THE FLYING FISH. 



W9. 



very gradual, so that the last ten or fifteen yards of 

 their course is almost parallel with the waves. It is 

 seldom that more than 200 or 300 rise up at once, so 

 that it is an exaggeration to magnify this numher to 

 thousands. Neither do they fly in all possible direc- 

 tions ; for their flight, with very httle variation, is always 

 straight forward, and they only diverge a little to the 

 right and left when they are about to re-enter their more 

 natural element. Sometimes, indeed, they fly off in an 

 obliquely angular direction from that which they at first 

 took. This is an important fact, because it proves that 

 this flying is not merely efi^ected by a leap, but by the 

 action of the fins and. tail, just as these members are 

 used to influence the course of birds : the forked struc- 

 ture of the caudal fin has an obvious connection with 

 this power, but in what manner the pectorals are used 

 we are altogether ignorant. We have no doubt that 

 more than double the number of species of Eococetus 

 really exist above those that have been described ; and 

 we have to lament the loss of three^ at least, that 

 formed part of our Brazihan collection : those of India, 

 the Mediterranean, and the Pacific, are probably all 

 different. Some of these^ forming our genus Cypsilurus, 

 are singularly characterised by the possession of simple 

 or lobed cirri or barbels proceeding from the lower jaw. 

 We here insert the cut of one of these (C. appendicu- 

 latus, jig, QS.^y described by Wood* as inhabiting the 

 American seas. 



* Jouraal of the Acad, of Nat. Sciences of Philadelphia, which for 

 brevity we cite as Amer. Trans, vol. iv. p. 283. 



