138 WYVILLE THOMSON, ON SYNAPTA INHiERENS. 



ened by the addition of external layers, and in the fully-formed 

 plate the outlines of the rounded spaces are serrated with 

 blunt calcareous teeth. 



The form of the mature anchor has been often described. 

 Young anchors have the arms curved and the centre de- 

 pressed, like a Cupid's bow (PI. VI, fig. 3, d, e, h, k). Sometimes 

 the crown projects into a point (fig. 3, /), and to have the throat 

 widened and perforated is a common malformation. The 

 flukes of old anchors are often serrated."^ 



Brownish rosettes, granular and with striae radiating, from 

 a centre, the larger "03 mm. in diameter in a state of rest, 

 and at irregular intervals of a quarter to half a minute, sud- 

 denly contracting to half that size, and then more slowly 

 expanding, are scattered in an irregular row beloAv the cal- 

 careous ring. It is very difficult to make out the position of 

 these rosettes. They are certainly not superficial, for the 

 outer surface of the sarcode moves independently over them. 

 I am inclined to think that they are imbedded in the inner 

 transverse muscular layer, and to regard them as the 

 sphincters of pores allowing the passage of water out of the 

 body ca\ity. 



Similar structures have been observed several times in 

 young echinoderms. 



The wall of the central ring of the ambulacral system be- 

 comes more defined, and shortly it may be traced as a special 

 contractile membrane. 



The Ccccal j)rocesses to the tentacles enlarge, and are like- 

 wise bounded by a special wall. The cavity of each tentacle 

 communicates with the ring by a perpendicular slit-like valve 

 (PI. VI, fig. 1, c), which remains open when the tentacle is 

 retracted or at rest ; but when the tentacle is in motion, the 

 caecum is turgid with fluid injected from the ring and Polian 

 vesicle, and the valve is closed. In the space between two 

 of the tentacular canals, a single large Polian vesicle is gra- 

 dually developed from the lower surface of the ring. The 

 vesicle hangs loose in the body-ca^dty, and communicates 

 freely, Avithout any valve, with the ring. Its walls are highly 

 contractile, and it changes form continually, now hanging, 

 long and vermiform, into the body-cavity ; and now injecting 



* Tor excellent figures of the anchors and auchor-plates of this species 

 see the papers of Dubeu and Koreu, and of Woodward and Barrett, cited 

 above. 



I did not happen to see Mr. Woodward's figures of the development of 

 the anchors till after mine were drawn, or I should have thought it unne- 

 cessary to repeat them. 



