THE STAEFISH S0LA8TEE ENDECA. 21 



advanced a degree of development as the succeeding members of the series, thus 

 offering the sharpest contrast with pouches VIII and IX, v?hich are still lagging far 

 behind. 



Although the details fall to be given later (p. 27), it may be stated here that the 

 ring-canal of the water vascular system is completed in such a way that the cavity of 

 the preoral lobe occasionally remains in continuity, though for a very short time, with 

 the axial sinus ccelom by means of a stalk, which, as in Asterina (Macbride, 15), 

 is encompassed by the hydrocoele ring. Soon, however, the cavity of the preoral lobe 

 disappears, and the sucker atrophies, migrating at the same time outwards beyond the 

 circle of the ring-canal. The last remnant of it is to be looked for as a stalked papilla 

 attached a short distance from the margin on the oral surface (PI. I. fig. 13). In four 

 cases out of five this papilla lies in interradius IX/I, while in the others it either lies 

 just outside of pouch IX or in the interradius between that and pouch VIII. The 

 variation is of no importance, and probably occurs in connection with the adjustment 

 that has to take place between the hydrocoele rays and the aboral arm-rudiments 

 (p. 22). 



The stone-canal opens into the ring-canal of the hydrocoele between the origins of 

 radial pouches I and II (p. 31). 



Considering the manner in which the radial pouches of the hydrocoele are developed, 

 and in particular the perfectly definite sequence in which the last four of them appear, 

 one can have no hesitation in numbering the series by the method adopted above — that 

 is, by commencing it with pouch I on the dextral * side of the preoral lobe as one looks 

 at the oral surface, and continuing in the same direction from that point. 



Leaving out of count such methods of numbering the rays of Asteroids as are based 

 on other than developmental grounds, we have to distinguish two systems. In one of 

 these the numbering proceeds as described above [e. g., Macbride for Asterina), yvh.\\.e 

 in the other the order is precisely the converse — that is to say, it begins with ray I on 

 the sinistral side of the preoral lobe as one looks at the oral surface, and continues the 

 series counterclockwise from this point (Ludwig, 12, and Goto, loa, for Asterina; 

 Goto for Asterias pallida, 9, 10 ; Masterman for Crihrella, 18). The order of develop- 

 ment in Solaster establishes a presumption that the former of the two systems of 

 numbering is the right one for Starfish in general. 



The short pause between the formation of the first five radial pouches and that of 

 the sixth is an indication that the Solaster family was derived from a five-rayed 

 ancestor. Further evidence on this point will appear in connection with the develop- 

 ment of the anus. I am indebted to Prof. J. W. Gregory for the note that an 

 apparently authentic member of the family, Solaster moretonis Forbes, occurs as far 

 back as the Oolite (Forbes in Mem. Geol. Surv. U. K. 1856, dec. v. pp. 1-3). 



* See footnote on p. 13. 



