In specimen A the pores went on for about 20 segments in regular alternation; in specimen B the regularity is wanting, and the following table, taken also from specimen B, shows the asymmetry of the pores on either side of the same segment. It gives the position of the pores in 14 segments near the posterior end of the worm.

Besides this alternation from segment to segment, there is an asymmetry on the right and left sides of the same segment—i.e., a segment may have a dorsal pore on the right side and a ventral pore on the left, and vice versa. This, however, is not invariable, though it is frequent. In some specimens examined the dorsal pore of the right-hand side was found more often opposite a dorsal on the left hand than opposite a ventral. In the majority of the worms examined, however, the asymmetry occurs.

Fig. 1. Anterior end of M. rosae (preserved in alcohol after potassic bichromate and acetic acid), showing the position of the anterior nephridiopores, the first of which (n.p.) lies in line with the prostomial furrow.

In the first six nephridial-bearing segments the nephridiopores are in one and the same line—i.e., they occupy the dorsal position on both right and left sides. In some cases the 8th of those segments was reached before any alternation commenced. A similar fact was noted by Benham (12) in M. uliginosus. But, excluding these anterior segments, there are more pores placed dorsally throughout the length of the worm (which contains about 170 segments) than ventrally. It may here be noted that in the table given by Borelli (13) of the position of the nephridiopores in various species of the European genera Allolobophora and Lumbricus there are, on the whole, fewer pores in the dorsal than in the ventral positions.

The position of the first pair of nephridiopores deserves particular notice, as it has been recorded hitherto for only a few other species. Each pore is at the extreme anterior margin of the 2nd segment, just behind the intersegmental groove, and slightly external to the furrow which limits the backward prolongation of the prostomium which crosses the 1st segment.

It has occurred to me that perhaps this groove alongside the base of the prostomium may have some function in connection with the nephridium. It is evident that the secretion discharged from the nephridiopore would readily flow along the groove which would lead it on to the food while this was being taken into the mouth. How far is this association of a “tanylobic” prostomium with a pepto-nephridium at the hinder end of the furrow widespread?

The only figure that I have been able to find showing this position is in Benham's account of Neodrilus monocystis, where it occupies the same position as I have here described. But he also notes that in Plagiochaeta punctata the “vestigial” nephridium of the 3rd segment opens by a pore m this position.

A few other instances may be noted here, for it is stated as a general rule in earthworms that the first nephridiopore lies in the 3rd segment.

Fig. 2.—A dorsal and a ventral nephridium, as seen by opening the worm in the middorsal line. Drawn from Maoridrilus uliginosus, -in which the caecum of the ventral nephridial bladder is longer than that of M. rosae, though the nephridia are, in general, very similar.

Beddard (P.Z.S., 1887) states that in Thamnodrilus gilielmi (p. 154) and in Plutellus (Cryptodrilus) fletcheri (p. 544) the first pore lies in the 2nd segment, as it does too in Argilophilus (Eisen) and one or two others. But it seems that little attention has been paid to details of this kind.

for dorsal shifting of the anterior pores has also been noted by Eisen for Deltania, where the nephridiopores normally lie in front of the 4th. He notes similar facts for Phoenicodrilus.

Bourne states that in Moniligaster grandis the first seven, pores (those of segments iii to ix) lie dorsad of the normal position; which is in line with the outer (i.e., dorsalmost) chaeta of its segment. Indeed, it appears that this shifting occurs in various genera.

The Nephridia.—On the worm being opened, the nephridia are seen to be arranged in two series, corresponding to the external pores. Those

nephridia which open on the dorsal surface are placed dorso-laterally; those which open on the ventral, ventro-laterally. The funnels, however, are all placed ventrally, in line. This condition has been described by Benham for M. uliiginosusm, and may be contrasted with that of the genera described by Hubrecht and Borelli, where there is only one position of the nephridia corresponding to three positions of the external pore. This position is ventral, and the terminal duct, in order to open dorsally, runs up between the muscular layers from the ventral to the dorsal surface.

