v^^n^r^^^- [^r^r^r^r^ "^r^r\f^^r^ ^SHHGaOBSak , r*ftW>»^/~ , Amqf^^^^^^?^ r^'^^^^rr^fVS'S' :r>0'^'^ ^^^^^^^^^ ir>r>r>^ ^0/^^^^o^^^«r^^^^Hl^" ,<^r^'< ^^^/^r;^'r>'f>U! .Lr r^rrrrrrr^;^^^^Pf^^^ ,^^^«^'^^ ««9«»PW???TP?« "T-T-?-! -y-'-f'-r-^T^f-f M?l9ftft«i^^Sj|^?«??5222 ^'iS'@w^; ^;5,r=^^ "^,^.^,^^^ ^m^ rrrp ^^nnr SS^^^r?^A'^>^M^«M.Lr^^ ^^leiMiR^jsSS:^^ ,j^^&^m, ^^yr^^^^^^^^2 ^^^^^^^ '^^^*i^,ii^ ^???ii8s??5?Sfi ^^^^r>^r^^i^^^s r^'^'^'^'^^n'^' ^^A^^m '^r^^ ^'~>^^^/^-fc^^ ' ' 1^ 1'^ w^ ' '-^i -| 'Uk^i iiy-' y: )^tr r 377 ] Olvlslsn ii7Flshe«, , U. S, Hafionsil MuMttnt ^ ri.^yy \^\ XIV. An account of the Fishes of the States of Central America, based on collections made by Capt. J. M. Dow, F. Godjian, Esq., and O. Salvix, Esq. By Albert GtJNTHEH, M.A., M.I)., Ph.I)., F.H.S., F.Z.S. Ecad March 22iid, 186-1, and December 13th, 1SG6. [Plates LXTII. to LXXXVIL] § 1. Introductory Historical Bemarks on the Collections forming the basis of this Memoir. Before proceeding to the enumeration and description of the fislies known to exist in the States of Central America, I may be permitted briefly to notice the circumstances which enable me to submit to the Society the results contained in the present memoir. Mr. Salvin started in the year 1859 on his second excursion to Guatemala, chiefly with the intention of working out the ornithological fauna of that country. But having had his attention directed by me to the fact that its cold-blooded vertebrates were almost entirely unknown, he made and brought home a small collection of reptiles and fresh- water fishes, which proved to be of sufficient interest to encourage him to pay still, more attention to this subject on a third excursion, which he undertook in company with Mr. Godman in the year 1861. By far the greater part of the materials which form the basis of this memoir were obtained on this occasion. Not only did the two travellers extend their excursions to various parts of Guatemala, but Mr. Salvin also visited Panama, where he met and collected in company with Capt. Dow, of the Panama Railway Company's Steamer ' Guatemala.' Capt. Dow, indeed, had commenced to collect fishes previously to this, lla^•ing sent several collections to the Smithsonian Institution in Washington, and to the Zoological Society of London, whence they were transferred to the British Museum; and foi^the last three years he has continued his researches with such zeal and liberality that I cannot abstain from acknowledging here the great services he has rendered to the cause of science. The collections made by these gentlemen contained not less than about 1500 examples, ni a perfect state of preservation, many of considerable size. In addition to these, I have examined a few which had been purchased of a dealer for the British Museum having been collected at Puerto Cabello in the Bay of Honduras, and, finally, those collected by Dr. Seemann, originally deposited in the CoUection of Haslar Hospital, and now in the Britisli Museum. The latter have lost much of their scientific value, as, unfortunately, no record was kept of the localities where they were obtained ; and only in a few cases have I been able to avail myself of specimens of this collection, viz. where the original label, with the name of the collector, has been accidentally preserved. VOL. VI. — P.VRT VII. ' 3 g 378 DE. GiJXTHEE ON THE FISHES OF CENTEAL AMEEICA. ^ 2. ToiwrjrapMcnl Features of the Localities explored. As regards the topographical features of the localities explored by Messrs. Dow, Godnian, and Salvin, I have been favoured by the latter gentleman, by whom also the accompanying map has been prepared, with the following notes : — Lakes. Amatitlan.