14 DR. J. F. GEMMTLL ON THE 



of these increased slightly in weight chiving the months of May, 

 June, and July. In control experiments, specimens of the 

 common Orossfish, Asterias rubens, kept under like conditions, lost 

 weight steadily and died, as a rule, in less than eight weeks. We 

 may note in this connection that the sucker-feet of Porania are 

 arranged only biserially, and that they are neither particularly 

 strong nor are they kept actively in use. At the Millport Marine 

 Station the Porania are never seen feeding on shell-fish etc., or 

 on their neighbours as other species readily do. Yet for several 

 seasons the Porania have remained healthy in the tanks as long 

 (nine months or thereby) as circumstances made it convenient 

 to keep them. 



In view of the data given under 1-10 above, it is, I think, 

 impossible to escape from the conclusion that ciliary feeding 

 plays a part in the nutritional economy of Porania. 



ASTROPECTEN IRREGULARIS (PI. I. fig. 2). 



It will be seen from the details given on pp. 6-7 that the 

 general arrangement of the cilia in Astropecten is suggestive of 

 ciliary nutrition. Strong oralward currents run in the ambu- 

 lacra! furrows, being fed by lateral streams coming from the 

 roof and sides of the disc and rays along the grooves between 

 the transverse rows of ambulacral spines. The interra.dii are 

 powerfully ciliated towards the mouth along avenues bordered 

 by short, thickly clustered^ spines. The whole aboral surface 

 underneath the great paxillary umbrella has its ciliation so 

 arranged that particles which manage to get through between 

 the tips of the spinelets into the' subpaxillary space can hardly 

 fail to be swept ultimately to the mouth- opening, by way of the 

 interradial or the ambulacral grooves. The denticles, or large 

 spines of the interradial angles, are ciliated from base to apex, 

 and frequently project right into the mouth. If a specimen, 

 after removal of the aboral body- wall and the roof of the stomach, 

 be placed mouth downward in a shallow dish of water with 

 suspended carmine particles, a very active upward eddy of 

 particles through the mouth will soon be found to occur. The 

 tube-feet of A.stropecten are pointed, and the animal habitually 

 crawls on or burrows in sand. It will, of course, find shell-fish 

 and other animals to feed on there, and, as a matter of fact, I 

 have taken out relatively large shells from the stomachs of dis- 

 sected specimens. But the sand will also contain abundance of 

 microscopic food-particles. The tiny spinelets which make up the 

 paxillae are weakly ciliated from the base to apex. The conjecture 

 may be hazarded that this ciliation, while keeping out debris, 

 will not prevent the more active microscopic organisms from 

 getting past the spinelets into the subpaxillary space. Once 

 there they will of necessity be swept towards the mouth, and 

 get the chance of passing into the stomach and being entangled 

 in mucus. An anus being absent, waste water would have to 



