550 DR. J. F. GEMMILL OX THE 



hand, in Solaster (see p. 560) it takes origin by a series of inter- 

 radial outgrowths as in Crihrella. 



(6) The 2>erUicemal ^^ouclies. — Perihsemal pouch I./II. has been 

 described (Masterman, 9, p. 392) as taking origin from the axial 

 sinus portion of the anterior coelom, being thus exceptional inas- 

 much as the rest of the pouches take origin from the posterior 

 coelom. I find that the exception is apparent, not real, and that 

 actually the pouch in question arises from the dorsal horn of the 

 posterior cceloin. At about the eleventh day the tip of this horn 

 bends leftwards, thus coming to lie between hydrocoele pouches I. 

 and II. A day or two later an opening is efiected between the 

 axial sinus region of the anterior coelom and the dorsal horn of 

 the posterior coelom a short distance back from the tip of this 

 horn. Two or three days afterwards the anterior coelom and the 

 dorsal horn of the posterior coelom again become closed ofi" from 

 one another, but in such a manner that the new septum cuts ofi' 

 the tip of the horn and leaves it for a time connected with the 

 anterior coelom, from which, however, it soon separates off to 

 become perihasmal pouch I./II. The process is illustrated on 

 PI. II. figs. 6-9. Occasionally one finds that the perihfemal 

 pouch in question remains longer in open connection with the 

 posterior than with the anterior coelom. 



MacBride (7, p. 360) and Goto (5, p. 235) in Asterina, and 

 Masterman (9, p. 392) in CribreUa, described perihaemal pouch 

 I./II. as arising from the preoral coelom. However, in Solaster 

 encleca (1, p. 35), Asterias ruhens (2, p. 260), Asterias, double 

 hydrocoele (4, p. 64), I found that the pouch arose from the dorsal 

 horn of the left posterior coelom, although, owing to the com- 

 munication between dorsal horn and axial sinus, it appeared at 

 first sight as if the pouch in question took origin from the last- 

 named cavity. In Solaster encleca (1, p. 35) and, I can now add, 

 in Solaster papposus, the origin of pouch I./II. from the posterior 

 coelom is perfectly definite. In Asterina, as in Cribrella, there 

 is a secondary communication between the dorsal horn of the 

 posterior coelom and the axial sinus. 



(7) Whether the larval posterior ooelom is morphologically single 

 or double. — So far as I can judge, in my early larvas of Crihrella 

 there are no signs of doubling of the larval posterior coelom. 

 Later, any features Avhich could be interpreted in this sense aflfect 

 chiefly the dorsal and ventral horns, and appear at a time when 

 the natural difierentiation of these horns leads to their dividing 

 more or less into right and left forks. In the case of the dorsal 

 horn the leftward fork is. at first the only one to appear, and 

 it becomes periha^mal pouch I./II. as described above under 6. 

 From the right one, wdiich appears later, there arises pharyngeal 

 pouch I./II., the genital pocket, and ultimately also the coelom 

 within arm-radiment II. In the case of the ventral horn, the 

 leftward fork burrows to the left or oral side of the stalk of the 

 preoral lobe in order to reach the internal aspect of hydrocoele 

 pouch I., while the right fork passes to the aboral side of the 



