184 PEDICTTLATI. 



naked, or covered with minute spines. The spinous dorsal is reduced to 

 three isolated spines, the anterior of which is modified into a tentacle, 

 situated above the snout. The soft dorsal of moderate length, longer 

 than the anal. No cleft behind the foui'th gill ; only one-half of the 

 anterior arcus branchialis is provided with lameDse. Pseudobranchiae 

 none. Stomach very wide ; pyloric appendages none. An air-bladder. 

 Seas between the tropics. 



There is scarcely another genus of fishes which offers so much 

 difiicidty in the discrimination of the species as the present. First 

 we may state the fact, that most of the species of this genus are 

 inhabitants of the seas between the tropics, Uving on floating 

 sea-weed, and enabled, by filling the spacious stomach with air, 

 to sustain themselves on the surface of the water. They are 

 therefore found in the open sea as well as near the coasts, and 

 being bad swimmers, are driven with the ciirrents into which they 

 hapiien to fall. Thus it is a natural consequence that at least 

 some of the species should have a very wide geographical range, 

 not only over the Atlantic, but also over the Indian Ocean. This 

 circumstance has been entirely disregarded by several naturalists, 

 who have readily admitted the identity of the species in specimens 

 from the He de France, New Ireland, and the Sandwich Islands, but 

 have opposed the idea of the same identity of fishes from the Brazil 

 coast, the Cape, and the He de France : they have constantly separated 

 the Atlantic from the Indian specimens ; and the great variabihty 

 to which the species of this genus are subject gave them ample 

 opportunity of affixing some characters to the fictitious species. 

 This variability is so great, that scarcely two specimens will be found 

 which are exactly alike ; it extends not only to the coloration, but 

 to the spines and cuticular appendages, which are more developed 

 in age, to the length and flexibihty of the second and third dorsal 

 spines, and to the form of the tentacle above the snout, which, being 

 tender and delicate, is necessarily often injured, and probably 

 reproduced. Whenever I had an opportunity of examining speci- 

 mens agreeing with the descriptions given, I was enabled to refer 

 them to the typical form ttf which they are merely individual va- 

 rieties, or to decide whether they shoixld really form a distinct specific 

 group ; but when I had no such opportunity, it would have been too 

 hazardous to treat them as merely synonyms, and therefore I have 

 preferred to admit them into the list of species, although I have 



3. Chironectes trisignatus, Bichards. Voy. Ercb. ^* Terr. Fishes, p. 15. pi. 9. 



Gg.l.— Hah. - — -? 



4. mentzelii, Cuv. ^ Val. xii. p. 417. — West Indies. 



5. Marcgr. p. 150. — Chironectes furcipilis, Cuv. Mim. Mus. iii. p. 329. pi. 17. 



fig. 1. — Cuv. cf- V(d. xii. p. 423. 



6. Seba, i. 74. 6. — Lophius marmoratus, Shaw, Nat. Misc. v. t. 176, and Zool. 



V. p. 386. pi. 165.— Pacific. 



7. Chironectes bicoi-nis, Lowe, Proc. Zool. Soc. 1839, p. 84, and Trans. Zool. 



Sor. iii. p. 10. — Madeira. 



8. Antcnnarius dorchensis, Blcek. Act. Soc. Sc. Indo-Nederl. vi., Nieuw- 



O'uinea, p. 21. — New Guinea. 



