GIBSON— CEPIIALOCHORDA : " AMPHIOXIDES." 247 



coiToctness of tliis view Avould be given, could it be sliown that the branchial muscles of 

 Ampldoxides are innervated from tlie left side only. 



AVilley (1906) has already pointed out the improbability of Goldschmidt's view as to 

 the respiratory mechanism and function of the lateral pharyngeal folds, and we may 

 now regard it as fairly certain that in the living AmpMoxides, as in the Brancluostom.a 

 larva, an expanded condition of the pharynx, in which the folds are obliterated, is the 

 normal one. With this admission, his views upon the function of the endostyle, and his 

 physiological explanation of the asymmetry of this organ and of the mouth, lose the 

 greater part of their value. 



Willey also calls attention to tlie fact that the peripharyngeal bands of the adult are 

 already represented in the BrancMostoma larva : I have shown this to be true, at any 

 rate for the left band', in Amphioxides. M. Legros has demonstrated to me thnt a i)and 

 of clear cells representing the future endostyle and a thickening marking the position of 

 the future right gill-slits are also present in the Branchiostoma larva before the beginning 

 of metamorphosis : although these do not seem to be present in AmpMoxides, their 

 significance is just as great in one as in the other. We might, of course, suppose that all 

 these are purely secondary structures, belonging to the adult type of organisation only, 

 but precociously developed in the larva: possibly even, as Goldschmidt does for the 

 endostyle, we might find in their asymmetrical position a physiological significance and 

 interpret it as primitive. The generally accepted view that they are primitively 

 symmetrical organs, ccenogenetically shifted into an asymmetrical position during larval 

 life, remains vastly the most probable. 



A priori, the primitiveness of the asymmetrical arrangements found iu Amphioxides 

 is rendered highly improbable by the general consideration of symmetiy and asymmetry 

 elsewhere in the animal kingdom. There can be no question but that asymmetry is 

 secondary in the vast majority of cases where it occurs, being arrived at either by 

 imequal growth and diversity of development of originally paii-ed equivalent organs, or 

 by the suppression of one such, or else by the displacement of primitively median 

 organs, only very rarely by the appearance of asymmetrical organs de novo. 



We may now pass to the explanation of the metamorphosis of " ylmphioxns" and of 

 the phylogenetic transition of which it is held to be the recapitulation, leased on the 

 supposition that the organisation of Amphioxides and of the Av/phioxus larva is 

 primitive. From the belief tiiat this supj)osition makes a satisfactory explanation of 

 the facts possible, where other hypotheses fail, Goldschmidt derives an increased 

 confidence in its validity. 



From the stage represented by A. pelagicus, the supposed evolution of which we have 

 already traced, the lines of evolittion leading to A. valdioice and to Aniphioxus are 

 supposed to coincide for a short distance. The musculature acquires a gi-eater ventral 

 extension and so necessitates the developraeut of a long stomodaeum, which causes a 

 torsion of the sagittal plane towards the right. In this way the right-sided position of 

 the gill-slits in A. valdivice and iu the Branchiostoma larva is accounted for. In point 

 of fact, however, we find a particularly slight ventral extension of the musculature, and 



33* 



