Tar[)on Fishing in Tlorida. 55 



sucker situated on their backs just behind the head, so that tlieir 

 bellies are turned outwards. The li^ht, therefore, is not cut off 

 from this part of the hsh to the same extent that it is with vertical 

 swimmers. Hence, were the belly white the fish would defeat its 

 object and become very conspicuous. It is, therefore, about the 

 same luminosity as that part of the skin of the fish to which the 

 sucker is attached. It might be expected that the side of the 

 sucker which is undermost would be of a lighter shade than the 

 side which is uppermost. This, however, is not so, as the head 

 of the sucker always points in the same direction as that of its 

 carrier, and so this would practically confine the sucker to one side 

 of their large friends. The construction of the sucking apparatus 

 is curious and interesting. It is about one inch long by three- 

 fighths broad, and is traversed by several thin flexible bars with 

 rough edges, after the manner of a closed V^enetian blind. These 

 bars are so constructed that the sucker cannot be pulled backwards 

 along the body of his large companion. In fact the harder it is 

 pulled backwards or the greater the speed of the carrier, the 

 stronjjer becomes the suction as these bars tend to assume a 

 position more perpendicular to the backbone of the fish, and so to 

 occupy a larger space than before. Vhv. sucker, therefore, is not 

 to be got rid of by any amount of speed on the part of the fish to 

 which it has attached itself. Kone of these can suddenly jump 

 backwards, so their only chance is to throw the sucker off in the 

 air. The sucking action is automatic and rec^uires no sustained 

 effort on the part of the remora, as it still continues after its 

 death. It is almost impossible to pull a dead sucker backwards 

 along one's wet hand and its weight is quite a sufficient backward 



