STUDIES ON THE PEOTOCHORDATA. 



325 



iu the wall of the branchial sac, and it is at the points of these 

 angles that the first two gill-slits simultaneously break through 

 into the atrial cavity (PI. XXX, fig. 12). 



It is this ingrowth of the wall of the branchial sac which I 

 interpret as a tongue-bar dividing a primitively single slit into 

 two. 



The following diagrams will render the mode of origin of 

 the first two slits as above described easy to understand. In 

 fig. 1 an atrial invagination is shown, causing a slight depres- 

 sion in the subjacent wall of the branchial sac (cf. PI. XXXI, 

 fig. 27). In fig. 2 this depression has deepened, and the wall 

 of the branchial sac has separated itself, except at the two 

 angles from the floor of the atrial invagination, the space thus 

 produced being continuous in front and behind with the body- 

 cavity. In fig. 3 the wall of the branchial sac has fused at the 

 two above-mentioned angles with the floor of the atrial cavity, 

 and, finally, in fig. 4 the two gill-slits (i and iv) have broken 

 through, and now open into the common atrial chamber. 



Ectoderm — ^=^ 

 Wall of branchial sac — ZZI 



Fig. 1 



Fig. 2. 



Fig. 3. 



Fig. 4. 



Atrial chamber. 



Tongue-bar. 



I have thus endeavoured to show that the first four 

 primary stigmata of Ciona, and probably of all the 

 Ascidise proper (A. mentula, mamraillata, &c.), repre- 

 sent and are derivatives of one primitive gill-slit. 



After the first four stigmata have become fully established 

 in the way just described, two more branchial apertures, v and 

 VI, arise behind them one after the other by independent perfora- 



VOL. XXXIV, PAKT HI. — NEW SEE. Y 



