40 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



side, leaving only its middle third uncovered. In the 

 dorsal part of the branchial sac the two ciliated pads 

 which further back lay one at each side of the narrow 

 epipharyngeal groove are now seen to have moved further 

 ventrally, and they continue as traced forwards to run 

 obliquely down the lateral walls of the branchial sac (see 

 figs. 7 and 8), being now in fact the peripharyngeal ciliated 

 bands of ordinary ascidians, till finally they meet in the 

 ventral median line in the region of the endostyle. In the 

 middle of the lateral surface of the body (fig. 8) the ecto- 

 derm is more than one layer deep. There are large 

 columnar cells,* in each of which several nuclei are present,, 

 while at the bases of these there are smaller triangular 

 cells with their apices running in between the adjacent 

 columnar ones (PL III. fig. 10). 



The tail in the last few sections while remaining of about 

 the same width from side to side has become a good deal 

 thicker dorso-ventrally, and the dorsal and ventral surfaces 

 in the middle third have become covered by a thicker 

 layer of ectoderm, while inside that and co-terminous with 

 it is a well marked layer of muscle fibres (see PI. III. figs. 

 8 and 9). In the centre of the tail and occupying all 

 the extent between the two muscle layers is the notochord 

 with its usual undulating outline, and on its left side the 

 delicate nerve cord (see PI. III. fig 7). The tail retains 

 this structure through all the sections forwards to and in 

 front of the oral end of the body proper. 



The next section figured, No. 216 (PI. III. fig. 8), shows 

 the body becoming rapidly smaller, while the ectoderm 

 cells on the lateral walls have become enormous, and 



* It is probably these cells which Sanders (]\[on. Micros. Jour., Ap. 1874, 

 p. 141) mistook for stigmata in his Olkophuro, from Torquay. Possibly the 

 supposed gill-slits in Moss's remarkable appendicularian (Linn. Trans. XXVII, 

 p. 299.) may have been either large ectoderm cells or large glandular cells of 

 the endostyle. 



