PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 265 



Professor Gill, in 1873, catalogued the east coast species as Lactoplirys 

 trigonus* 



Poej^, ill liis " Synopsis Piscium Cubensiuni,"t follows the lead of 

 Bleeker, accepting his subgenera though not bracketing them into the 

 midtlle of the binomial names as was the practice of the Dutch zoolo- 

 gist. In a later work, the " Enumeratio Piscium Cubensium," he adopts 

 the genera Ostraciofi, Acanthostracion, and Lactoplirys. Jordan uses the 

 name Lactophrys quadricornis, accepting provisionally Swainson's ar- 

 rangement. 



After studying the group, as represented in the collections of the 

 National Museum, I am unable to recognize any characters sufficiently 

 persistent to serve in dividing the typical Ostracions into genera. The 

 most dissimilar forms are connected by others, intermediate in character, 

 and a series of specimens in various stages of growth of a single species 

 like 0. quadricornis or 0. concatenafus shows great age-variation both in 

 shape of carai^ace and in size and distribution of spines. 



As has been remarked, Lacepede and Bleeker regarded the position of 

 the spines as the most reliable character for classification. 



Hollard sniiis up his observations on the specimens in the Paris Mu- 

 seum as follows : 



" The diversity of species at a first glance appears greater than it is 

 in reality ; at least it is easy to reduce them to a small number of typi- 

 cal forms. The true types are those based upon form. The absence, the 

 presence, and the number of the spines, large or small, with which many 

 of the Ostracionts are provided, aftbrd characters of very secondary and 

 simply specific value. These spines in fact are present or absent with- 

 out regard to more important characters. * * * They should be 

 subordinated to other diiferences between which no known examples 

 form connecting links.":j: 



Kaup and Swainson, on the other hand, adopted the form of the cara- 

 pace as the most important character. This was considered by Hol- 

 lard as of but little value for generic diagnosis, and by Bleeker is dis- 

 posed of most summarily. " The triangular or quadrangular form of 

 the body," he remarks, " appears to have no real value (for the separa- 

 tion of genera) since it depends simply upon the greater or less con- 

 vexity or elevation of the dorsal plane of the carapace. If, for example, 

 we place an Ostracion triqueter, L., by the side of an Ostracion teiragonus, 

 L. (= 0. cubicus), we have before us two well-marked types, one with a 

 triangular the other with a quadrangular carapace [Ostracion, Kaup, 

 and Cihotion, Kaui)). But if between these two extremes we place an 

 Ostracion guineensis, Bllvr., and an Ostracion nasus, BL, we cannot decide 

 whether "■we are dealing with a triangular or quadrangular form, for the 

 dorsal surface is elevated in the shape of a roof, presenting two faces 

 which descend from a central crest to unite at an obtuse angle with the 



* Eep. U. S. Com. Fisheries, part i, p. 793. 



t Repertorio Fisico-Natnral de Ma tie Cube, ii, 1888, pp. 439-442. 



t Annales des Sciences Naturelles, vii, p. 140. 



