.
Göttingen, xxi., 1876; (14) Barrois, Annales des Sciences Naturelles, vol. ix..
1880; (15) Balfour, Comparative Embiyology, London, 1880, vol. i,, p. 242 ; (16)
Morse, "On the Systematic Position of the Brachiopoda," Boston Soc, Nat. Hist.,
vol. xv., 1873 ; (17) Lankester, "Remarks on the Affinities of Rhabdopleura,"
Quart. Jour. Micr. Sei., vol. xiv., new series, 1874; (18) Harmer, "On Loxo-
soma," Quart. Jour. Micr. Sei., April 1885. (E. R. L.)