— The Lake of Amatitlan is situated in lat. 14° 29' N., long. 90" 35' W., in the Republic of Guatemala. Its elevation above the sea-level is about 4500 feet. Being only a short distance on the southern side of the main ridge, it collects the waters of a few small streams, which it discharges at its southern extremity, into the river Michatoya, a mountain-torrent for half its course, then expanding, like all the rivers of Guatemala which flow into the Pacific, into a broad shallow stream with a shifting sandy bed. The lake is very deep, and its water clear. The volcanoes of Pacaya and Agua rise amongst the mountains of its southern border, the whole forming a landscape of great beauty. Fish are caught during the rainy season near the outlet into the river Michatoya, and are sent to the market of the City of Guatemala. Atitlan. — The Lake of Atitlan is elevated 5000 feet above the sea. Like the last- mentioned it lies in Guatemala on the southern side of the main ridge, in lat. 14° 43' N., long. 91° 14' W. It has no visible outlet. The Avater is clear and fresh, and the lake of great depth. The hills on three sides attain to a height of 2000 feet above the lake. On its southern border the two large volcanoes of Tollman and S. Pedi'o rise, theii' bases being washed on one side by the lake, giving one the idea that one of them (that called Toliman) has in rising acted as a dam and stopped the outflow of the waters of a mountam-valley. A few small streams enter the lake, the water of which rises during the rainy season, to fall again in the dry. On the mountain-slope below, several streams take their rise, supplied probably by the filtration of water from the lake ; but it would appear, from the alteration of the water-level in accordance with the season of the year, that it is chiefly influenced by evaporation. A number of Indian villages surround the lake ; at one of them, Panajachel, a small collection of fish was made. Fish never seem to grow to any size in this lake, the Mojara (Ileros) being quite diminutive. The Indians fish with round nets amongst the reeds that grow at the mouths of small streams. The lake itself is about twenty-two miles long, and twelve miles wide. DuENAS. — This lake is little more than a depression in one of the elevated (5000 ft.) plams forming the tablelands of Guatemala. Its depth is nowhere more than 6 feet, and its banks are everywhere clothed with reeds. A small stream connects the lake with the river Guacalate. Here, too, fish are caught by the Indians in romid nets, which are held by both hands, pushed in amongst the reeds, and suddenly brought to the surface. HuAMUCHAL. — This name applies properly to a series of small lakes situated in about lat. 14° 32' N.,long. 92° 13' W., close to one another, about six miles from the mouth of the river Tilapa on the Pacific coast. The place is not shoA\Ti on any map ; but it is near DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 379 the large Lake of Taniachian, with which, in the rainy season, all these smaller lakes are connected. During this period of the year the river Tilapa overfows its banks and inundates tlie whole country round. In the dry season water remains in depressions of the land, forming the lagoons of Huamuchal ; but in years of great drought even these dry up, the fish being destroyed ; but a fresh supply finds its way from Lake Tamachian during the next inundation. The water is slightly brackish. The fish are taken in drag-nets, salted, and sold to Indians coming from the Altos of Guatemala. Maxagua.— According to Mr. J. Bailey this lake is about fifty or sixty miles long, by tliirty-fi\e miles wide. Its depth varies from 2 to 10 and 15 fathoms, but in its deepest part reaches to as much as 40 fathoms. Its elevation above the sea is 156 feet. On its south-western border the lake is separated from the Pacific by a series of comparatively- low hills, the lowest section of which, through the Plain of Leon, is only 230 feet above the ocean-level. The high mountains of the Republic of Honduras approach the nortli- eastem border of the lake. On its south-eastern side an opening communicates with the Lake of Nicaragua. Commencing with the Fall of Tipitapa, of 22 feet heiglit, the river widens into the Estero of Panaloya, and thence into the larger lake. Nicaragua. — The same authority gives a length of one hundred and five miles to this lake, and a width of about forty-five, its depth being