544 THE ANATOMY OF INVERTEBRATED ANIMALS. 



the excavation of the, at first, solid mesoblast. The perivis- 

 ceral cavity thus developed is what I have termed a sehizoecele. 

 But whether there is any fundamental difference between an 

 enterocoele and a sehizoecele is a matter for further inquiry. I 

 have referred above (p. 485) to the case of an Ophiurid, in 

 which the hollow diverticula of the archenteron, characteristic 

 of the Echinoderms, are represented by solid outgrowths of 

 the hypoblast. From this condition there would appear to be 

 an easy transition to that presented by the embryos of those 

 Oligochceta and Hirudinea, in which, though the mesoblast 

 is a product of the hypoblast, it contains no continuation of 

 the alimentary cavity, but eventually splits into a visceral 

 and a parietal layer, the interval between which is the peri- 

 visceral cavity; and there is much probability in Kowalew- 

 sky's suggestion that the longitudinal bands (Keimstrei- 

 fen) in which the mesoblast makes its appearance may be 

 homologous with the diverticula of the alimentary, cavity of 

 Sagitta. 



In this case, the sehizoecele will be an advance upon the 

 enterocoele, and the development of the perivisceral cavity 

 in Sagitta may represent the primitive mode of development 

 of all invertebrate perivisceral cavities. On the other hand, 

 it must be remembered that between the endoderm and the 

 ectoderm, in the disk of a Medusa, or in the bod} T of a Cte- 

 nophoran or Turbellarian, there is a gelatinous mesoderm 

 which occupies the position of the primitive blastoccele. 

 Now, this mesoderm may be, and probably is, a product of 

 the endoderm ; but any cavities which appear in it, such, for 

 example, as the water-vascular canals of the Turbellaria, can 

 have nothing to do with an enterocoele. 



Again, in the Tunicata, as we have seen, the atrium is a 

 kind of " perivisceral cavity," which is formed either by an 

 invagination of the ectoderm, in which case it may be termed 

 an epiccele ; or else it is a true enterocoele. Assuming the 

 former alternative, for the moment, to be that which ought to 

 be adopted, what is called a "perivisceral cavity " ma} T be one 

 of four things : 



1. A cavity within the mesoblast, more or less represent- 

 ing the primitive blastoccele. 



2. A diverticulum of the digestive cavit} T , which has be- 

 come shut off from that cavity (enterocoele). 



3. A solid outgrowth, representing such a diverticulum, 

 in which the cavity appears only late (modified oiteroccele or 

 sehizoecele). 



