hi<;marks on the developmi';nt of amphioxus. 355 



nocoele often appears in sections as a solid wedge of cells ; 

 whilst in the posterior region the splanchnoccele is not 

 completely formed; the ventral portions of the coelomic sacs 

 Bjve, it is true^ nearly shut off from the dorsal portions^ but 

 the coalescence of the ventral divisions to form the splanch- 

 noccele is not complete. The youngest larvfe in my collection 

 caught by the tow-net are too advanced to throw further 

 light on the subject. 



About five or six sections behind the only gill-slit formed 

 at this stage the collar cavities somewhat abruptly cease, and 

 the form of the section changes; the ventral ectoderm becomes, 

 so to speak, moulded on the gut (fig. 3) and the ventro-lateral 

 ridges disappear. The ventral part of the splanchnoccele can 

 just be made out as a cord of cells, intervening between gut 

 and ectoderm. 



When later larvos are examined, in which several gill- slits 

 have been formed (figs. 4 and 8), the collar cavities can again 

 be recognised. Owing to the extraordinary inequality of 

 growth of the two sides of the pharynx so characteristic of 

 these larvse, the first gill-slit, which was originally situated in 

 the mid-ventral line, becomes shifted up on the right side. 

 In this region the ventro-lateral ridges appear as before on 

 each side of the gill-slit; they have here, in accordance with 

 the changed position of the slit, become lateral. Further 

 back (fig. 8), however, the hinder gill-slits are still in the 

 middle line, and the ventro-lateral ridges containing the 

 collar cavities are distinctly seen on each side. 



In still later larvae (fig. 5) the ventro-lateral ridges become 

 still more accentuated. On the right side it can be distinctly 

 seen that the portion of the collar cavity contained in the 

 ventro-lateral projection is cut off from the rest. This portion 

 we may designate as metapleural coelom. A similar cavity 

 exists on the left side, and has doubtless a similar origin, but 

 owing to the disturba.nce caused by the appearance of the 

 mouth this could not be demonstrated. The greater part of 

 the ridge which Ave may now designate as atrial ridge is, 

 however, caused by a peculiar thickening of tiie ectoderm 



