'^Madreporic System'''' of Echinodermata. 323 



the left vasoperitoneal sac as constituting a left nepliridium, 

 the right having failed to receive a duct *. Of the numerous 

 niadreporic systems of some Holothurians I believe the deve- 

 lopment has not been made out. If the madreporic system 

 be really of nephric nature it would seem a 'priori unlikely 

 that the ciliary action should be reversed, despite the acquire- 

 ment of other functions by the nepliridium itself. These con- 

 siderations determined the following experiments, which 

 fully bear out the views which I have expressed. 



Exp. I. — A fresh strong Echinus splicer a was opened and 

 the madreporic canal dissected out and cut off at either end, 

 and then a longitudinal slit was made in one side ; on exam- 

 ining it in the perivisceral liquid I could easily see a strong- 

 inrush of particles through the slit and a corresponding out- 

 rush through the cut distal end of the tube, i. e. that turned 

 towards the madreporite. This was repeated and confirmed 

 on six specimens. The stony canal of Asterias was found 

 unsuitable for similar experiments. 



Exp. II. — From a live Echinus I cut out the madreporite 

 with a short stump of duct attached, and examined it in 

 sea-water to which a little charcoal powder was added. The 

 disk lived for over sixteen hours with active ciliary currents 

 and movements of the spines and pedicellarige. The flakes 

 of charcoal were carried about in the currents, but never 

 reached the surface of the disk except close around the bases 

 of the spines, where there are no perforations. They seemed 

 repelled from it, instead of settling down by gravitation ; and 

 this could only be due to an outward current through the 

 pores. 



Exp. III. — I excised the madreporite of a live Starfish 

 with a short stump of the sand-canal, and examined it in the 

 perivisceral liquid of Echinus to which a little carmine 

 was added, the coelomic surface of course downwards. 

 Examination was here possible by strong transmitted light (the 

 full aperture of the Abbe condenser) as well as by reflected 

 light. The surface is marked by radiating ridges (imper- 

 forate) , and if there were any indraught the carmine particles 

 should be attracted towards the centre of the disk and the 

 bottom of the grooves ; if there were merely no outrush they 



* It is interesting to recall the development of the nephridium in Peri- 

 patus : — " The ventral half of each somite remains distinct, and consists 

 of a small vesicle, leading from which is a small coUed tube (nephridium) 

 which acquires an external opening " (Haddon's summary after Sedgwick, 

 Embryology,' p. 59). This is comparable to the division of the vaso- 

 peritoneal sac into two in Echinodermata, the inferior (sc. ventral) portion 

 going to form the ambulacral system. 



23* 



