V. Echinoidea. 21 



glandulaire, projecting into it. It is quite closed and is bounded by an 

 upward prolongation of the envelope of the gland which becomes soft and spongy 

 and terminates round the triangular hollow beneath the madreporite. The pro- 

 cessus glandulaire terminates in this space, being attached to the madreporite by 

 fine threads of connective tissue. It is covered with ciliated epithelium and tra- 

 versed by minute canals, most of which open at its apical extremity. On the side 

 of the anus the wall of the submadreporic space is crowded with thin walled ve- 

 sicles which sometimes contain the refractile amoeboid cells, and sometimes 

 ciliated cells like those of the coelom. The latter are probably produced in this 

 tissue; while the processus glandulaire is often crowded with amoeboid cells 

 which escape into the submadreporic space. The colourless cells with long pseu- 

 dopodia are especially characteristic of the gland and seem to pass out from it into 

 the coelom through its superficial canals. They are less abundant in ill nour- 

 ished animals, and the gland, together with the oral lacunar ring, is probably 

 the seat of their formation. The materials necessary for this plastidogenic func- 

 tion are supplied by the water entering through the madreporite and by the con- 

 tents of the visceral lacunar system. The water-vascular and the blood-lacunar 

 systems are entirely distinct and nowhere communicate with one another ; though 

 their oral rings are in close relation and can exchange contents by osmosis and 

 diapedesis ; neither does the blood-lacunar system communicate with the exterior 

 through the madreporite, as supposed by Perrier. Oxygen is conveyed to the 

 tissues by the coelomic fluid which receives it from the ambulacral system ; and 

 Stewart's organs immersed in this fluid serve for the respiration of the con- 

 tents of the closed jaw-chamber in the absence of external branchiae. The 

 author maintains, in opposition to Perrier, the truth of his previous statement 

 respecting the independent origin of the ovoid gland and the genital rudiments of 

 the young Strongylocentrotus [see Bericht for 1887 Ech. p 13]. - The two sub- 

 apical plates of Spatangus enclose a spongy tissue which receives the dorsal 

 extremity of the ovoid gland, corresponding to the processus glandulaire of D. 

 Koehler called it the madrejporic canal and believed it to communicate with the 

 exterior through the madreporite, which is not the case. The water-pores lead 

 solely into the tube lined by columnar epithelium, which was recognised by Teu- 

 scher and Hamann as the stone-canal. At the extremity of the madreporic apo- 

 physis it is joined by the duct of the ovoid gland (canal annexe) and towards 

 the adoral end of the gland it loses its columnar epithelium and becomes a rami- 

 fying canal which terminates blindly in the oesophageal region. Some little way 

 from its end is the blind extremity of Koehler's sinuous canal which arises from 

 the water- vascular ring ; but the two canals do not communicate, so that the 

 ambulacral system is not in direct relation with the exterior. At the side of the 

 sinuous canal is a second vessel, K.'s canal glandulaire, which arises from 

 the oral blood-lacunar ring, ramifies over the walls of the ovoid gland, and ter- 

 minates in a plexus within the membrane enclosing the genital apparatus. The 

 internal marginal lacuna of the gut communicates neither with the oral water- 

 vascular ring, nor with a perineural space [schizocoel ring, Hamann] , but with a 

 definite blood-lacunar ring, which with its radial extensions has been generally 

 overlooked. - - The author also describes the segmentation and gastrulation 

 of the ovum of D., which begins to assume the Pluteus-form about the 10 th day 

 and matures in 3 months. The two posterior arms are the first developed, to be 

 followed in succession by the anterior, antero-lateral, and antero-internal pairs ; 

 of these the posterior and antero-lateral pairs are long with smooth trelliswork 

 spicules; and lobes appear on the ciliated band, but there are no ciliated epau- 

 lettes. One abnormal larva had a supplemental pair of arms, while another of 



