WATEE-VESSELS OF ECHINODEEMA. 221 



The portion corresponding to tlie stone-canal is not always con- 

 nected with the perisome. In the Holothuroi'da the connection is 

 broken close to the dorsal pore of the larva ; the latter disappears, 

 and the stone-canal hangs freely in the body-cavity, whence it takes 

 up water by a very complicated and porous terminal apparatus. 



There are further complications of the water-vascular system, 

 due to the formation of contractile diverticula of the water-canals 

 projecting into the body-cavity; these must be mentioned in addi- 

 tion to the ari-angenient, just sketched. These diverticula vary 

 greatly in character; on the circular canal they form large pear- 

 shaped vesicles (Polian vesicles) {ajy) ; where the ambulacral canals 

 pass into the sucking feet they form small ampullee {a), which 

 always project into the body-cavity, and which may be regarded 

 as enlargements or diverticula of the branches of the ambulacral 

 canals. They are cavernous in structure. Both these kinds of 

 oi'gans serve as receptacles for the fluid passing into the canals, and 

 owe then* structure to their adaptation to the function of this 

 vascular system ; that is to say, when the suckers are drawn in, 

 their ampullte are always filled, and when the suckers are pro- 

 truded the contents of the ampuUte swell them out. What the 

 ampullae are for the separate suckers, the Polian vesicles of the 

 cu'cular canal are for the whole system of canals ; that is, they 

 allow of a much more rapid action of the ambulacral structures, 

 whether these are pushed out or drawn in, than would be possible if 

 the quantity of fluid needed for the erection of each separate sucker 

 must be first taken in either by the stone-canal or the madreporic 

 plate. This activity of the ampullae of the suckers and of the 

 Polian vesicles of the circular canal is due to the contractihty of 

 their walls, in which a muscular layer has been made out. The dis- 

 tribution of the fluid is also regulated by muscular fibres, which 

 here and there enclose the canals. In addition to this the ciliated 

 epithelium which is found throughout the water-vascular system 

 serves to distribute and continually change the water, and so without 

 a doubt render it efiicient as an organ 

 of respiration. 



§ 178. 



The arrangement already sketched 

 in a general way applies most com- 

 pletely to the Asteroida. In them 

 the stone-canal is always inserted 

 into a madreporic plate, which, as a 

 rule, is i^laced interradially on the 

 dorsal surface of the body. In some ^^;.^:^Ti:^^::t 

 cases there are several (2-5) madre- aurantiacus (after E. Teuscher). 

 poric plates, and the number of stone- 

 canals is then also proportionately increased; this condition, how- 

 ever, does not remain constant in the species of the same genus. 

 It is to be regarded as the more primitive one ; it is important 



