324 Prof. M. M. Hartog 07i the True Nature of the 



sliould gravitate towards the bottom of tlie grooves. But 

 this did not occur ; on the contrary, the carmine settled round 

 the edge of the disk ; and in one or two places on the ridges, 

 where, from their convergence^ an eddy must necessarily 

 exist : not a particle entered the grooves. Three madreporites 

 were examined, all with the same results, even after three 

 hours. 



Eai). IV. — A Comatula {Antedon rosaceus) was examined 

 disk upwards in sea-water with charcoal powder, the arms 

 being removed to facilitate observation. During three hours 

 no charcoal particles reached the surface of the disk, except 

 along the imjperforate ambulacral grooves ; on the contrary, 

 as they floated down towards the disk they seemed arrested 

 above its surface by an invisible screen, which could have 

 been only due to an outward current through the coelomic 

 pores. 



Exp. V. — The same observations on eviscerated dishs of 

 Comatula gave the same results. 



The above experiments show clearly that the perforations 

 of the madreporite in Echinus and Asterias, and of the disk 

 in Comatula, are purely excretory, and serve to eliminate the 

 excess of water taken up by the body. 



It may now be urged, " How, then, can the Echinoderm 

 take up the liquid that fills its perivisceral and ambulacral 

 cavities?" One might as well ask how a Vertebrate takes 

 up the liquid in its coelom, blood-vessels, and bladder. The 

 answer is by osmosis, through the walls of the gut (respira- 

 tory siphon especially), the tube-feet, and the "gills." It 

 will easily be seen that when a starfish protrudes its tube-feet 

 rapidly the arm becomes limp from the evacuation of the 

 ampullse, and when it retracts them the arm regains its tur- 

 gidity , owing to the refilling of the ampullar. In Echinus the 

 problem appears complicated by the close rigid shell, which 

 would seem to prevent any ingress and egress of liquid 

 from its cavity ; but, in the first place, the soft peristome is 

 protractile and retractile, and quite large enough to balance 

 by its movements very considerable alterations in the capacity 

 of the ampullas ; and in the next place the intestine, through 

 which water is constantly streaming, is also dilatable. 



In the majority of the Holothuria we find that the madre- 

 porite has lost its connexion with the surface, and opens into 

 the ccelom. This admits of a ready explanation. The cloaca 

 is rhythmically contractile, and receives the excess of the 

 coelomic liquid by what are physiologically nephrostomes 

 attached to the respiratory trees, an arrangement physiolo- 

 gically the same as the nephridial apparatus of Rotifera. Theel 



