W\^ILLE THOMSON, ON ASTERACANTHION VIOLACEUS. 103 



clawed dra^ving-room table. Circulation of granules takes 

 place rapidly in the peduncle and appendages^ but pressure 

 applied to the star-fish will no longer send the granular 

 contents of the disc into the peduncle, •while pressing the 

 peduncle does not inject the general cavity of the star, but 

 only renders turgid the circular canal and the radial ambu- 

 lacral vessels. A change now begins to take place in the 

 peduncle. It becomes more flaccid, and frequently portions 

 of the tubular feet are separated by deepening constrictions, 

 and shortly all that remains is an inflated sac hanging to the 

 under surface of the disc. The further development of the 

 disc, however, has in the mean time made the connections 

 of this sac more apparent. The integument has been in- 

 verted in the centre of the disc, and the inversion, gradually 

 deepening, has formed a mouth communicating with the 

 digestive cavity, and the vascular ring surrounding the 

 mouth has become more distinct. Five Avell-marked vessels 

 branch from this ring, each to the end of a ray, and the sac is 

 distinctly seen to join the ringbetween two of the radial vessels. 



A delicate opaque thread may be traced along the inferior 

 surface of the circular and radial vessels, and to end near the 

 point of each ray in a bright-red, double granular spot. 



I have mentioned before that early in the development 

 of the star-fish, five small plates make their appearance at 

 the re-entering angles of the rays on the apical surface of the 

 disc. These plates, without extending much in diameter, in- 

 crease in thickness by the development of an irregular 

 upper layer of calcareous tubing, so as at length to form five 

 porous calcareous masses. No diff"erence is perceptible at 

 this stage in the structure of the five masses. Tufts of 

 paxillae appear at the ends of the arms and above the 

 axillary and central calcareous plates, and the rows of 

 spines begin to be developed, which afterwards fringe and 

 fold over the ambulacral grooves. Up to this period the 

 integument of the doi'sal surface, between the calcareous 

 plates, is continuous, and uniformly granular. There is no 

 anal aperture, and there are no apparent respiratory pores. 

 The five inter-radial plates appear to rise through the mem- 

 brane, and I imagine they allow the water to filter through 

 them into the general cavity of the body. It is not till long 

 afterwards that one of these plates is developed into the 

 madreporic tubercle, and becomes connected by a special 

 membranous tube with the central ring of the circulating 

 system. 



It is difficult to form an accurate idea of the length Qf 

 time occupied by this process of development. la the 



r-f- 



