110 EDWJN S. GOODRICH. 



siuus" is ''a completely closed vesicle" with an internal 

 lining of distinct epithelium (13). 



De Selys Longchamps (16), on tlie contrary, denies the 

 existence of a complete anterior septum. He finds " des 

 formations mesenchymatiques . . . particulierement deve- 

 loppees entre le ganglion nerveux central et le ' subneural 

 gland ' . . . constituent en ce point une membrane, ratta- 

 chaiit la parol dorsale du lobe preoral a la parol de I'oesophage 

 . . . il ii'y a pas, pour cela, separation entre les deux cavites." 

 Further, he rightly insists that there is a single haemal pre- 

 septal cavity: " Daus la cavite anterieure, commune au lobe 

 preoral et a la region collaire, un mesenchyme surtout 

 abondant dans la premiere de ces deux regions ; dans la 

 cavite posterieure du corps, un mesoderme epithelial, repre- 

 sente par une splanchnopleure et une somatopleure, et un 

 mesentere sous-intestinal typique." In fact, the Belgian 

 observer confirms Caldwell's account. 



Ikeda's statements, accompanied by elaborate figures, are 

 even more convincing (9). He shows that muscular strands 

 pass from the sides of the ganglion to the oesophagus and 

 to the body-wall, that a membrane stretches across these 

 muscles, forming the dorsal limit of a ''posterior recess" of 

 the pre-oral region. This recess is a sort of pocket bulging 

 backwards beneath the brain. It opens forwards into the 

 spaces of the pre-oral hood, which communicate widely with 

 the main space in the " collar " region. Blood-corpuscles 

 may be found in all these spaces, and they are of a haemocoelic 

 nature. 



My own results entirely confirm those of Ikeda on this 

 question of the anterior septum. A subneural pocket is 

 present (figs. 1, 4, 12, and 9) ; it is lined with a conspicuous 

 epithelium, and when seen in frontal sections may present 

 the appearance of being a completely closed cavity (tig. 12). 

 But, as I have endeavoured to show in figures 1 and 9, the 

 pocket is widely open ventrally, and its cavity is merely a 

 portion of the general pre-septal cavity. It is this clear space 

 which allows of the ganglion being withdrawn by the muscles 



