194 MEMOIRS OF THE CARNEGIE MUSEUM. 



of G. carapo, the second species of the Chjmnotince, as has ah-eady been intimated, 

 may be the result of its color-markings and active, predaceous life. The other 

 two subfamilies, the Sternopygince and Sternarchince, show a relatively high per- 

 centage of injured individuals. This seems to be due to two causes: (1) the 

 exposure of a large amount of tail to injury; (2) the survival of the injuries received 

 in the region of the tail. 



The longer the caudal portion exposed the greater the chance of its being 

 attacked and injured by other fishes. Size in itself, other things being equal, may 

 be quite a factor in determining the liability to injury. This statement is borne 

 out by an analysis of the injured specimens of Sternopygus macrurus. A larger 

 percentage of the large fishes have been injured than of the small ones. 



Sternopygus macrurus (Bloch and Schneider). 



All the specimens mentioned in this table were taken at the same time in one 

 catch from the Botanic Garden, Georgetown. Of course the element of time in 

 addition to that of size enters into the comparisons in this table. The larger 

 examples, being the older, have been exposed to injury for a longer period than the 

 smaller. 



The long caudal portion, which contains no viscera or vital organs, may be 

 mutilated without killing the fish. Specimens having been mutilated in this 

 region are therefore in evidence. Species of a shorter type with the viscera occupy- 

 ing relatively much more of the body, if injured, would be more liable to be fatally 

 affected, consequently fewer mutilated specimens would be found in a large col- 

 lection. The presence of so many mutilated specimens among the collections of 

 Gymnotidce does not necessarily mean that the Gymnotidce are more frequently 

 injured than other species of the same habit subjected to the same conditions, but 

 it does show the injuries to be less frequently fatal. The frequent injury to the 

 caudal portion seems due to the elongated tail. Since the Gymnotida' survive 

 these injuries because of the elongate tail and the extreme cephalad position of 

 the viscera the question arises whether they are not "protectively shaped." 



Regeneration is probably of little importance to the first two species, E. 

 eledricus and G. carapo, as compared with its value to the Sternopygince and Stern- 

 archince, for in these two subfamilies it tends to restore the protective shape. The 

 power of regeneration seems about equally developed in both species subject to 

 frequent injury and those not often injured. The same parts are regenerated by 

 both groups and with about an equal degree of completeness. 



