238 s. GOTO. 



taken pains to examine closely if tlie idea of MacBride can not 

 be applied to the entire perilieemal system in this sense, that 

 the cells forming the masses that give rise to the radial peri- 

 hœmal spaces wander out from the wall of the central portion 

 of the system, which arose, as shown above, from the entero- 

 cœl. I have not, however, found any evidence to support this 

 supposition, but on the contrary could trace all the stages from 

 the mass of a few cells to a definite vesicle, such as the one 

 figured in fig. 16. We must therefore conclude that the peri- 

 hsemal system {mihi) arises partly from the enterocœl and partly 

 from the mesenchyme. We may notice that a similar composite 

 origin of evidently homologous structures are known in other 

 groups {Balanoglossus, Ascidia). 



If now we institute a comparison between Ästerias pallida 

 and Ästerina gihbosa, we see that there are striking similarities as 

 well as marked differences in their development. The formation 

 of the epigastric enterocœl and of the secondary left posterior 

 enterocœl takes place in exactly similar ways in the two species. 

 The periœsophageal enterocœl and the stone-canal originate in 

 essentially the same way. As to the "dorsal sac", it arises also 

 from the same cavity in the two species, although it subsequently 

 undergoes difterent changes, as already described. On the other 

 hand, the circular enterocœl and the central portion of the peri- 

 hœmal system present markedly different modes of origin in 

 the two forms ; the circular enterocœl being formed in Ästerias 

 from the anterior enterocœl and tlie entire perihsemal system 

 from the mesenchyme, while in Ästerina the former is derived 

 both from the secondary left posterior, and the anterior, entero- 

 cœl, and the peripheral portion only of the latter from tiie 

 mesenchyme, the central portion having the same origin as the 



