METAMOKPHOSIS OF ASTERINA GIBBOSA. 235 



same way as in Asterias pallida, as may be followed in the 

 accompanying plate, and will therefore be passed over. The 

 periœsoj^hageal enterocœl originates exactly as in Asterias, and 

 as described by MacBride. As to the formation of the stone- 

 canal I also find the statement of the author just named 

 confirmed by my observations, which are therefore opposed to 

 those of Ludwig, according to whom it originates from the hy- 

 drocœl and grows out towards the pore-canal, with which it 

 finally unites ['82, p. 38]. 



We shall now pass ou to the formation of the perihsemal 

 system, on which I have to record some observations that do 

 not agree with those of MacBride. In the following description 

 I shall retain the words I used in my former paper, and speak 

 of the "circular enterocœl" and the ''j^erihœmal system." The 

 former is equivalent to the "inner perihœmal ring" of authors 

 and the latter to the "outer perihsemal ring" plus the radial 

 perihœmal spaces. I have shown that in Asterias the circular 

 enterocœl is formed from the anterior cœlom and that the whole 

 perihsemal system originates in the form of mesenchymatous 

 spaces. In Asterina gibbosa the process is still more complicated, 

 for, anticipating the results of my observations, I may state at 

 once that a portion of the perihsemal system (mihi) is found to 

 have an enterocœlic origin. In MacBride's opinion the entire 

 perihsemal system of authors originates as five interradial out- 

 pocketings from the secondary left posterior enterocœl. A 

 thorough examination of my series of sections has, however, 

 shown that there are only four such diverticula, and that in 

 the inter radius 4/5 '^ the diverticulum is replaced by the 

 posterior extension of the anterior enterocœl. Diverticula 1/2, 



-1/2 of MacBride. 



