which the prefent drawing is made : here it is per- 

 fedly ftraight, two inches and fix tenths of an inch 

 long, and tapers gradually from the bafe to a fine 

 point, the upper mandible being a little longer than 

 the lower, and bent over it. Scopoli * is the only 

 modern author I know who has obferved this^ and 

 thus accurately expreffed it, Rojirmn re&umy mandi~ 

 bula fuperiore longiore apice deflexa ; Willoughby's 

 defcription, though not fo particular, implies nearly 

 fhe fame, Rojirum gracile inJeElis lancinandis aptuniy 

 probably taken from Aldrovandus, whom he refers 

 to, not having himfelf feen the bird. The rude, 

 but not ill proportioned figures of Sibbald and Gef- 

 ner f correfpond, in this particular, with the de- 

 fcriptions of Scopoli and Willoughby, as well as 

 with my fpecimen, better than any others I have 

 feen. 



May I prefume to ofi'er a conjedure how a miftake 

 in defcription might have been firft produced by a 

 flip of the pen of the great Naturalift ? He probably 

 had the two words craffus and acutus, in idea, and 

 accidently inferted the one infl:ead of the other ; for 

 I cannot conceive that the accurate Linn^us could 

 have intended that as a fpecific difference, which 

 was part of the Generic character ; whereas the de- 

 viation of rojirim apice acutius, from the Generic 

 charader, ohiiijum^ affords a mofi: ftriking fpecific 



* Annus I. Hiftorico-naturalis, 1769. 

 t Icones AmmaUum, 1560. 



diflinc- 



