206 ZOOLOGY. 



seem to be identical, it seems probable that the coelomic 

 fluid may be the carrier of katastases to the nephridia. 

 Thus the function of the coelom in Amphioxus would seem 

 to be that of a sort of auxiliary vascular system a func- 

 tion perhaps still retained, but in ever lessening degree, as 

 we pass up to higher and higher Vertebrates. 



We have seen that the coelom is to be regarded as 

 primarily the generative cavity, but as to its mode of 

 origin in evolution we can only speculate. It may have 

 been a portion of the original enteric cavity which occurs 

 in ccelenterates, separated off, and this view is supported 

 by its mode of origin in Amphioxus and certain other 

 cases. The other possibility is that it was a new cavity 

 developed in the mass of mesoblastic cells which were 

 developed between ectoderm and endoderm. The question 

 is whether the schizocoel or the enterocoel is the more 

 primitive. The latter alternative seems more probable. 



Besides the nephridia, the whole atrial epithelium appears to be 

 excretory in function, and there are a pair of " brown funnels" in 

 the roof of the atrium, just by the posterior end of the pharynx, 

 that are supposed, but have not been proved, to be excretory. 



14. Reproductive System. This is extremely simple, 

 consisting only of the gonads, already mentioned ( 1 and 

 fig. 104). These appear to be developed on the side- walls 

 of the atrium, into which they project ; hence a further 

 apparent resemblance of the atrium to a ccelom. But the 

 resemblance is just as illusory as in the case of the liver. 

 Really the gonads are in coelomic cavities in the atrial wall 

 (gonadic pouches), on the lining of whose walls ova or 

 spermatozoa are developed and fill up the cavity. 



If the atrium were a coelom the gonads would be deve- 

 loped as solid outgrowths from its wall into its cavity, and 

 the ova and spermatozoa would escape at the surface of 

 these solid outgrowths, as in higher vertebrates, In the 

 adult Amphioxus this appears to be the case, but study of 

 development shows that the myocoelomic pouches send off 

 prolongations into the atrial wall, and that the ovaries and 

 testes arise in the walls of these prolongations. 



There are no gonoducts, and when the ova or spermatozoa 



