224 PERCY SLADEN TRUST EXPEDITION. 



is further remarkable for the possession of 30 gill-slits (no other specimen being found 

 with more than 28). In everything else, however, they show so exact a correspondence 

 with the normal specimens that it is impossible to regard them as specifically distinct. 

 I am inclined to attribute their peculiarities to rather greater age : possibly they may 

 represent the stage immediately preceding the onset of metamorphosis. 



I have also found one specimen of A. valdlvice in which a few of the anterior gill-bars 

 are everted and appear as simple rods of uniform thickness separating tlie wide rectangular 

 slits. This specimen showed side by side the two very different conditions which the bars 

 can assume : they are so different that one can readily understand Goldschmidt's belief 

 in an essential diversity of structure, in the absence of evidence showing them to be 

 alternatives, dependent solely on the contraction or relaxation of the branchial muscles. 



The disposition of the folds into which the branchial epithelium is thrown in the 

 contracted state, with its characteristic differences in the two species, is presumal)]y 

 determined by mechanical exigencies, which have led to the formation of definite lines 

 of weakness in the pharyngeal wall ; it can hardly be directly influenced by the branchial 

 muscles, since these ai'e inserted solely into the somatopleure, and no splanchnopleuric 

 fibres occur. Their formation follows as an obvious mechanical necessity on the 

 contraction of the muscles : the folding would be even more marked were it not for the 

 additional contraction of the sphincter muscle, reducing the outer opening of the 

 originally rectangular gill-slit to a small circular orifice. Since no folding of the outer 

 body-wall occurs, we must presume that it possesses considerable elasticity. It is evident 

 that Goldschmidt's view that the transverse muscles act as dilators of the giU-slits, 

 adopted in ignorance of the true expanded condition, is incorrect : the transverse and 

 sphincter muscles must act in harmony with, not, as he supposed, in opposition to one 

 another. The mechanism of dilatation is not so easy to understand as that of contraction ; 

 probably it is effected by the pressure of the coelomic fluid, as soon as the muscles have 

 relaxed. 



"With regard to the finer structure of the gill-bars, I am obliged to differ from Gold- 

 sehmidt on certain j^oints. The first of these concerns the course of the muscle-fibres. 

 He writes : " lunerhalb des Kieraenbogens verlaufen die Fasern vorwiegend dicht unter 

 der Korperwaiid . . . Aber sie steigen auch zwischen die Falten der entodermalen 

 Kiemenbogenwand empor " ; and he depicts in fig. 21 {A. valdlvice) fibres running 

 apparently in the dorsal (splanchnocoelic) wall of the coelomic canal, and in fig. 22 

 [A. pelagicus) in its lateral walls — also apparently splanchnocoelic. As I have said, no 

 fibres are inserted into the splanchnopleure ; and the only places where they can be said, 

 in any sense, to pass between the folds are at the edges of the slits, round which the 

 ectoderm is drawn in towards the interior and shares to a certain extent in the folding of 

 the endoderm : it is to this inward extension and folding of ectoderm that the complex 

 shape of the inner opening of the slit, so clearly figured and described by Goldschmidt, 

 is due. The structure of the gill-bars is practically impossible to make out in whole 

 preparations viewed from the side, and here he was probably misled by appearances ; if 

 real fibres were seen in the position depicted, probably (as in his fig. 19) they were some 

 which had slirunk away from the somatopleure. 



