THE DEVELOPMENT OF SAGITTA. 543 



with a large cleavage cavity, or blastoccele. One face of the 

 vesicle thus constituted now becomes invaginated, with the 

 effect of gradually obliterating the blastoccele, and converting 

 the spherical single-walled sac into a hemispherical, double- 

 walled, cup-shaped gastrula. The cavity of the cup is the 

 future digestive cavity ; the layer of invaginated blasto- 

 dermic cells which lines this cavity is the hypoblast, which 

 will become the endoderm ; and the outer layer of cells is 

 the epiblast, and will become the ectoderm. In this condi- 

 tion the embryo resembles that of the Leech in its early 

 state. The embryo elongates, and the aperture of invagi- 

 nation, or blastopore, eventually ceases to be discernible. 

 Whether it becomes the anus, or whether the anal aperture 

 is formed anew, is not certain. The nervous ganglia result 

 from the modification of cells of the ectoderm. The anterior 

 end of the primitive alimentary cavity, or archenteron, is at 

 first closed. It soon sends out an enlargement on each side, 

 so that the archenteron is divided into a central and two lat- 

 eral divisions. The central division opens externally and an- 

 teriorly by the development of the oral aperture ; and, as the 

 body elongates, it becomes the tubular intestine. The lat- 

 eral diverticula at first communicate with it, but they are 

 eventually shut off, and constitute the right and left perivis- 

 ceral cavities, their walls becoming converted into the cellu- 

 lar and muscular lining of those cavities. It results, from the 

 mode of development of the perivisceral cavity of Sagitta, 

 that this cavity, like the perivisceral cavity of the Brachio- 

 pods, and the " peritoneal " cavity of the Echinoderms, is an 

 enteroccele, comparable to that of the Hydrozoa and Actino- 

 zoa y but which, instead of remaining in communication with 

 the alimentary cavity, is shut off from it, its wall becoming 

 the mesoderm, and its cavity the perivisceral cavity. 1 



Nothing of this kind is known to occur in the Turbellaria, 

 Annelida, JVematoidea, or Rotifera / but when a perivisceral 

 cavity exists in these animals, it appears always to result from 



1 Kowalewsky's account of the development ofSagitta has been confirmed 

 by Butschli,* who has further determined the origin of the reproductive or- 

 gans, -which arise as outgrowths from the hypoblast; and the division of each 

 primitive enterocoele into two sacs— one for the head and another for the body. 

 It appears probable that the latter becomes subdivided by a transverse parti- 

 tion between the ovary and testis. Bi'itschli suggests that the segmentation 

 of the mesoblast, which forms the walls of the enterocoele, is a point of approxi- 

 mation between Sagitta and the Annelids. 



*"Zur Eutwickelungs^eschichte der Sa^Uta." (Zeitschrift fitr wiss. Zooloqie, 

 1873) 



