246 PERCY SLADEN TRUST EXPEDITION. 



secretion of the club-shaped gland, serves to convey food-particles into the " pars 

 nutritoria." Thus the asymmetry of mouth and endostyle is primitive, not secondary : 

 both orsrans arose de novo in the position in which avc now see thcjm, at the time when 

 the ancestor of Amphioxides first acquired gill-slits. The position of the formerly- 

 existing mouth is not specified. Incidentally, it is assumed that the rostral extension 

 of the uotochord was already present at the pliylogenetic stage when these changes took 

 place, in opposition to AVilJey (1891), who rcgtirds it as an adaptation to life in the sand. 



Eefore considering the supposed course of evolution from Amphioxides to ylnq^hioxiis, 

 I may offer a few criticisms on the direct evidences of primitiveness derived from the 

 structure of the former. Of such evidences, Goldschmidt insists jorincipally on that of 

 the gill-slits. Eumetamery of the gill-slits is, no doubt, a primitive character, but in 

 view of the extreme unlikelihood that it really obtains in A. valdivice (vide supr-a, 

 p. 227), no very great importance can be attached to its apparent occurrence in 

 A. pelagicus. The bilaterally symmetrical structure of the gill-bars, if it really occurred, 

 would be an imj^ortant argument for their jirimitive character. Goldsclimidt lays 

 considerable stress upon it, and writes : — " Man miisste den Thatsachen schon ziemlich 

 Gewalt antvm, wenn man etwa annehmen wollte, dass durch die sekundare Verlagerung 

 einer linksseitigen Kiemenspaltenreihe in die Symmetrieebene sie auch sekundiir eine 

 innere Symmeti'ie angenommen habe." I cannot but think that far less violence to 

 facts is needed for this supposition than for Goldschmidt's own as to the manner in 

 wdiich the perfect symmetry of the gill-slits of the adult Amphioxus was secondarily 

 established. But, in fact, the symmetry of the gill-bars in Amphioxides is of the 

 slightest, and such as it is may perfectly easily have arisen secondarily. I hope to have 

 shown above that the assumption of a paired structure of the endodermal portions is 

 uujustifitid. The symmetrical disposition of the muscles is no indication of paired 

 structure, being conditioned solely by mechanical needs. Moreover, in A. valdivice, at 

 any rate, the position of the gill-slits is far from symmetrical : and I cannot find entirely 

 satisfying tlie explanation that here the sagittal plane of the whole pharyngeal region 

 has been pushed to the right, owing to the development of a long stomodaeum, consequent 

 upon the increased ventral extension of the muscles. 



Another fact which weighs against the primitiveness of the median series of gill-slits 

 is the occurrence of an asymmetrical series of excretory tubules, on their left side only 

 (the right series being represented, if at all, by mere " Anlagen," far up on the right- 

 hand side). Whether, with Boveri, w^e homologize these tubules with the pro-ncphric 

 tubviles of the craniates, or regard them, with Goodrich, as the probable homologues of 

 the annelid nephridia, or even if we accept neither hypothesis, there can be little doubt 

 that they are essentially and primitively paired structures. Goldschmidt's apparent 

 acceptance of Boveri's homology, and his views as to the relationship of Amphioxides to 

 the craniates, undoubtedly demand their occurrence in a double series symmetrically 

 disposed with regard to the median slits. Their occurrence as actually determined 

 points strongly to the conclusion that the slits have secondarily assumed the median 

 position during the temporary supjires'sion of their fellow series. Positive proof of the 



