330 ARTHUR WILLET. 



deux portions bien distinctes, une portion prechordale et une 

 portion subchordale." 



When the larva of Ciona fixes itself the above-mentioned 

 anterior body-cavity which, as has been indicated, is at first 

 almost filled up by a compact mass of mesoderm proceeding 

 from the two mesodermic plates, swells up enormously, and 

 then presents the aspect of a large cavity bounded by ecto- 

 derm, and containing loose scattered mesoderm-cells. The 

 position and relations of the prseoral lobe with its contained 

 anterior body-cavity are shown very plainly in the figures 

 on PI. XXX. 



The adhering papillae flatten out shortly after fixation, and 

 give place to a thickened disc of ectoderm at the anterior 

 extremity of the prseoral lobe. 



The prseoral lobe has been figured by Kowalevsky (20) in 

 the case of Phallusia mammillata, and by Kupffer (24) in 

 the case of Ascidia canina, but was for them nothing more 

 than a peculiar " Haftapparat." 



Kupffer says (p. 158), ''Die prononcirt dorsale Stellung 

 der Mundoffnung darf nicht ausser Zusammenhang mit dem 

 am vordern Ende entwickelten specifischen Haftapparate 

 beurtheilt werden. Durch diesen wird die dorsale Lage der 

 Mundoffnung bedingt." The dorsal position of the mouth in 

 the Ascidian tadpole will be considered later. 



As to the endostyle, it is a remarkable fact, the sig- 

 nificance of which has not yet been properly recognised, that 

 its primary position in the free-swimming larva, and also for 

 some time after fixation, is at right angles to its definitive 

 position. In fact, the endostyle lies at first praeorally and 

 dorso-ventrally, and eventually post-orally and antero-pos- 

 teriorly. 



Kowalevsky (21) and Kupffer (24) both stated that the endo- 

 style extended from the neighbourhood of the mouth in front 

 to the hinder extremity of the branchial sac, and Kowalevsky 

 figures it accordingly along the whole ventral side of the 

 branchial sac of the larva. Kowalevsky's account of the endo- 

 style in the larva, which was repeated by Balfour (1, p. 18), 



