W R M 683 ratus of Eudnlus. On tlie right side the spermatheca has been cut away to show the contorted ovi duct ov. o, ovary ; sp, sperma theca; gl, gland opening into con joined duct of spermatheca and oviduct. 14. They are surrounded by layers of muscles, and the interior is divided by trabeculse into numerous compartments ; directly con tinuous with each ovary is a contorted tube which passes through the septum into the 13th segment, and then again passes through the septum, and opens on to the exterior in the 14th segment, in common with a large spermathecal pouch and a small glandular body. The oviduct opens by a wide mouth into the interior of the ovary, and is lined throughout with a ciliated epithelium ; it is covered by layers of muscular fibres, which are continuous with those of the ovary. The continuity of the ovary and its duct is unknown in any other Oligochieta, but is analo- gous to the condition of the ovaries FlG . 6 ._ Fema]c Rcproductlve Appa- and their ducts in the leeches. It is possible that the muscular wall of the ovary is to be regarded as an hypertrophy of the peritoneal cover ing of the ovary in Lumbricus, &c., and that this has involved the ovi duct; or the muscular coat of the ovary may be regarded as the receptaculum ovorum, which, as in Stylaria, is developed in the same segment as the ovary; this supposition is strengthened by the fact that the oviduct traverses the septum between the 13th and 14th segments. Spermathecal. Of these organs there are from one to eight pairs ; they consist of spherical or pear-shaped vesicles, often furnished with accessory diverticula. They receive the semen during copu lation and by their epithelium is fabricated the generally chitinous spermatophore in which the spermatozoa are enveloped, and in which they are conveyed to the neighbourhood of the clitellum of another individual. The genital ducts have been compared with nephridia by many writers : there is not only a considerable anatomical resemblance, but a developmental similarity ; the vasa cleferentia (and the ovi ducts ?), like the nephridia, consist of a funnel, of a more or less elongated tubule, and of a distal vesicular part. The intrctccllular duct of the nephridium and the intercellular duct of the vas deferens may be explained by the different functions which the organs perform. 1 The spermathecae, on the other hand, are com parable to the vesicle of the nephridium, which is formed in both cases by an ectodermic ingrowth. The fact that the funnel and the tubule in both cases are developed independently, but both from the mesoblast, is a striking point of similarity between the vasa deferentia and oviducts and the nephridia. The belief of Claparede that in the "Limicolse" the genital ducts were the homo- logues of nephridia, but not in the " Tcrricolae," was based upon the erroneous assumption that nephridia are absent in the genital segments of the former. This has been shown by Vejdovsky to be an error : nephridia are present at first in the genital segments, but degenerate and disappear when the genital ducts are formed. Moreover, enough has been already said to prove the extreme unlikelihood of a non-homology between the generative ducts in earthworms and those in the "limicolous" forms. Lankcster put forward the theory that there were primitively two pairs of nephridia in each segment, each series being connected with one of the pairs of seta ; the genital ducts were supposed to be the remains of one series which had aborted, except in the genital segments. This theory was at first espoused by Terrier, who found that the nephridia were sometimes connected with one series of setae and sometimes with the other, and sometimes (Plutcllus) alternated from segment to segment, opening in one segment by the ventral and in another by the dorsal setoe ; these facts appeared to show that one or other of the two series of nephridia was partially or entirely retained in different genera. In his later publications Terrier was not inclined to attach much weight to this hypothesis, inasmuch as it did not satisfactorily refer the genital ducts to one or other of the presumed double series of nephridia ; the apertures of the genital ducts and nephridia were found occasionally to co incide at the same pair of setse. The discovery that each segment of a worm may contain numerous nephridial pores disposes of any a priori difficulties as to the homology between nephridia and genital ducts, though the question is far from being settled. Classification and Affinities. It has already been stated that the Oligochscta form a group which cannot be subdivided into Limicolie and Tcrricolse as was proposed by Claparede. The genera of Oligo- chcxla have been arranged in families by Vejdovsky (14), to whom the reader is referred for a classification, which is satisfactory as regards the "limicolous" forms, and some of the families (e.g. , LuinbricidsR and Pcrichietidse.) of earthworms. Since that time, how- 1 The difference between the nephridia and genital ducts in this respect cannot be regarded as being of importance; the nephridia of 1 olyc.hxta differ from those of Oligochxta in having an intercellular lumen. Lnng, moreover, has shown (" Die Polycladen," Naples Monographs) that in some 1 hinarians the reproductive ducts, like the nephridia, may have an t nf/ accllular lumen. ever, a large number of new genera and species of earthworms have been described, which cannot at present