616 



case with the larva of Asterina gibbosa, had one insufficient material, 

 and if consequently many stages were missed out, that I cannot imagine 

 this explanation to be the true one. It is very improbable that a creeping 

 larva should have a more circuitous development than a pelagic one. 



The points to which I wish to draw special attention are as 

 follow. 



Mr. Goto asserts a) that the sac denominated by me the right 

 hydrocoele, is merely a portion of the axial sinus separated compa- 

 ratively late and originating on the left side of the larva, 



b) that the radial perihaemal canals with the outer perihaemal 

 canal which connects them are formed by the hollowing out of solid 

 masses of mesenchyme and have no genetic connection with the body 

 cavity or coelom. 



Now with regard to a) the earliest stage which Mr. Goto descri- 

 bes corresponds to a late one of the organ described by me in Asterina 

 gibbosa. I traced the organ back to the young larva and found it there 

 originating as a bud on the right side of the anterior coelom: I further 

 gave what seemed to me to be most decisive evidence as to its real 

 nature by describing a number of cases where it had undergone a 

 more or less similar development to the' left hydrocoele. After a re- 

 examination of my preparations, I am fully convinced of the truth of 

 every word I published in 1896. One of the best of my preparations 

 on this point was shown at the meeting of the International Zoological 

 Conference in Cambridge 1898. After once separating from the anterior 

 coelom, the right hydrocoele never has any further connection with 

 it. Mr. Goto's statement that the right hydrocoele is a part of the 

 axial sinus is easily explicable by the method of preservation which 

 he employed. He preserved his larvae in corrosive sublimate and glyce- 

 rine and embedded them in paraffin. When commencing my studies on 

 Asierina gibbosa I had material preserved by similar methods and I 

 also embedded in paraffin. Eut I got no reliable results until I used 

 material drenched with osmic acid and embedded in celloidin. I was 

 several times tempted to state that the right hydrocoele had a con- 

 nection with the axial sinus and also with the general coelom, but 

 with better methods the entire isolation of the cavity after its first in- 

 ception could be clearly seen. 



With regard to b) also Mr. Goto has in my opinion entirely mis- 

 sed the first stages in the development of the structures in question. In 

 Asterina gibbosa the first stages are found in stages E and i^when the 

 metamorphosis is just beginning and then the connection with the 

 coelom is clear and unmistakeable. I found it in every larva of the 

 proper age examined. Mr. Goto found the rudiments of the perihae- 



