Moreover the nephridia themselves in this instance have remarkable peculiarities. There may be 3-4 funnels to a nephridium; two adjacent nephridia may be connected by a tube; and the duct of each nephridium divides into two main branches, and each branch into 4-5 minute tubules which probably open externally by separate pores. The nephridia of the aquatic Oligochaeta situated in the genital segments atrophy before the sexual ducts develope; they persist in the terrestrial Oligochaeta side by side with the genital ducts.

1 Cosmovici's account of the nephridia in Polychaeta is followed in the text. He applies the term 'organ of Bojanus' to the vesicle, and that of 'segmental organ' to the ciliated tube. The vesicle is probably formed by invagination of the hypodermis, as Vejdowsky has found it to be in aquatic Oligochaeta, The presence of the vesicle in some somites, its absence in others, in one and the same animal, e. g. Ophelia, is a remarkable fact. A pair of vesicles, very large in size, opening on the head in Serpulidae, was supposed by Claparede to secrete the material for the tube. That author figures a stream of mucus passing out of the branchial funnel of Myxicola. See Supplement aux Annelides Chetopodes du Golfe de Naples in Mem. de la Soc. de physique et d'histoire Nat. de Geneve, xx. 1869, p. 512, PL xiv. Fig. 2. The position of the apertures (double, according to Cosmovici, single, according to Claparede) in this animal and in other Serpulidae is noteworthy. See note, p. 596, ante.

Cosmovici points out that the nephridial funnel opens into the median chamber of the coelome containing the digestive tract in all Errantia and in those Tubicola which are distinctly segmented. It opens in other Tubicola, e. g. Arenicola, into the lateral coelomic chambers (p. 597)- The nephridium itself is usually applied in the Errantia to the posterior face of a septum; sometimes to the anterior face as in Cirrhatulus. For a good figure of the nephridium and its papillate external pore in Polynoe clava, see A. G. Bourne, Tr. L. S. ii. PL xxxvi. Fig. 22. The pores are papillate also in Notomastus (Capitellidae).

The Oligochaeta are hermaphrodite, the Polychaeta of separate sexes with a few exceptions, such as certain Hesionidae and Nereis massiliensis among Errantia, and among Tubicola certain Serpulidae (Amphiglena, Laonome, Salmacina, Protula, Spirorbis Pagenstecheri, Sp. communis). The sexual organs in Oligochaeta are contained in the anterior somites, and there is either a single pair, or in terrestrial forms two pairs, of testes and a single pair of ovaries placed in different somites, the ovaries behind the testes, except in Plutellus. The organs are attached either to the posterior face of a septum or to the nerve-cord (Naidomorphd). The corresponding organs of Polychaeta are closely attached to a blood-vessel 1.

1 Cosmovici is followed in this account. He found the sexual organs developed in Tubicola on a vessel in relation with the nephridial funnel and tube (? Arenicold); on branches of a median ventral vessel (Terebella gigantea); or of an infero-lateral vessel (Spirorbis); on the intestinal walls (C/iaetopterus); or on the anterior aspect of a septum vis a vis to the nephridium (Hermella). As to the Errantia, he found the glands attached, in Nereis, to the supra-nervian vessel; in Marphysa (Eunicidae) to the branchial vessels proceeding from the supra-nervian, on the section intervening between that vessel and the nephridial funnels; in Hermione and Sthenelais (Aphroditidae) to a vessel on the posterior aspect of the septa; in Cirrhatulus to a vessel on the anterior aspect of the septa. The organs are present in a large number of segments in Errantia.

The number of organs present appears to correspond with the number of nephridia in many Tubicola. The testes and ovaries are usually very similar, and probably always originate from peritoneal or sub-peritoneal cells. They are scarcely visible save at the period of sexual activity, and consist in Oligochaeta of a single mass; in Polychaeta of a number of grapelike masses in which the developing ova and spermatozoa are contained within a membrane. Cell-limits are not always traceable in the young organ1. The genital products appear to be set free into the coelome, and to ripen there in Polychaeta and the Oligochaete Chaetogastridae. The testicular cells of other Oligochaeta ripen in special pouches developed from the septa - the so-called vesiculae seminales - paired in the terrestrial, unpaired in the aquatic, forms. The ovarian cells ripen in a similar way in some of the aquatic Oligochaeta, but in the terrestrial are detached from the ovary only when mature. Special efferent ducts - vasa deferentia and oviducts - are present in all Oligochaeta, and in Eudrilus the oviducts are continuous with the ovaries, a feature not observed in any other Chaetopod (Beddard). Accessory organs are also present; special copulatory setae which replace the ordinary setae; a clitellum at once a copulatory organ, and an organ for secreting the cocoon, produced at the sexual season by the hypertrophy and increase in number of the cutaneous glands in certain somites; (one only or several) receptacula seminis which closely resemble nephridial vesicles (p. 602, and note), and the vesiculae seminales above mentioned.

1The anterior somites of the hermaphrodite Microphthalmus (Hesionidae) are male, the posterior female (Bobretzky, Z. A. iii. 1880, p. 139). So Protula. In Spirorbis communis it is the reverse. Hesione sicula and Tyrrhena, on the contrary, have hermaphrodite organs, which take the form of longish cylindrical processes, with a blood-vessel in the centre. The surrounding peritoneal cells change into ova and spermatozoa commingled (Eisig, Mitth. Zool. Stat. Naples, ii. 1881, p. 298).