STUDIES ON THE PROTOCHOEDATA. 331 



is quite wrong, as may be seen at a glance from PI. XXX, 

 figs. 1 and 2, especially the latter. 



The most anterior portion of the wall of the enteric cavity 

 of the larva of Ciona — namely, that which lies in front of the 

 mouth and immediately behind the prseoral lobe — becomes the 

 endostyle, the primary position of which is, therefore, clearly 

 anterior and dorso-ventral (fig. 1). At a later stage the long 

 axis of the endostyle becomes changed in its direction owing 

 to a rotation through an angle of 90°, which the whole body 

 of the Ascidian undergoes with the praeoral lobe for the fixed 

 pivot about which the rotation takes place. In figs. 3 and 4 

 a stage is drawn in which this rotation of the body of the 

 Ascidian has just commenced. While in fig. 2 the long axis 

 of the endostyle was at right angles to the praeoral lobe, in 

 figs. 3 and 4 the former lies obliquely with reference to the 

 latter. Fig. 5 shows a stage long after the completion of the 

 rotation, and the praeoral lobe is posterior and has relatively 

 smaller proportions than at first ; the mouth, at first dorsal, is 

 now anterior and terminal ; and the endostyle, at first praeoral 

 and dorso-ventral, is now post-oral and longitudinal. The 

 long axis of the adult Ascidian is, therefore, at right angles 

 to the long axis of the larva. 



At a later stage the praeoral lobe commences to branch and 

 send out secondary processes, but the constancy of its primary 

 position, relations, and form is absolute. 



In Clavelina the peculiar behaviour of the adhering appa- 

 ratus (Haftapparat) has its efiect indirectly on the axial rela- 

 tions of the endostyle. In the free-swimming larva of Clavelina 

 the endostyle, owing to another hastening in the development, 

 is already in its definite antero-posterior position. It, how- 

 ever, lies at first dorso-ventrally in the unhatched larva as 

 described by Seeliger (31); but the change in the axis is 

 effected in a somewhat different way from that described above 

 for Ciona before the hatching, and therefore long enough 

 before the fixation of the larva. This is what Seeliger says 

 (pp. 33, 34) : '' An der vorderen Wand tritt der Endostyl 

 immer mehr als diflferenter Abschmitt des Kiemeudarmes 



