548 PHORONIDEA. 



Of these new facts the second and third derive any importance they may have 

 in the discussion from the names which Masterman has given them. The state- 

 ments under (1) are more important, but they are supported by evidence which 

 is very insufficient when one remembers the size of the edifice which he erects 

 upon them ; e.g., the external openings of his proboscis pores are not shown in 

 section, nor is either of the openings of the collar nephridium. Further, one 

 sees no reason for speaking of a collar region of the body ; and lastly, the 

 development of these most important spaces has not been examined. All we 

 know on this subject is based on Caldwell's observations that there are two pairs 

 of coelomic cavities in the embryo. 9 The method of development of these spaces 

 in Adinotrocha is a highly important one from Masterman's point of view, and 

 should, in our opinion, have been examined by him. For, in the Balanoglossus 

 larva, the collar coelom and the trunk coelom are paired, whereas the spaces in 

 Adinotrocha with which he compares them are not paired, but, however they 

 may arise in the embryo, are continuous in one case dorsally and in the other 

 ventrally to the alimentary canal. 



