GIBSON — CEPHALOCIIOEDA : " AMPHIOXIDES." 229 



If, as I can hardly doubt, the bands of cells are true sense-organs, we may see in 

 them a possible representative of the lateral-line organs of the lower craniates. The 

 possibility is of considerable interest in connection with Willey's suggestion (1894, 

 pp. 42-45) that the metapleures of Amphioxus and the hypothetical lateral fin-folds of 

 the craniate ancestor aie homologous, and that the lateral-line system arose in 

 conjunction with the primitive fin-fold to subserve the function of balance. 



The band of cells is best developed on the right metapleure of A. valdivite ; it runs 

 along the ventral edge of the fold, and extends beyond it anteriorly along the ventral 

 edge of the so-called " rostral fin " or " right cheek," a fact tending to support the 

 belief that this and the metapleures are primitively continuous structures. It dies out 

 towards the hinder end of the gill-region. The band on the left metapleure is not so 

 well-developed, and can be traced only from the level of the hinder end of the mouth 

 to about the middle of the gill-region. 



The general relations of the right metapleure in this species are well shown in 

 Goldschmidt's figs. 55-62. But he makes a mistake in his location of the anterior end 

 of the left metapleure. According to bis fig. 59, it would appear that this passes up 

 onto the left side into the upper lip of the stomodseum. Its true course is very 

 diiferent ; instead of ending dorsal to the mouth, high up on the left side of the body, it 

 ends actually on the right hand of the topographical sagittal plane, in close proximity 

 to the 7th and 8th gill-slits. It shares, in fact, in the torsion which alfects the gill-slits 

 and right metapleure in this region, and throughout its course, as in the Aniphioxits 

 larva (Lankester and Willey), runs strictly parallel to its fellow. The whole of the 

 anterior part of the fold is very poorly developed — it hardly stands ofi" from the body-wall 

 as a genuine ibid before the level of the 17th or 18th gill-slit ; in this anterior region it 

 is represented only by the band of clear cells and by its cavity — the pterygocoele, — here 

 minute, lying immediately internal to these. Its appearance and position immediately 

 beliind the mouth (at the opening of the 10th gill-slit) can be seen from PI. 15. fig. 12. 

 Opposite the 8th gill-slit, the band of clear cells dies out ; the pterygocoele can be traced 

 a little farther forward as a minute canal. This appears to communicate with the 

 splanchnocoele at its anterior end, but the openings are irregular and not above 

 the suspicion of being artefact. Such communications as appear to exist between the 

 right pterygocoele and splanchnocoele appear to me to be even more suspicious. 

 Goldschmidt states that openings exist in both cases, but it is clear that he did not see 

 the true anterior termination of the left pterygocoele at all. I cannot regard the 

 question as settled one way or the other. 



In A. pelaglaus the general relations of the metapleures are very similar, with the 

 differences that the right one is not carried up onto the right-hand^side in the mouth-region, 

 and that the left one is much better developed anteriorly than in valdivice. The right 

 fold, however, forms a shelf projecting over the anterior slits exactly as in A. valdivice, 

 M'ith the difference that instead of hanging down on the right side, it is inserted nearly along 

 the middle line and projects upwards towards the left. Fig. 9 (PI. 15), taken through 

 almost the same region as Goldschmidt's fig. 73, shows this. On the outer face of this 

 shelf the thickened ejjidermal band is conspicuous in most whole preparations, forming 



SECOND SERIES. — ZOOLOGY, VOL. XIII. 31 



