GIBSON — CEPHALOCHORDA :' " AMPHIOXTDES." 239 



anus, and here presents the ap2icarancc of a lateral structure, the anus itself lying in the 

 middle line. Fihrillae are conspicuous in the gelatinous substance of the tins. 



The gonads {vide Vl. 15. fig. 18) are in exactly the same condition as that already- 

 described for Anij)hioxides, appeai'ing, on the right side only, as small masses of cells in 

 the hinder corners of the 22nd-41th myotomes (therefore belonging in reality to the 

 23rd-4uth). They still retain the stalk of connection with the parent myotome, and do 

 not form any projection into the atrium. Traces of gonads can be seen in about tea 

 more anterior segments as small cellular thickenings not yet evagiuated from the wall 

 of the parent myotome. 



WHAT IS "AMPmOXTDES'"! 



Golflschmidt has now adopted the view, and there can be very little doubt of its 

 correctness, that Amjjhioxicles is no adult, but the larva of some undetermined member 

 of the Eranchiostomidse. Most conclusive in favour of this view are the detailed corre- 

 spondences which AmiMoxides presents with the well-known larva of Branchiostoma 

 lanceolatum in the structure and relations of its pre-oral organ, mouth, endostyle, gill- 

 slits, metapleural folds, and other organs — correspondences which could hardly occur in 

 such perfection between a true primitive form and a larva displaying a mere recapitula- 

 tion of the same primitive organisation. This objection only iiolds, of course, against 

 the supposition that Aniphioaides is an adult of primitive character, not against 

 Goldschmidt's alternative hypothesis that it is racially or individually neotenic. Other 

 considerations, however, \^ill show us that there is not sufficient ground for the latter 

 hypotheses. In the first jilace, only one of his 29 specimens showed the gonads in at 

 all an advanced state of development, and only two others had even beginnings of them: 

 in the 190 specimens vhich 1 have examined, only a few showed any traces of them, 

 and in these they were in a very young stage. There is therefore no evidence to show 

 that " Amphioxides " can become sexually mature in the condition in which we know it, 

 and not sufficient ground for assuming that it does so without undergoing meta- 

 morphosis *. iSecondly, the occurrence of Amphioxides in great numbers in the 

 neighbourhood of land is strong jiresumptive evidence that it is the larva of some 

 littoral, and therefore metamorphosed, form. On the supposition that its distribution 

 was purely pelagic, it was quite justifiable to suspect, if not to assume, the power of 

 neotenic development, since the metamorphosed condition is an obvious adaptation to 

 a littoral existence, and the change thereto might well be inhibited by a pelagic one. 

 Now, even this ground for the supposition disappears. 



In his later paper (1906) Goldschmidt adopts the view that Amphioxides is a 

 " neotenisch weitergebildete Larveuform." This would seem to imply that the 

 " Weiterbildung " which gives rise to Amphioxides is an exceptional, or at any rate an 



* It is impossible to be certain as to the true nature of the animal, 20 mm. long, described by Cooper (1903) as 

 BrancJiiostoma pelar/icum. It does not look at all like A. pelayicus, and is described as having only 60 myotomes. 

 The 'Challenger' specimen, on the other hand (Giinther, 1889), is probably a genuine A. pelayicus. 



32* 



