the Convergence of Types by Pelagic Life. 87 



numerous bands are almost always formed by the Noctilucce, 

 Medusa 1 , Ctenophora, Sagittce, Copepod Crustacea, Mysides, 

 Pteropod Mollusca, &c. 



It is evidently the resemblances of adaptation that Sagitta 

 presents to Amphioxus, to the Heteropoda, and to Tomopterts 

 and other Annelida, that have determined zoologists to place 

 the Chcetognatha sometimes among the Vertebrata, and some- 

 times with the Vermes, at a time when neither their anatomy 

 nor their embryogeny was sufficiently known. 



The arrangement of the nervous system evidently removes 

 all possibility of an immediate approximation of Sagitta and 

 the Vertebrata. Leydig and Kowalevsky have justly indi- 

 cated that the nervous system resembles that of the Mollusca. 

 It may also be compared to that of the Annelida ; and in 

 this there is nothing surprising, from what we now know of 

 the close relationship (demonstrated by embryogeny) between 

 the group of Annelida and that of Mollusca (Brachiopoda, 

 Chitons, Dental ia, &c). 



On the other hand, this same nervous system removes the 

 Sagittoi from the true Nematoids ; and their attempted approxi- 

 tion to Chcetosoma does not appear to be completely justified 

 by what we know of the organization of the latter. 



The presence of chitinous setse is another character in 

 common with the Annelides ; and indeed, from the anatomical 

 point of view, the only serious argument that we can oppose 

 to those who would unite the Sagittal with ringed worms is 

 the absence in the former of any metameral structure, even in 

 the embryo. This character, on the other hand, approximates 

 the Chaetognatha to the phylum of the Mollusca, or, in a more 

 general way, to the ancient animals from which have been 

 derived on the one hand the Mollusca and on the other the 

 Annelida. The presence of the vibratile disk and of the 

 lateral invaginations also reminds us of the arrangement 

 observable in groups allied to those inferior types of which we 

 are speaking — for example, in the Rotifera or in the embryos 

 of certain Annelids. 



The very peculiar embryogeny of Sagitta (formation of a 

 secondary general cavity), however, does not allow of out- 

 placing them directly among either the Mollusca or the 

 Annelida. It is, in fact, a dilated embryogeny (without the 

 formation, either primitive or secondary, of a nutritive vitellus) 

 which is the indication of high antiquity of the type. It may, 

 however, be the case that the development of the Annelides, 

 which is not sufficiently known, represents the condensed 

 form (with nutritive vitellus) of the evolution of the Chaeto- 

 gnatha. 