In Maoridrilus rosae the terminal bladder opens directly to the exterior (at the points ne.p.d, ne.p.v, fig. 2). In general arrangement and position in the body-cavity they are very similar to the nephridia of Neodrilus monccystis as figured in a transverse section by Benham (11), pl. 15, fig. 4, and in M. uliginosus (12). It is probable that in both worms their minute structure is similar to that of M. rosae.

The nephridium of Maoridrilus is formed essentially on the same plan as that of Lumbricus, although the former is somewhat simpler in structure, and differs in external appearance from the latter. Lumbricus has only one series of nephridia, which are therefore all alike; but in Maoridrilus, as in several New Zealand earthworms, the two series exhibit slight differences of form in accordance with their position. The chief difference between the dorsal and ventral series (see fig. 2) concerns the terminal bladder (bl). In the ventral nephridia this has a large caecum directed dorsally, while in the dorsal series the caecum is smaller, the bladder being continued only a little way beyond the external pore.

The nephridium is made up of a small preseptal and a large post-septal region. The preseptal portion (fig. 2) consists, as usual, of a funnel, a nephrostome, and a funnel-duct, lying more or less parallel to the septum, with its opening directed upwards. The preseptal part of the nephrostomial duct is a delicate tubule which curves round from the funnel, passes immediately through the septum, and joins the post-septal portion of the nephridium.

The Funnel.—It is by no means easy to dissect out a nephridial funnel in this species, owing to its position close against the septum, and its proximity to the nerve-cord, so that its structure was studied chiefly in sections. But I was able to obtain a good view of one sufficiently isolated to exhibit its general form; and it will be noted (fig. 3) that the marked difference in the height of the dorsal and ventral lips which is familiar to us from the usual figures of the Lumbricus nephrostome is here absent.^*

The funnel is not distinctly marked off from the funnel-duct, which is at first nearly as wide as the funnel itself, and slowly tapers towards the septum (fig. 4), which it perforates. The large lateral flaps formed by the prominent rounded lip in Lumbricus, &c., are here absent, and at a first glance the whole preseptal portion looks like a very long funnel; but further Study shows that this really consists of the true nephrostome and its preseptal duct.

The actual opening is comparatively wide, and surrounded by a rather thick ciliated lip, formed by marginal cells. The thickness of this lip is almost uniform all the way round, except that it is interrupted on the ventral surface; though in the figure this margin is shown as it actually appeared in the preparation, undulating, so that the height of the lip appears to be

irregular. The examination of sections, however, enables one to rectify one's ideas on this matter.

The marginal cells are much shorter and less specially arranged than in other genera, but are, as usual, provided with cilia which project into the coelom and are shorter than those borne by the central cell. The marginals pass round the sides to the front of the funnel, decreasing only very slowly and very slightly in height, and here join the cells of the canal-wall.

The dorsal lip is formed by the central cell (figs. 4, 5, n.cc), capped by the short marginals; the ventral lip merely by the cells of the canal-wall which here, of course, cease abruptly.

Below the margin there is a thick layer of connective tissue, as is usually the case, but in- Maoridrilus it comes right up to the marginal cells, and, tapering downwards, gives the nephrostome a resemblance to an ordinary glass funnel, or, with its funnel-duct, to a convolvulus.

Fig. 3.—Outline of a funnel, from the side, as seen in a living nephridium. The long cilia were moving rapidly, and some of the cilia of the marginal cells, together with those of the central cell, were sweeping strongly down the funnel.^*

Fig. 4.—Longitudinal section through a funnel, showing the nephrostomial duct passing through the septum. The slowly tapering character of the funnel-duct is well seen; it joins the nephrostomial duct at the point z, and the latter enters the nephridial folds at the point y.

Fig. 5—Transverse section of a funnel from front to back. The cilia of the central cell are directed against the front wall of the funnel.

This connective tissue is variably developed: in some funnels I find that it is continued down the sides of the preseptal canal much farther than in other cases; in the section figured (fig. 4) it was feebly developed.

The dorsal lip contains very little connective tissue, but the lower lip is supported by a considerable amount. The nuclei (as has recently been described by Rosen (23) for Lumbricus) are here crowded, and in consequence this part of the funnel appears denser than the upper lip, which is composed only of the central cell (fig. 4).

