FISHES W MASSACHUSETTS. 55 



diate between the two einds is so slight that it is easily torn, 

 or even entirely worn away by use during life," — if this ex- 

 planation be correct, it is not a little singular, that the appear- 

 ance of the entire dorsal ridge should, in my specimen, have 

 been so uniform : no rays partially worn or broken in the in- 

 terval between the extremities, as we should suppose would be 

 produced by a gradual change, but an entire absence of rays, 

 from the eighteenth, which is perfect, to the posterior rays, 

 which are also all perfect. 



Yarrell observes, that the edges of the sword are " finely 

 denticulated." In that portion of the edge of the sword before 

 me which is perfect, it is entirely smooth ; the greater portion 

 of the edge is broken at intervals irregularly, unnaturally, evi- 

 dently by use. 



Trachinotus. Lacepede. 



Generic characters. Free spines on the back, and two oth- 

 ers, also free, before the anal ; body elevated, and the tail loith- 

 out the lateral carincB ; profile vertical ; and the dorsal and 

 anal fins tapering to points more or less long. 



T. argenteus. Cuv. The Rudder Fish. 

 Cuv. et Valenc. Hist. Nat. des Poiss. t viii. p. 413. 



Dr. Yale sent me from Holmes Hole, three specimens of a 

 fish, generally known at the Yineyard by the name of " rudder 

 fish,''"' which, after considerable hesitation, I have concluded to 

 be the " T. argenteus,'''' Cuv. The smallest specimen, two 

 inches in length, answers perfectly to the figure, in Mitchell's 

 "Fishes of New York," incorrectly called by him '■'• Cory- 

 phmna hippurus.''^ This small specimen was taken at a wharf 

 at Holmes Hole. The two large specimens were caught about 

 fifteen miles from land. Dr. Yale writes me, that ''this fish 

 is generally found at sea, but is common in oiu waters. It 

 follows vessels, or keeps near old casks or plank that are float- 