about 15 fathoms. The surface of the lake is studded with numerous islands, some of them, as Omotepec, being volcanic cones. The elevation of the lake above the mean ocean-level is given as 128 feet. The same line of low hills which divides Lake Managua from the Pacific separates Lake Nicaragua from the same ocean ; but at no point is the elevation so low as at that above indicated. The river San Juan, a deep stream with several rapids, flows out of the south-eastern end of the lake, and falls into the Atla:,;'- OrPnn. nf the port of Gn-v- town, or San Juan del Norte. Petex. — The Lake of Peten is situated in lat. 17° W ^., long. l'0° ^^ ., and is one wi several lakes formed at the base of the Promontory of Yucatan. Its length is about thirty miles, its width eight miles, and elevation above the sea 500 feet. The water is quite fresh, clear, and of considerable depth. Neither the Lake of Peten nor the adjoining Lake of Yasha has any outlet ; and in both the water is rapidly mcreasing in expanse — so much so that several streets of the toAvn of Flores, which stands on an island in Lake Peten, have been absorbed within a few years, and the posts of huts, Avhich formerly were on dry ground, may now be seen standing in deep water. This increase of water can onlv be accounted for by supposing that a common subterranean outlet has been stopjied up, or that the land of this district is experiencing a gradual subsidence. All the fish obtained here were caught with a hook and line, or speared. All the natives, even quite small cliildren, are very expert in using a light spear formed of bamboo cane with an iron barb at tlie end. YzABAL.— This lake, which is also called the Golfo Dulce, is about thirty or forty miles long, and ten to fifteen miles wide, and has a tolerably uniform depth of about 35 3 G 2 380 DE. GtJNTHER ON THE FISHES OF CENTEAL AMEEICA. to 40 feet. It is situated in lat. 15° 30' N., long. 89° 15' W., at the bottom of the Bay of Honduras. One large river, the Polochic, enters this lake ; and it has a narrow but deep outlet to the sea, called the Eio Dulce, which is navigated by small schooners plying between Belize and the town of Yzabal. It was near this last-mentioned place that a few species of fish were obtained. Bivers. BayAx\o. — lliis is a river which rises in the narrow part of Central America, and flows into the Pacific a little to the southward of the Bay of Panama. Cahabon. — The town of Cahabon, where a few fishes were obtained, is situated on an affluent of the river which bears this name. The main stream rises in the same marsh as the Polochic, but takes another valley, in Vera Paz, and again joins the Polochic, when they both flow into the Lake of Yzabal, and thence into the Atlantic. Chagres. — This is the principal river of the Isthmus of Panama. It flows into the Atlantic. The fish were obtained near the railway bridge at Barbacoas, about halfway across the isthmus. Chisoy. — Of the numerous names this river bears, I have chosen this for the principal stream which forms the large river that flows out into the Laguna de los Terminos, in the Bay of Campeachy. This branch is also known as the Rio Negro ; and after receiving the water of the Rio de la Pasion, or Rio de Santa Isabel, as it is also called, the two are usually called the Usumacinta. Fishes were collected from this river near the Indian village of Cubulco ; and a number were also procured by poisoning with herbs a small stream near Saouchil, an Indian village below the town of Cohan, in Vera Paz. GuACALATE. — Is One of the numerous rivets which drain the southern watershed of the main ridge into the Pacific. It flows past Antigua, the old capital of Guatemala. Fishes were obtained about 3500 feet aboA^e the sea, where the river is still quite a torrent. MoTAGUA. — This river, the second largest in Guatemala, rises in the main ridge, and flows, with high mountains on either side, nearly due eastward into the Atlantic. Fishes abound in this river ; and nearly every year a considerable length is poisoned, and a large quantity