In this figure the central cell, described in Lumbricus for the first time by Benham (10), and subsequently recorded in the funnels of other earth-worms

of various families, is indicated by its large nucleus, surrounded by granular protoplasm. But what makes this funnel of Maoridrilus rosae remarkable is the great tuft of cilia borne by this central cell, and directed down the nephrostomial canal; for in other earthworms this central canal is not ciliated^*

These cilia represent probably the original cilia of the flame-cell which terminated the embryonic nephridium; they have become somewhat reduced in size, and increased in number with the increasing size of the cell, but have still retained their original direction down the funnel. If this be so, the cilia represent an interesting case of the survival of an old structure in the specialized organ which has been developed from the older and simpler form. The presence of these cilia on the central cell is most unusual.

On reference to Vejdovsky's figures of the development of the nephridia of Rhynchelmis (25) we find that the embryonic pronephridium terminates in a large cell containing a vacuole, in which the flagellum forming the “flame” projects downwards towards the nephridial lobe, composed of a string of cells without lumen. As the embryo grows into the adult worm the vacuole acquires an opening to the coelom and becomes the cavity of the funnel while the terminal cell forms the central cell of the funnel of the adult. In the case of Maondrilus the develpment is probably similar, and the cilia have persisted on the central cell, which forms the chief portion of the upper lip, as does the terminal cell in Rhynchelmis.

Lumbricus, on the other hand, has a terminal cell in its embryonic pronephridium devoid of vacuole or flame; here, then, is perhaps the reason for the absence of cilia on its central cell.

The post-septal portion of the nephridiun (figs. 6 and 8) consists of two folds, anterior (a.f) and posterior (p.f) respectively, a large terminal bladder (bl), and a smaller fold, or “spur” (sp), which comes off from the ventral end of the posterior fold.

The anterior and posterior folds are connected by a single tubule, the “bridge” (b), which is ciliated throughout its length (fig. 8); and the anterior fold is connected with the bladder by a single wide duct, the “outlet-duct” (o.d).

Fig. 6. Outline of dorsal nephridium mounted in glycerine. The covering of coelomic epithelium (c.ep) is shown in part; the bladder is turned over, so that the pore appears to be facing the posterior instead of the anterior end of the segment.

The anterior and posterior folds consist of tubules embedded in connective tissue; they lie parallel to each other, and transversely across the body. I have used the nomenclature given by Eisen (17), but these two folds together correspond to the second or middle loop of Lumbricus (10). The single

tubule which connects them corresponds to the short “ciliated middle tube” of Lumbricus, or the “bridge canal” of Eisen. The third fold, or “spur” (sp), represents the third loop of Lumbricus, in which genus it is, however, much longer. In Maoridrilus it is short and curved outwards, and lies approximately at right angles to the posterior fold (sp). The nephrostomial duct, after passing backwards through the septum, traverses the body for a short distance, and joins the ventral end of the anterior fold at the point where the outlet-duct leaves it. This latter duct lies parallel to the folds and anterior to them; it crosses the coelom independently, and enters the bladder.

The shape of the bladder, or the muscular duct, differs in accordance with the position of the nephridium. In the ventral series it s long and straight, somewhat club-shaped, and comparatively narrow (fig. 2, bl) At its inner end it bends down into the intermuscular gap of the body wall through which the chaetae (ch) project, and opens at this point to the exterior. The pore is terminal, and the outlet-duct enters it at about the middle of its length, or slightly nearer the dorsal end. The extension of the bladder beyond the point of entrance of the outlet-duct I have called the caecum (fig. 2, caec).

In the dorsal series, however, the bladder appears to be a direct continuation of the outlet-duct, very much enlarged (figs. 2, 6). It is wider than the ventral bladder, but not so long. It is continued beyond the pore as a short dilated caecum (caec).

The nephridia lie practically freely in the body-cavity, being joined by blood-vessels and occasional delicate strands of connective tissue to the body-wall. There is no trace here of the large coelomic sacs or peritoneal masses on which the nephridia of some earthworms are supported.

