178 Proceedings of the Royal Physical Society. 



branches which are in relation with the marginally placed pores, while com- 

 municating with the first set of branches and also with the central chamber 

 at points all round the convex side of the horse-shoe, are continued more 

 directly in a downward direction, and gradually flowing together form the 

 upper or aboral end of the axial sinus. The lining epithelium of these 

 cavities, except at the parts to be afterwards noted, consists of low cubical 

 epithelium, and the marginal collecting trunks nearly all contain a consider- 

 able number of corpuscles of leucocytic nature. 



The central chamber mentioned above is horse-shoe shaped, with the 

 concavity of the horse-shoe looking to the sinistral side, as well as slightly 

 outwards, ie., towards the margin of the disc. The horse-shoe itself lies 

 practically in the horizontal plane of the starfish, though, if anything, the 

 sinistral limb is a little higher (i.e., more aboral) than the dextral limb. 

 The roof of the chamber receives the openings of the collecting branches 

 from the central set of pore canals, while the sides of the chamber, except 

 within the bend of the horse-shoe, have free communication by nine or ten 

 apertures with those collecting branches which are in connection with the 

 marginal pore canals on the one hand, and with the axial sinus on the other. 

 These apertures are all of the same general character, but two, slightly larger 

 than the rest, are found a little to the dextral side of the middle of the 

 horse-shoe. The floor of the chamber is deficient. The blank left is the 

 commencement of the stone canal, and the horse-shoe cavity of the chamber 

 passes by gradual transition into the simply folded lumen which, in this 

 species, is characteristic of the stone canal. As it goes deeper, the canal 

 rotates clockwise through about half a right angle, so that its concave 

 or folded side now looks outwards towards the margin of the disc, as well 

 as slightly to the sinistral side. 



On those parts of its inner surface the curvature of which is concave to- 

 wards the lumen, the stone canal is lined by greatly elongated cells carrying- 

 long and powerful cilia. This lining passes upwards for a short distance on 

 the sides of the central chamber. It may also be continued into some of the 

 main branches of the chamber in the form of broader or narrower tracts, 

 which are usually to be found along the floor of the branches in question. 

 It merges, by gradual transition, into the cubical epithelium mentioned above 

 as lining the intramadreporic spaces generally. 



Careful examination of the whole series leads me to the conclusion (1) 

 that there are no channels connecting the external surface with the axial 

 sinus which do not communicate within the madreporite directly or indirectly 

 with the central chamber ; and (2) that one cannot speak of any single large 

 communication between the stone canal and the axial sinus, unless one means 