obtamed. On one of these occasions a collection was made a little below the bridge over which the highroad from Guatemala to Vera Paz passes. Another collection came from lower down the stream, below the village of Tocoy. San Geronimo. — Is a tributary of the Chisoy before mentioned. A small collection was made near the village of San Geronimo, in a plain at the foot of the mountains whence it takes its rise. Santa Isabel. — A small stream flowing into this river, one of the principal branches of the Usumacinta, was poisoned, and a number of small fishes obtained. San Salvador. — A few small fishes were caught by Capt. Dow in a warm stream near the capital town of this republic. DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 381 Marine localities. Belize. — All fishes from Belize were from the market, and were caught amongst the coral reefs which line this coast. Cardox' Island. — Is situated at the mouth of the fine harbour of Realejo, in Nica- ragua. Fishes were found at low tide in the pools amongst the rocks, and caught with a landing-net. CliiAPAM. — The whole coast of Guatemala, bordering the Pacific Ocean, is studded with a number of lagoons formed at the mouths of the numerous rivers which How down from the neighbouring mountains. All these rivers are charged with volcanic sand, which is thrown back by the heavy surf that rolls in on this coast. The body of water brought down during the dry season is often insufficient to reduce this sandbar; and it frequently happens that all outlet to the sea is stopped. The accumulation of water during the rainy season breaks this barrier ; but it again forms when the water subsides. About the period of the cessation of the rains the natives cut an artificial channel, which, at first widening of itself, often remains open some months, each tide bringing a great quantity of fishes into the lagoon, whicli are there netted by drag-nets. The water is almost salt, but varies in this respect according to the size of the river which enters it. A few fishes were also obtained by a hook and line from a canoe in the open sea. LiBERTAD. — This is an open roadstead, the port of the City of San Salvador. A\'hilst we were lying at anchor here a few fishes were caught with a hook and line. Panama. — Most of the fishes taken in the Bay of Panama were found in the pools amongst the rocks at low tide. A reef running out from the town was an excellent locality ; one spring tide Capt. Dow and I secured twenty-four species in the course of half an hour. San Jose. — Is the port of Guatemala on the Pacific side ; a few fishes were caught here in the open sea in a canoe. § 3. Definition of the Boundaries of the Fauna treated of in this Mtrnoir. Although we may presume that our account contains a tOier; hly complete list of the species inhabiting the localities visited, particularly as on several occasions poison (the best means for securing a complete series of the fishes of a certain locality) was resorted to, yet there is still a wide field for future explorers in a country where several forms (such as Heros, Pimelodus, and the Cijprinodontes) are so much developed and specialized. Of the fishes of Yucatan we still know absolutely nothing. The list of the marine fishes of the Atlantic coast will, without doubt, be considerably swelled, as the gentlemen mentioned paid much less attention to the Atlantic marine fauna (which would have yielded comparatively few novelties) than to the freshwater fiiuna. And knowing how little advantage is derived from, and how much confusion is caused by, receivuig into a ' Tliis name is misspelt " Cardova" in several places in the 3rd volume of the ' Catalogue of Fishes.' — A. G. 382 DE. GUNTHEE 0\ THE FISHES OF CENTEAL AMEEICA. fauna species which may be expected to belong to it, although they are not yet disco- vered within its limits, I have excluded all species not actually known from Guatemala, although they have been obtained north and south of it. A collection made by Mr. Godman at Belize was of great value in determining this part of the