THE STAEFISH S0LA8TEE ENDECA. 35 



(see p. 32). I must, however, leave these points in some doubt meantime, as certain 

 discrepancies appear in my difterent series. The pockets or diverticula in question 

 give rise to the so-called pharyngeal coelom which, in the adult, surrounds the lower 

 portion of the gastric wall. 



During early metamorphosis the stalks of origin of these diverticula seem to 

 become obliterated, but they themselves, extending in the manner previously described, 

 have already met and united on the side away from the madreporic interradius. Later 

 they complete the circle of the pharyngeal coelom by becoming continuous also 

 over the region of the madreporic interradius. The sheet of tissue which now separates 

 the pharyngeal and hypogastric coeloms may be called the circular hypogastric 

 mesentery. This mesentery becomes perforated in each interradius, leaving a flat band 

 or ligament connecting the stomach- wall with the floor of each of the rays. Still 

 later, these connecting bands or ligaments also become perforated, each in its mid- 

 radial line. The hypogastric and the pharyngeal coeloms now communicate with one 

 another both by radial and by interradial apertures. The remaining parts of the 

 circular hypogastric mesentery then break up into more or less isolated ribbons, forming 

 the numerous ligamentous strands which in the adult pass downwards and outwards 

 from the stomach-walls to an attachment over the junction of the ambulacral and the 

 adambulacral sets of ossicles. 



The External Oral Circular Sinus. — The external oral circular sinus or circular oral 

 canal of the perihsemal system, together with its extensions along the rays, is formed, 

 as Macbride first described for Asterina, from a series of small diverticula, all but one 

 of which arise from the posterior coelom, and which interdigitate with the radial pouches 

 of the hydrocoele. Of these, the first three in Solaster, i. e. those between pouches 

 I and II, II and III, and III and IV, become evident two or three days before fixation ; 

 the fourth soon follows, while the rest are delayed in correspondence with the later 

 formation of the pouches between which they lie. Thus, at a stage in metamorphosis 

 when the larval sucker is no longer functional, the perihsemal diverticula between 

 pouches VII and VIII and between VIII and IX are still in course of being 

 budded off from the ventral horn (PL IV. fig. 45). 



The origin of perihaemal pouch IX/I from the anterior coelom has been referred to 

 on p. 31, and also the possible contribution to this pouch on the part of the posterior 

 coelom. In the case of pouch I/II, which in Solaster undoubtedly is to be described 

 as arising from the posterior coelom, I have noted an instance (towards the very end 

 of the free-swimming stage) of a fine canal connecting the pouch in question with the 

 dorsal angle of the axial coelom, but believe that this condition is exceptional. 



In Asterina (Macbride, 15) and Cribrella (Masterman, 18), by the same method 

 of numbering, pouch I/II arises from the anterior coelom, while all the others 

 (including pouch V/I, which corresponds, in position at least, with IX/I of Solaster) 

 arise from the posterior coelom. The differences are thus sharp, but are, perhaps, 



f2 