be satisfactorily arranged. Terrier divided earthworms into three groups: (1) Prcclitdlians (e.g., Lumbricus), where the male pores nro situated in front of the clitellum ; (2) Intraditdlians (e.g., Eudrilus], where the male pores are within the clitellum ; and (3) Postditdlians (e.g., PericJiseta), where the male pores open behind the clitellum. Advancing knowledge has shown this classification to be untenable, for two principal reasons : first, because it separates some species of Acanthodrilus which are postclitellian from others which are intraclitellian, and it separates the intraclitellian Megascolcx from the postclitellian Pcrichieta, between which genera there are numer ous points of affinity ; and, secondly, because this classification is based on the assumption that earthworms can be considered as a group apart from other Oliyoch&ta, whereas it is now impossible to draw any line of division between any tAVO such groups. The simplest forms of Oligodiwta are the genera ^olosoma and Ctenodrilus. dSolosoma has a "head," consisting of one segment which contains the pharynx. The nervous system consists of cerebral ganglia, which are placed in the first segment, and retain connexion with the epiblast ; the apparent absence of a ventral nervous cord is in all probability to be looked upon as evidence of degeneration. The body possesses considerable traces of the primitively continuous ciliation ; upon the head are a pair of ciliated pits. In all these particulars, as well as in the characters of the vascular system, SEolosoma agrees with Ctenodrilus ; the latter form has, however, a ventral nerve cord, which, like the cerebral ganglia, is lodged in the cpilermis. In most of these points sEolosoma and Ctenodrilus resemble larval forms of the higher Oligoch&ta. The reproductive organs of Ctenodrilus are not known, but those of ^Eolosoma are constituted on the Oligochsetous plan. These two genera are also closely allied to the Archianndida, of which a short description is appended, as they are not treated elsewhere in this work. ^Eolosoma therefore retains certain Archiannelid characters, but is to be referred to the Oligoch&ta on account of (1) the limitation of the reproductive glands to two segments and the presence of special efferent testicular ducts ; (2) the paired arrangement of the seta} bundles, which agree in their structure with those of other Oligo- cli&ta ; and (3) the relative complication of the nephridia, as com pared with Archiannelida. The affinities of Ctenodrilus are not so plain ; but, in the absence of any knowledge respecting the generative organs, it is impossible to refer them definitely to ilic.Oligochfeta or to the Polychwta. Archiannelida. This subclass includes certain small marine Archian- worms which were formerly placed among the Polych&ta. Hatchek uelida. originally created the group for the reception of Polygordius and Protodrilus ; and recently Fottinger has placed Histriodrilus* (an animal formerly referred to the Discophora) in the same group. The name A rchiannelida implies that these forms stand at the base of the Annelida, and that they represent most nearly the common ancestral form from which both the Polychxta and OligocTiseta have been derived. Their structure, which bears out this supposition, suggested the name. Polygordius is found on the northern and southern coasts of Europe ; it is a small slender worm, varying in length (according to the species) from 30 m. to 1 decim. The segmentation is hardly marked externally. The head segment is divided as in the Chxtopoda generally into a prostomium and a peristomial ring; the prostomium gives rise to two tentacles. On either side of the prostomium is a ciliated pit, which is of special interest, as it occurs on the one hand among the NEMERTIXES (q.v.), and on the other among certain Polych-ssta (e.g., Ophdidse), and in the lowly organized Oligochcet sEolosoma. Cilia are found in certain species in the neighbourhood of the mouth and on the anal segment. The ciliation of the body is more marked in Protodrilus, where there is a continuous ventral groove lined by cilia ; there are rings of cilia also on each segment. The partial ciliation of Polygordius is to be derived from this by reduction. Certain Oligoch&ta retain traces of the primitive ciliation of the body (e.g., JEolosoma). There are no traces of sete in Polygordius or its allies. The segmentation of the body, although hardly marked externally, is very clearly marked internally ; the body-cavity is divided by suc cessive septa into a series of segments ; there is no such fusion of the anterior segments as is met with among the Polychicta and Oligochiela, and the segments are less differentiated among them selves than is usually the case in the higher Chxtopoda. The head segment contains the pharynx, which is limited to this segment ; the nephridia, which consist of simple tubes (probably but not certainly formed of a chain of "drain-pipe" cells, as in OligocJixta) are found in all the segments except the first and last. The nephridia do not lie freely within the coelom, but in the thickness of the parietal peritoneum ; the funnel only just opens into the cavity of the segment preceding that which bears the external pore ; that is to say, it is chiefly contained within the thickness of the septum. 2 The correct name of this worm is Ifistrinbdella. The altered termination is
to express the change in the view regarding its affinities.