fauna. Numerous species of fishes have been described from Mexico' ; and if we were better acfiuainted wdth their geographical distribution, it would have been useful to treat at least of the southern portion of them, in conjunction with the Guatemalan species. Unfortunately but a small proportion of tlie exact localities are known, so that at present no line can be drawn to indicate where the preponderance of nearctic types over tropical ones terminates. Thus, confining myself to the fishes occurring between the political boundary of Guatemala in the nortli and the Isthmus of Darien in the south, I would repeat that, previously to the receipt of the collections forming the basis to this Memoir, only a small number had been described, as will be seen from the following remarks : — (^S 4. Historical account cjf Publications previous to this Memoir. It would be of but little advantage to enumerate the few isolated species incidentally described in general works or memoirs as occurring in Guatemala or Panama. How- ever, I must mention that the first traveller who collected fishes in these states appears to have been Baron von Friedrichsthal. I am not aware that any account of his travels has been published ; but in a paper published by the late Jacob Heckel in 'Annalen des Wiener Museums,' vol. ii. 1840, a single species is described, which is stated to be from Friedrichsthal's Central-American Collection, and which I have recognized as belonging to the Lake-Peten fauna {Ileros friedrichsthalii). The greater part of the collection made by this gentleman evidently remained unpublished until 1864, when Dr. F. Steixdachner determined from it four other species (Denkschr. ^Akad. Wiss. Wien, xxiii.), viz. : — Hcros urophthcdmus (Gthr.), Heros triagrawimi--H. salmm{Gt\iv.), Heros melanopogon, and Fctenia splendida (Gthr.). As we have received four of these species from Lake Peten, it is very probable that Baion Friedrichsthal visited and collected in that locality. In the second place I have to mentioirD*', Sekmann, who, as naturalist attached to the expedition of the ' Herald,' brought to England a collection of Central-American fishes. These, as I have mentioned above, were originally deposited in the collection of Haslar Hospital, but no record as regards the origin of the specimens was kept, so that most of them are lost for the purposes of this Memoir. In the year 18G1 I received the first collections from Mr. Salvin and Capt. Dow. Tlie species belonging to the families treated of in the 3rd volume of the ' Catalogue of Fishes ' were described therein ; and a separate account of those sent by the latter ' Prof. Troschel enumerates some 130 frcsliwuter and marine species in Miiller's ' Eeisen in den Yereinigten Staaten," &c. DE. aUNTHEE ON THE FISHES OF CENTEAL AMEEICA. 38:3 gentleman from the Pacific Coast of Central America was published in the Society's ' Proceedings ' for 18G1 (Nov. 26) ; it contained fourteen species, ten of which were new. In the following year the 4th volume of the ' Catalogue of Fishes ' was publislied, containing the descriptions of those species of Pharyngognaths and Anacanthines whitli had arrived from our travellers, who were then engaged in collecting. In the year 1863 Mr. Gill published a descriptive enumeration of a collection of " Fishes from the western coast of Central America, presented to the Smithsonian Institution by Capt. J. M. Dow." He distinguished in it the following twenty-tive species, of which I consider eighteen to have been new to science (Proc. Ac. Nat. Sc. Philad. 1863, p. 162):— 1. I)ia2)teri(s dowii, sp. n.= Ger>rs dovii. 2. Pomacanthodes zonipectus. Gill. 3. Centro])Oiniis armatus, sp. n. 4. Epinephelus analogus, STp.n.=Serrani(s analog us. 5. Promicropterus decoratiis, sY).n.=Phi/2)t9Cus decorattis. 6. Bairdiella armata, s\).n. = Corvina armata. 7. Opkioscion typicus, s]).n. = Corvina ophiosdon. 8. Amblyscion argenteus, sp. n. 9. Caratwc imnamensis, G\]l,=-Carana- speciosus (Forsk.). 