254 On the (Esophageal Teeth of the Stromateidse. 



the general direction of the branchial arches. As in the last 

 case, tlie toothed epithelium of tlie upper branchial elements 

 projects backwards as a stout lobe into thcoesophaoiis, between 

 the opening of the sacs, and is followed by several series of 

 smaller lobes which line the interior of the sac, as in Psenes 

 capenais. 



Material was not available for the further examination of 

 this and other types, which would no doubt throw further 

 light on the structure and homology of these sacs, which we 

 may, however, reasonably conclude, from the above evidence, 

 are not strictly oesophageal, but are derived from an extension 

 backwards of pharyngeal epithelium in the form of two 

 jiouches. There seems to be at least two distinct types, in 

 w^hich the lining of the sacs is derived from the toothed 

 (■|)ithelium of the pliaryngo-brancbials, and the other in 

 which it is derived from the raker-bearing elements of the 

 gill-arch. The distinction may also prove to be of systematic 

 importance, in which case species resembling Psenes capensis 

 would be generically separated from species resembling 

 P. natalensis in respect of the nature of their oesophageal 

 teeth. 



The origin of the conspicuous paired saccular outgrowths 

 lined by pharyngeal epithelium is of interest, as they may 

 date from a time in the phylogen}'- of the Teleosts when the 

 gill-slit behind the fifth branchial arch began to close up, and 

 may now be all that remains of this gill-slit. Their develo))- 

 ment and further comparative study of their structure in 

 various groups (they also occur in the Tetragonuridoe) might 

 throw some light on this point. Certainly in Nomeiis tiiey 

 bear a striking, if superficial, resemblance in position to a 

 gill-slit. 



The physiological significance of the 'oesophageal teeth 

 which are found in these fishes is of interest, and has doubtless 

 some connexion with the nature of their food. The teeth of 

 the jaws are poorly develojied, and in some there are gill- 

 raker-like structures below the pseudobranchia?. It is known 

 that some feed on medusae. Nomeus — the well-known 

 Portuguese man-of-war fish — is said to find protection from 

 its enemies by hiding under the poisonous tentacles of the 

 Portuguese man-of-war Physalin, and perhaps securing the 

 jackal's share of its food ; but it may be suspected that its 

 object there, among medusre-pruducing gonophores as well as 

 pf>isonous tentacles, is not such an innocent one. The nature 

 of the food may again be associated with another peculiar 

 feature, well developed in some. Pores are described as 

 occurring on the surface of the body, and these, when traced 