The description given above refers to the nephridia from about the middle of the length of the worm. There is no alteration of form, beyond details of the coiling of the folds, towards the posterior end, though the muscular bladder appears to become larger in comparison with the folds; nor is there any decrease in size of the nephridia, which occupy, therefore, a greater part of the cavity of the segments, which are smaller at this end of the worm. The septa in this region are greatly thickened, and obscure a generl view of the nephridia.

In Maoridrilus there are no proctonephridia. Their absence was established by examination of a series of horizontal sections through the posterior end of the worm. Fig. 11, taken from such a series, shows the opening to the exterior of the third nephridium from the posterior end. There is no trace of a longitudinal duct connecting the terminal portions of the posterior nephridia, and opening by one common opening into the alimentary canal, such as has been recorded by Beddard (6), who discovered the anal nephridia of Octochaetus multiporus, and found that they opened not only to the alimentary canal, but also to the exterior, and in each casse by many pores, as is characteristic in that genus.

I could not find in the wall of the rectum of Maoridrilus any indication of a network of nephridial tubules opening by intercellular ducts into its cavity.

The nephridia of the clitellar region are similar to those of the postclitellar. In the preclitellar region, and particularly in the first six segments, the nephndia are much larger, and present a distinctly different appearance (figs. 7 and 8). In segments 3–6 the septa, being inclined backwards,

lies on the anterior or outer side of the anterior fold (a.f), and is straight: there is no branching of the narrow duct in Maoridrilus as there is in Microchaeta (Benham), in Moniligaster (Bourne), and in Argilophilus figured by Eisen, where the duct breaks up into a spongy mass of tubules encircling the other canals.

In the dorsal lobe (d.l), where the anterior and posterior folds become continuous, the outer narrow canal commences to loop round and wind in and out of the inner narrow canal (a.i). This looping often continues for a considerable distance down the posterior fold (p.f), though in the majority of specimens examined it was restricted to the dorsal lobe. The outer narrow canal continues down the outer (posterior) side of the posterior fold, and enters the “spur,” in which it crosses over (at point X, fig. 8) from the posterior to the anterior side.

At the apex of the spur the duct turns back along the anterior or inner surface, and without change of diameter returns along the posterior fold as the “inner” narrow canal (a.i) lying parallel to the “outer” canal. In the dorsal lobe this part of the canal is ciliated at certain points. After the “windings” in the dorsal lobe it is continued into the anterior fold, where it passes across the wide tube (c), and and comes to lie on the inner side of the fold, separated by the wide tube from the outer canal.

In some of the nephridia examined the inner narrow canal was, in this fold, of slightly greater diameter than the outer canal; in the majority, however, both ducts were approximately of the same dimensions.

Fig. 8. An anterior dorsal nephridium from one of the segments 3 to 7, greatly enlarged, to show the course of the nephridial canal through the various folds.

At the point where the nephrostomial duct enters the nephridium—i.e., at the lower end of the anterior fold—the inner canal turns sharply back, as shown in fig. 8, to form what I have termed the “bridge” canal (b). This does not lie in either of the nephridial folds,

funnel is ciliated throughout; not only the marginals, but even the central cell, as well as the epithelial cells surrounding the lumen in the proximal part, bear cilia. The nephrostomial duct from the funnel to its point of entrance into the anterior fold is finely ciliated. From sections, these cilia appear to be in two rows, but I could not ascertain the exact arrangement from a living specimen. The narrow canal has few ciliated regions, and these are confined to the dorsal lobe. In the inner narrow canal (a.i) cilia, were observed moving in three short lengths as shown in fig. 8, the direction of the excretory current being towards the anterior fold. No other cilia could be seen in the narrow duct. The bridge canal is ciliated from its commencement up to the ampulla (fig. 8), the cilia being directed towards the ampulla. The cilia in this canal form the chief means of distinguishing it and tracing its direction and limits. At the sharp bend (fig. 8, B), where the narrow tube is continued into the bridge canal, are two peculiar structures; they have the form of two combs of cilia (fig. 9, co), which lie almost at right angles to the course of the tubule. They are formed, as can be seen from the figure, of fused cilia, some of which are free, or have become frayed out to form the teeth of the comb. Their function is evidently to create a stronger current than could be given by separate cilia at the sharp turning; and on this account they may be compared, I think, with the membranella or undulating membrane of Paramoecium, which is often frayed at the edge. I thought at first that they could be compared to the “undulating structures composed of bundles of long cilia-like filaments” of Moniligaster grandis (15, p. 339), but these, though placed obliquely across the lumen of the excretory duct, are attached at both ends, and have no comb-like appearance; while the waves of undulation work in the direction of the current—i.e., from the nephrostome outwards.

