GIBSON— CEPHALOCIiORDA : " AMPillOXfOES." 243 



posterior end of the mouth-openiag and of Hatschek's pit both lie relatively farther back 

 in the metamorphosed animal than in the larva. Fnrther, both metamorphosed animals 

 are considerably smaller than the largest larvae obtained. 



But on the whole these difficulties seem to me to be outweighed, once again, by a 

 distributional coincidence. Por Andrews figures, as a stage in the life-history of 

 Jsymmetron, not only a larva whicli can hardly be other than A. pelagicus, but also 

 a young metamorphosed animal extremely similar to mine, as a comparison of his tig. 3 

 with my PL 15. fig. 2 will show. Thus we have in the Bahamas and Maldives closely 

 similar parallel series of three stages, with undoubted correspondences between them to 

 suggest that all are stages in the life-history of a single species. That they do, in fact, 

 belong to one and the same species is, I think, a justiriable deduction. 



The connection is perhaps weakest between the two younger stages, where, however, 

 tlie intervening metamorphosis would naturally be expected to bring about a con- 

 siderable change. That the young metamorphosed animals belong to A. lucayamini is 

 indicated by their corresponding myotome formula, their smooth oral cirrhi, and the 

 continued absence of ventral (in-ray spaces — in the corresponding stage of Braiiohiosloma 

 laiiceolatuni the latter have definitely made their appearance. The number of distinctly 

 developed gonads is not yet so great as that possessed by the adult Asyvimetroti ; 

 this is only what we should expect in a young specimen. The posterior limit of the 

 gonads coincides, within two segments, in Amphioxides pelagicus, the metamorphosed 

 animal, and Asymmetron lacayanum, a further argument for their specific identity. 



( Asymmetron lacayaimm (Andrews). . 15— i:3 



Segments in which gonads I Metamorphosed animal 23— ii 



Amphioxides pelagicus (1) . . . . 18-42 



(.2) .... 20-43 



are distinctly developed. 



< 



Andrews's specimen, to judge from the length of the liver-cfecum and the more 

 tapering tail, is rather older than those which I have described. It has apparently 

 15 left gill-slits, most of them completely divided by tongue-bars, thus resembling the 

 smaller of mine. 



The exact metamorphic history of the pharyngeal region is a little difficult to 

 reconstruct from imagination. In a typical A. pelagicus there are 27 median gill-slits, 

 the last corresponding to the division between 27th and 28th segments. In the larger 

 but — ^judging from tlie number of oral cirrhi — the younger of my metamorphosed 

 animals, the 21st and last left gill-slit lies opposite the ventral end of the 28th myotome. 

 We must therefore suppose that, with the closing of the most posterior slits during 

 metamorphosis, those in front have been shifted bodily backwards, a fate which the 

 mouth and Hatschek's pit would seem to have shared. In my other specimen, and in 

 Andrews's since there are only some 15 left slits, a larger number of the posterior ones 

 would seem to have atrophied ; a certain amount of irregularity in this respect is found 

 in Branchiostoma pelagicum, and need not therefore surprise us here. 



Until more intermediate stages are discovered, it would no doubt be premature to 

 state dogmatically that Amphioxides pelagicus is the larva of Asymmetron lucayanum. 

 We can only express the opinion that it probably is so. It is certainly remarkable that 



