THE DEVELOPMENT OF AMPHIOXUS 41 



The rudiments of the secondary gill slits enlarge and become 

 perforated, and then an additional perforation appears at each 

 end of the series making eight in all. As these enlarge, chiefly 

 in the ventral direction, the primary gill slits below them are 

 moved down to the ventral side, and then over on the left side 

 of the pharynx to their proper position (Figs. 14, C; 15). As 

 a result of this migration the dorsal and ventral margins of the 

 primary gill slits are reversed, and it is evident that the mor- 

 phological ventral margins of these actually appeared first to 

 lie dorsally. As this migration is going on, first the secondary, 

 then the primary gill slits become divided by a downward 

 extension of a tongue-like process from the dorsal wall of 

 the slit which finally reaches the ventral side, divides the 

 originally simple opening into two, and forms the tongue bar 

 of the completed structure of the adult (Fig. 15). These 

 tongue bars appear in all the secondary slits except the first, 

 and in all of the primary slits except the first two and a vari- 

 able number of the last. As the primary gill slits approach 

 their final position the most anterior disappears completely 

 (Fig. 15) and soon after, the last five also disappear, reducing 

 the number in the primary series to eight, the same as the 

 number originally appearing in the secondary series. This is 

 the arrangement of the gill slits at the critical stage, the sym- 

 metrization of these structures being one of the important 

 phases of metamorphosis. After the reduction of the gill 

 slits to eight symmetrical pairs, arranged metamerically, 

 additional pairs form slowly, posterior to the primary and 

 secondary. These are the tertiary gill slits. They are not 

 metameric in their appearance and they displace anteriorly 

 the primary and secondary slits, so that the whole gill slit 

 series loses its metameric arrangement. 



Regarding the actual details of the formation of the gill 

 slits little is known except in certain of the secondary series, 

 or those of the right side proper. In these, and presumably 

 in most of the primary series also, true gill pouches seem to be 

 formed first, purely of endodermal cells. These are drawn 

 out ventro-laterally into small tubes closed by a very thin 