Fig. 9.—Enlarged drawing of the commencement of the bridge canal (b), showing the combs of cilia (co) lying across the entrance to the canal. The arrow indicates the course taken by the excretory vesicles.

Fig. 10.—Enlarged view of the apex of the spur, showing the entrance of the bridge canal into the ampulla.

The combs of Maoridrilus work unceasingly with an undulation along the comb—i.e., across the current—while the cilia forming the teeth work strongly backwards and forwards, sweeping the contents of the tube on towards the nephridiopore. The inner narrow canal (a, fig. 9) contains excretory droplets or globules (ex.v), and as these arrived at this point they were seen to be gradually forced round the bend by the action of the combs. The cilia worked more energetically on coming into contact with the droplets, while the undulating motion also became brisker. It was some time before the larger globules were forced round, the smaller ones being more easily disposed of and brought round by the sweep of the cilia into the

The nephridium is supplied by two vessels—one connected with the dorsal, the other with the ventral, blood-trunk (fig. 12). The former trunk gives origin to a dorso-parietal vessel, which courses down the body-wall, and at a little distance below the lateral line divides into three branches—(a) to body-wall, (b) to septum, and (c) to the nephridium. The ventral system is built up on the same plan. It is inadvisable to use the names “afferent” and “efferent” in regard to these vessels, for although no doubt one of them takes blood to the organ and the other one takes it away, yet we are in ignorance as to which does which. A comparison of the literature on the matter shows that Harrington^* and Bourne agree to differ from the views expressed by Benham for Lumbricus. The statement of the latter as to the course taken by the blood in the dorso-parietal and ventro-parietal vessels is diametrically opposed to the views of Harrington, who made a most careful and elaborate study of the vascular system in Lumbricus.

On the other hand, the only earthworm which agrees with Maoridrilus in being without a subneural blood-vessel in which the vascular system has been pretty fully described is Microchaeta (Benham), and in the disposition of the vessels it agrees quite closely with Maoridrilus. So far, however, as the course of the vessels on the nephridium are concerned, we do not know the details. But in Maoridrilus the arrangement of the nephridial vessels is so similar to those in Lumbricus that it is unnecessary to reproduce the figures which I prepared for my thesis. It will suffice to give a brief account, and a figure representing the origin of the vessels in transverse section (fig. 12).

The nephridial vessel arises from the dorso-parietal vessel, and runs alongside the septum towards the ventral end of the nephridium, where it enters the “spur”; it divides into two or three branches, one of which traverses the “folds” to reach the “dorsal lobe.” Here it breaks up into a mass of small vessels (capillaries), which subdivide again and again so as to cover the winding ducts, and then unite to enter a corresponding vessel connected with the ventral blood-trunk. This vessel lies parallel with the former. Similarly, the other branches referred to, which go to the spur and to the outlet-duct, divide into capillaries, all of which enter corresponding vessels connected with the ventro-parietal vessel. This passes through the septum at the level of the nephrostome, and thence passes to the ventral blood-trunk.

Fig. 12. — Diagram of the blood-supply to the nephridia in all parts of the worm with the exception of the most anterior ones

There is a great fascination in tracing out the capillaries of this elaborate blood-supply because it is remarkably beautiful under the microscope, and because, on account of the parallel arrangement of afferent and efferent vessels, one may be traced with ease into the other. Perhaps I may be permitted to quote here a sentence from Bourne (15, p. 327) which seemed, as I read it, so peculiarly applicable to Maoridrilus. Speaking of the peripheral vascular networks of Moniligaster grandis, he says, “Perhaps the