290 ME. J. E. DUERDEN ON" 



stitutes the lining o£ the adult cavity. In fig. 19, e, f, y, A, the 

 limiting-layer appears, as it were, creeping round the ridges siill 

 remainiag, and thus separating them for disintegration. 



The original coelomic spaces in conaection with each radial 

 division, whether above or below the oesophageal region, enlarge 

 as the vacuolated tissue disappears and as the polyps increase in 

 size, and ultimately become the endocoeles and exocoeles of the 

 adult. These latter are thus shown to have their origin in, 

 primary spaces connected with an archenteron, exactly as occurs 

 in the formation of the coelome in the higher Metazoa, except 

 that partial disorganization supervening, the spaces are never 

 cut off from the central chamber. 



In all the larvae free, zooxanthellge and fragments of the dis- 

 organized tissue occur in the gastro-coelomic cavity. It seems 

 likely that these are ultimately expelled through the mouth 

 of the polyp. I did not observe such in Lehrunia ; but in 

 Aulactinia and other embryos of about the same stage 1 

 have watched extrusions of this character going on from the 

 interior ; small irregular masses of mucus-lilre matter, mixed 

 with what seemed to be yolk-particles and zooxanthellae, would 

 at times be passed out through the oral aperture. 



The mesenterial filaments along the free edge of the mesen- 

 teries probably represent a divided digestive tube. In which 

 case the space central to the mesenteries, the result of the dis- 

 integration, will be gastric, and the peripheral portion of the in- 

 ternal cavity will be coelomic. Hence the term " gastro-coelomic " 

 more nearly expresses the true morphological conception of the 

 whole of the internal cavity of the Scyphozoa than either 

 " coelenteron " or " gastro-vascular cavity." 



It follows that the entire imperfectly-chambered iuternaJ 

 cavity of the adult Lebrunia is onto genetically both coelomic and 

 gastric, and is a secondary formation having its origin in two 

 very different phenomena : — Firstly, in a primary system of 

 radiating, archenteric diverticula or coelomic spaces ; and 

 secondly, in the disintegration of a primary, undifferentiated 

 tissue. The former gives rise to the mesenterial chambers, 

 distinct from one another in the oesophageal region, but im- 

 perfect below ; the latter gives origin to the space (gastric) 

 central to the free edge of the mesenteries, and results in 

 the imperfect character of the mesenterial chambers below the 



