1018 



in tlie region of the head. The term "liead-cavities" was introduced 

 by Balfour, who discovered them in the development of the Selachii. 

 In tlie branchial region of Amphioxus tlie coelom is divided by tlie 

 gill-clefts into as many "head-cavities" as there are clefts, and the 

 asyinmetij of these cavities necessarily coincides with that of the 

 gill-clefts. Here there is no new oddness to be explained. 



But Skdgwick (textbook part II, 1905, p. 34) wrongly understands 

 under "head-cavities" only the foremost entoderm sacs of Hatschek, 

 which are placed before the mouth, as it seems, as each others' 

 antimeres. Soon however they take up a mesial position, whereby 

 the right sac, a true head-cavity, is shifted before the left. In its 

 further development the left sac has not a single character from which 

 it could be deduced that it should be considered as a section of the 

 coelom. Since 1893 — the last time more in extenso in 1914 (I.e. 

 p. 63) — I iiave defended the opinion that the two sacs are only 

 apparent antimeres, that they originally must have lain not opposite 

 to, but before each other, and that the apparent antimeric occurrence 

 must be the result of the asymmetry of the larva. In this opinion 

 I however remained alone. The later investigators attached so much 

 importance to the tirst appearance — certainly an argument of great 

 ^veight — that they did not consider the important differences 

 which soon arise. Undoubtedly the symmetrical situation of 

 the foremost myotome (which must be considered as the second 

 of the row; the first is included in the right entoderm sac) played 

 a role in this. Yet in the rest — with exception of the third of the 

 i-ow — each myotome of the left side reaches, since its first appear- 

 ance, half the length of a myotome more rostrally than its antimere 

 of the opposite side. 



As this wry symmetry is found only in the region behind the 

 mouth, in front of it however not, it was apparently taken for 

 granted that the true gut in the snout would indeed be properly 

 symmetrical. This is however not the case. 



When the first gill sacs of the right side make their appearance 

 during the period of metamorphosis, the foremost of these does not 

 lie opposite its antimere, but opposite the second gill cleft of the left 

 side (cf. my paper 1914, fig. 3, 4, 5). In the fore-end of the 

 branchial region the left side of the gut has been displaced rostrally 

 over the whole length of a branchial metamere. ^) 



1) In the adult animal, (and also in the more caudally situated clefts of the 

 larva) this removal is not so large. Tlie foremost half of a left cleft then, as is 

 known, does not lie the whole, but half the length of a cleft rostrally to its 

 antimere of the right side. 



