Much variety exists in the disposition of these parts (pp. 205-8). Though strong structural resemblances exist between the genital ducts and nephridia in Oligochaeta, it is by no means certain that the ducts are modified nephridia. As to Polychaeta, the escape of the genital products by the nephridia has been observed in Hermella, Arenicola, Terebella, and it probably occurs in other instances as well. It may take place also by rupture of the body-walls2. Accessory organs of generation are rare in this order, so far as is known, and are (perhaps) to be regarded as modifications of nephridia or parts of nephridia 3.

2 Cosmovici found ova in the nephridial vesicles of Arenicola, and witnessed their escape by the nephridiopores in Terebella and Hermella. A. G. Bourne states that the slightest irritation of the body-walls of Polynoe clava, when tense with accumulated genital products, causes rupture and discharge. But the products were never found in the nephridia. Kallenbach, however, states that they escape by the nephridia in Polynoe cirrata.

3 A pyriform vesicle is attached to the nephridia of the sixteenth and following somites in Alciope, and is filled with sperm in male and female alike. It represents possibly a nephridial vesicle. In the hermaphrodite Microphthalmus there is a pair of penial papillae between the second and third setigerous somites. The male ducts open on the apices of these papillae. The ducts are long, coiled, and furnished with ciliated funnels, and may be modified nephridia. The nephridia in the

The spermatozoa of Oligochaeta are contained in spermatophores, as is the case also in Spio Mecznikowianus, the sole instance among Polychaeta 1 The ova are invested by a vitelline membrane, and are generally coloured. They are impregnated externally to the body, or in transit (?), except probably in a few viviparous forms (Marphysa sanguinea, Syllis vivipara, a Cirrhatulus); and in Oligochaeta are laid in a cocoon secreted by the clitellum. They are carried under the elytra in Polynoe cirrata; attached to a dorsal or ventral cirrus in the Exogoneae (of Langerhans) among Syllidae, and in the female Autolytus cornutus ( = Sacconereis) they are carried and develope in a ventral sac. In Spirorbis Pagenstecheri ( = Sp. spirillum of Pagenstecher), and some other Sabellids 2 the ova develope in a cavity of the operculum; in Spirorbis spirillum (of Gould), and Manayunkia they are contained within the tube; in Terebella, Dasychone, etc. they are attached externally to the tube; and in Phyllodoce, Aricia, Ophelia, they are laid in a mass of jelly-like mucus.

Segmentation is with few exceptions, e. g. Serpula, unequal. The Oligochaeta have a direct development. A ventral ciliated furrow appears to be common, and in Lumbricus trapezoides its cilia extend round the mouth also. The young embryo of the same worm divides into two, and each half proceeds to develope normally. The Polychaeta have a more or less pronounced metamorphosis with the exception of the Syllidian Exogoneae. The larva is a Trochosphere with a prae-oral lobe of variable size, a small body somewhat pointed posteriorly. The body of the adult is formed by the growth and segmentation of this pointed region. The Trochosphere is ciliated, and the differing modes of arrangement of the cilia have given rise to various descriptive terms, to which a classificatory value has sometimes been given. One and the same larva may have at different times differing arrangements, and the larvae of allied species may be completely unlike3. The terms alluded to are as follows: Atrochae with the cilia scattered uniformly at first, though disappearing in places subsequently, but never so as to form bands; Mono- or Cephalo-trochae with a prae-oral band; Telotrochae with both prae-oral and peri-anal bands; female somites develope at the sexual period a saccule contiguous to the external aperture, which becomes filled with sperm.

The male and female Capitella capitata possess a pair of ciliated saccules between the seventh and eighth somites, opening both externally and internally. They are filled in both sexes by sperm ; ova appear to issue by them in the female, and there are special setae close to their external apertures in the male. The homology of these saccules is doubtful. Their external aperture is close to the intersegmental furrow, and they appear relatively late (Eisig, Mitth. Zool. Stat. Naples, i. 1879, p. 114). The structures termed 'elliptical or seminal pouches,' occurring in the sexual somites of Syllideae, are perhaps nephridial vesicles (cf. Viguier, A Z. Expt. (2), ii. 1884, p. 86), unless they are, as Robin appears to have found them to be in Pionosyllis, a collection of glands (Bull. Soc. Philomath. Paris (7), vii).

1 Claparede and Mecznikow, Z. W. Z. xix. 1869, p. 171.

2 Sp. laevis, Salmacina aedificatrix, Psygmobranchus caecus, and Pileolaria sp. ?

3 E. g. Terebella nebulosa is telotrochous, T. chonchilega nototrochous.

Polytrochae with bands appearing on the somites between the two telo-trochous bands and subdivisible into (I) Polytrochae proper with complete bands; (2) Nototrochae with only dorsal half-bands; (3) Gasterotrochae with only ventral half-bands; and (4) Amphitrochae with both dorsal and ventral half-bands which do not correspond; Mesotrochae with one or two median bands but devoid of the two telotrochous bands (the Chaetopteridae alone). A flagellum or a bunch of cilia are often situated at the apex of the prae-oral lobe, and there may be an anal patch of cilia. The prae-oral band is frequently double, and the mouth is then situated ventrally between the two. A stomodaeal and proctodaeal invagination probably always occur, at least during growth. The blastopore is said to persist as the primitive oral opening in some Oligochaeta. It closes in Serpula in what corresponds to the median ventral line, and the mouth and anus are formed in positions coinciding severally to its two extremities. The archenteron is more or less saccular, and the anus posterior. The larval digestive tract makes a curve concave ventrally.