lc«C3.] PEOCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 245 



Tropidichtiiys Bleekcr, Nat. Tyds. Nederl., lud., iv, 1854 {valentini.) 



AxosMius Peters, Wiegia;ran's Archiv, 1855,274 (twniatwi, etc. )• 



Les Eynchotes (Rhynchotus) Bibron, Revue de Zoologie, 1855,279; {Gronovii; mar- 



garitatiis, sU-iolatiis : peronii, laterofasciatus.) 

 Rhynchotus Hollard, £tudes sur les Gymuodoutes, 1857, 320 (jyeroni.) 

 Anosmius Giiutber, Cat. Fishes, Brit. Mus., viii, 300, 1670. 

 Canthigaster Bleeker, Atlas. Ichth. Gyiun., 1867.80, etc. 

 I'SILONOTUS Gill, Proc. U. S. Nat. Mus., 1884, 421 (rostratits.) 



Type Tetrodon rostratus Bloch. 



This genus is strongly marked as to its external characters, and still 

 more peculiar as to its skeleton. Professor Gill has lately raised it to 

 the rank of a separate family under the name of Pdlonotidcc. 



The proper name of the genus has been inv^olved in some confusion, 

 owing to the use by Swainson of two different names for it. 



On page 194 of his miserable work on the classification of fishes, 

 Swainson gives an analytical key to the genera, and applies to the 

 present group the name of Canthigaster (correctly written Acanthogas- 

 ter). No species are here mentioned by Swainson, but in this case his 

 diagnosis is accurate and sufficient. On page 328, these genera are 

 again defined, the present one in nearly the same way, but under the 

 name oi P.silonotus. Two species (rostratus : eJectricus) are here men- 

 tioned as types. 



Professor Gill has preferred to adopt the last-mentioned name, re- 

 garding Canthigaster as unidentifiable except through the medium of the 

 species mentioned under the diagnosis of Fsilonotus. Dr. Bleeker has 

 preferred to take the earlier name of Canthigaster. In this case it is 

 certainly true that no doubt could exist as to what Swainson intended 

 to include under Canthigaster, even had the second diagnosis been 

 omitted; moreover, the name Psilonotus is preoccupied. We see, there- 

 fore, no sufficient reason for setting this name aside, objectionable as it 

 is. 



The species of Canthigaster are somewhat numerous and belong mostly 

 to the East Indian fauna. As a rule, they are smaller in size than most 

 of the other Tetrodonts. 



ANALYSIS OF SPECIES OF CANTHIGASTER. 



a. No black ocellus under tbe dorsal fin. 



t. Upper parts dark brownisb, Avithont white spots; candal fiu with its upper and 

 lowtjr margins abruptly black; the middle of the fin pale, immaculate; an 

 irregular brownish baud or series of dots from pectoral to upper margin of 

 caudal, continuous with the dark band on the latter, lower baud of caudal 

 continued on lower side of tail; adult specimens (always?) with chestnut- 

 colored bands on the caudal peduncle and on the chin ; dark streaks about 

 eye; abdomen with very small, two-rooted spines; body otherwise smooth ; 

 snout moderately produced, rather more than twice the concave interorbital 

 space; caudal fin slightly lunate ; dorsal rays about (5 Rostratus, 13. 



ib. Upper parts everywhere behind head covered with round whitish spots, not 

 larger than the pupil, and separated from each other only by a network of 

 the brown ground color ; fins all unmarked, pale ; spines on the belly and on 

 the front part of back; snout produced, its upper profile slightly concave. 

 Dorsal rays, 9 Puxctatissimus, 14. 



