THE MOEMYRID^, NOTOPTEKIDiE, AND HTODONTID^. 215 



Mormyrides vis-a-vis des Albulides, c'est a dire qu'ils peuvent 

 en etre coBsideres comme modification excentrique." In con- 

 sidering the possibility of evolution of the Notopteritlse from the 

 ancestral Hyodon, one must not lose sight of the fact that 

 Notopterus — in the presence of the large lateral cranial foramen 

 bounded by the squamosal, epiotic, and exoccipital, in the 

 attempt (a futile one, it is true) of the thin scale-like supra- 

 temporal to cover it, and in the presence of a paired tendon- 

 bone of considerable size projecting down from th.e side of the 

 second basibranchial — exhibits characters strikingly constant in 

 the Mormyridse, but not possessed by Hyodon. 



Although in both Notopterus and Syodon there are vesicles 

 of the swim-bladder on the lateral face of the otic region of the 

 cranium, it does not necessarily follow that these structures 

 have had a common origin. The connection between the swim- 

 bladder and the ear must not be relied upon too implicitly as 

 indicating close relationship between such fishes as jjossess it. 

 That it has arisen independently in different groups is evident 

 from the remarkable difference between the methods by which 

 the result is arrived at. Compare, for instance, Clupea on the 

 one hand and the Ostariophysi on the other. Stannius (Handb. 

 d. Anat. d. Wirbelth. i. p. 2) mentions that there is a connection 

 between the swim-bladder and the ear in the Macruridae among 

 the Anacanthini, and in the Berycidae and Gerriidse among the 

 Acanthopteri ; while Sagemehl (Morph. Jahrb. x. 1885, p. 51, 

 footnote) observes that it occurs in the Gadidoids Physiculus 

 and TJraleptus, and in the Scleroderm JBalistes. 



In conclusion, I wish to point out with some emphasis that 

 the remarks in this paper apply to the skull only. I have 

 purposely avoided all reference to other parts of the skeleton 

 and to the soft structures of the body, since these were foreign 

 to the investigation undertaken. The above deductions as to 

 inter-relationships, therefore, cannot in any sense be regarded 

 as final. It is hoped, nevertheless, that they may prove valu- 

 able as a contribution to a discussion of the affinities of the 

 families concerned. 



