STUDIES ON THE PEOTOOHORDATA. 323 



perforation, and while their primary long axis is still at right 

 angles to their future long axis (being longitudinal instead 

 of vertical), the tongue-bars form by a dovvngrowth from the 

 dorsal walls of the slits, while the primary slits have already 

 existed for several weeks without any sign of a tongue-bar. 

 The tougue-bars of the secondary slits of Amphioxus are, 

 therefore, considerably hastened in their development. A 

 little more and the two halves of the slit would break through 

 separately. Such a precocious development of a tongue-bar 

 has recently been described by Morgan (29) in Balanoglossus, 

 and this is what occurs in connection with the first two gill- 

 clefts of Ciona, as the following account will demonstrate. 



We must first, however, consider Kowalevsky's account 

 of the origin of these slits (21). He says, namely, that the 

 two first gill-slits of each side arise by the fusion of two out- 

 growths from the wall of the branchial sac with the floor of 

 the corresponding atrial invagination, and he figures the two 

 slits in his fig. 37, which was reproduced by Balfour. But it 

 is practically certain that what Kowalevsky figured as the 

 gill-slits were not the slits at all. If reference be made to 

 Balfour's 'Comparative Embryology,' vol. ii, p. 11, fig. 6, it 

 will be seen that in this figure the connection of the oesophagus 

 with the branchial sac is not even indicated. But the opening 

 of the oesophagus into the branchial sac is very large indeed, 

 the former arising as a direct continuation of the latter. Now 

 that portion of the wall of the branchial sac in which the gill- 

 slits occur, which may be called for convenience the branchial 

 tract, lies at first precisely at the niveau of the junction 

 between the branchial sac and the oesophagus, and Kowalev- 

 sky has apparently mistaken the small portions of the lumen 

 of the branchial sac which are visible on each side of the 

 branchial tract for the gill-slits themselves (compare my fig. 3 

 with the above-mentioned figure of Kowalevsky's). This mis- 

 take, it should be added, is, on account of the extremely small 

 size of the larvae and of the primary crowding of the organs 

 very easy to make and very hard to rectify if later stages are 

 not examined. 



