MINUTE ANATOMY OF THE BRACHIATEcECHINODERMS. 175 
through the pores in the madreporite. Greeff recognised 
that the so-called genital vessels (¢.e. periheemal canals) 
surrounded the glands; but he supposed the glands to open 
intothem,and that the sexual products passed out by the inter- 
brachial pores, which would thus place the blood-vaseular 
system in direct communication with the external water. 
Ludwig’s later observations have shown that definite 
genital openings are present in all Starfishes, varying in 
number from ten upwards. They sometimes extend out on 
to the arms, so that the presence of individual openings for 
the numerous separate glands in the arms of Brisinga is 
no longer such a striking peculiarity as it was formerly 
Supposed. Each genital pore (fig. 8, g.p.) leads into a 
distinct efferent canal, which pierces the perihzemal canal 
and its contained vessel, and enters the cavity of the gland, 
to which it serves as a duct. In most cases these pores 
are on the dorsal surface of the arm, but in Asterina 
gibbosa they are ventral (25). 
It will be seen from what has been said above, that the 
pores of the madreporic plate were supposed by Hoffmann 
to lead into the tubular space around the stone-canal as well 
asinto the stone-canal itself. He likewise believed the marginal 
pores of the madreporite to lead directly into the celom. 
This connection of the water-vascular and the blood-vas- 
cular systems through the pores of the madreporite with 
the exterior and with one another was also believed in by 
Greeff and Teuscher. Their views rested principally upon 
the results of injection not checked by the section-method, 
which are necessarily liable to much error. Ludwig, how- 
ever, after making sections in three planes through the 
madreporic plate, satisfied himself of the truth of the older 
views of Sharpey, L. Agassiz, Miiller, and Tiedemann, 
viz. that the pores of the madreporite lead simply and 
soley into the stone-canal. The interior of the plate is 
traversed by pore canals, which correspond in position with 
the radiating furrows on its upper surface, and communicate 
with them by short vertical tubules. These are lined by a 
pavement epithelium, which becomes columnar and ciliated 
at their openings on the surface, and likewise in the actual 
stone-canal itself, into the top of which the radial pore 
canals open (fig. 7,p.). At the aboral edge of the attach- 
ment of the stone-canal to the madreporite is a lateral 
diverticulum of the former, into which some of the collect- 
ing tubes of the madreporite open (12, 21). It is occa- 
sionally double or even triple, but its wall.is never calcified 
like that of the stone-canal, and it is lined by pavement 