10. Caravgoides dorsalis, sp. n. 11. Carangus marginaftcs, Gi[\,=: Caranw hippos, L., var. 12. Oligoplites inornatus, sp. n. = Chorinemus inornafus. 13. Kvoccefus dowii, sp. n. 14. aJUdactylus, sp. n. '? = ^. bahiensis (Kanz.). 15. Upeneus grandisquamis, sp. n. 16. Trichidion opercularis, sp. n.=iPolynemus opercularis. 17. approocimans=Polynemus approximans (Lay & Benn.). 18. Mugil guentlierii, Gill, = il/. brasiliensis (Agass.). 19. Batrachoides pacijici=. Batraclms pacifici ((jthr.). 20. Dormitator microphthahms. Gill, = Eleotris macidata (Bl.). 21. Leptarius dowii, sp. n.=^n»s dovii. 22. Sciades troschelii, sp. \\.^=Arms troschelii. 23. jElurichthys panamensis, sp. n. _ - 24. Atractosteus tropicus, s^.n.-=Lepidosteus tropicus. 25. JJrotrygon mundus, sp. n. At later periods Mr. Gill has described some other species incidentally, which will be referred to in the gen(>ral list. A small collection made by Prof. M. Wagner on the Isthmus of Panama, between T and 9' lat. N., and 77^" and 83" long. W., was examined by Messrs. Kxkr & Stkix- 384 DE. GtNTHEE OX THE FISHES OF CENTEAL AMEEICA. DACiLVEK, who gave a preliminary account of it in ' Sitzgsber. bayer. Akad. Wiss.' 1863, pp -220-230, and more detailed descriptions in 'Abhandl. bayer. Akad. Wiss.' 1864(1865), „,, l-Ol. Prof. M. Wagner added, besides, a detailed account of the hydi-ographical peculiarities of this part of Central America (pp. 65-92). The species treated of in these Memoirs are the following : — 1. Pristipoma humile, sp. n. 2. Dajaus elongatus (K. & St.)=Ag(mostoma msuUim (Gthr.). 3. Dajaus monticola (C. & V.). 4. Acara ccerideopunctata, sp. n. 5. Heros altifrons, sp. n. 6. Heros sieboldii, sp. n. 7. Meotris pictus, sp. n. 8. Engraulis macrolepidotus, sp. n. 9. poey?-, sp. n. 10. Xijjhophorus gillii, K. & St., = P(edlia, sp. ? 11. Macrodon brasiliensis^K. & St.,=l/. microlepis (Gthr.). 12. Saccodon wagneri, STp. n. — , l^c 13. Pseudochalceus lineatus, sp. n. 14. Chalcino])sis striatulus, sp. n. 15. chagrensis, sp. n. 16. Chalceus atrocaudatus, sp. n. 17. Tetragonopterus ceneiis (Gthr.). 18. gronovii (C. & V. "?). 19. Bagrus {l)arioides, sp. n.=Anus multiradiatns (Gthr.). 20. Pimeludus modestus (Gthr.) 21. dnerascens (K. & St.)=P. wagneri (Gthr.). 22. Loricaria uracantha, sp. n. 23. lima (Kner). 24. Hypostomus plecostomus (K. & St.) =Plecostomus, sp. 25. Ancistrus drrhosus (C. & V.). 26. Acanthias vulgaris (Rissol). Finally, having received in 1864 the last collections made by Messrs. Godman & Salvin, 1 gave preliminary notices of the new species in the ' Proceedings ' of this Society, em- bodying the numerous contributions to our knowledge of the Siluroids and Characinoids in the fifth volume of the ' Catalogue of Fishes,' to which were added the Cyprinodonte.s and Scombresocides in the sixth (1865-66). § 5. General List of Central-American Fishes. After these introductory remarks on the contributions to the ichthyology of Central DR. GtNTHER OX THE FISHES OF CENTEAL AMERICA. 385 America preceding this Memoir, I at once proceed to give a list of all the species known to exist in these countries. There are comparatively few which I do not know from autopsy; their names are printed in italics. An asterisk (*) marks those which are described or remarked upon. The second column contains chiefly the names of the localities where they have been found within the limits of Central America ihe localities of species occurring on both sides of the Isthmus are prmted m italics; ot these I shall treat again subsequently. Finally, the letter M signifies that a species is marine, B that it is known fi-om brackish, and F that it is from fresh water. ACANTHOPTERYGII. Fam. PERCID^. Centropomus, Cuv. t? s at 1. *appendiculatus, Poey . • • Chagres R. (Cuba, Mex., Suriu.) ^Sc^L 2. *mediiis, Gthr Chiapam 3. *nigi-escens, Gthr Chiapam 4. *parallelus, Poey Chagres R. (W. Indies, Bahia) .... 5. *armatus, Gill Chiapam 6. *ensiferus, Poey Belize (Cuba, Jamaica, Guyanas) . . . Centropristis, Bris. de Barnev. 7. *maeropoma, Gthr Panama Serran'us, Cuv. 8. *creolus, C. 6c F. Atl.S^-Pac 9. striatus, Bl. ...:.■ Atlant 10. coronatus, C. ^ F Atlant 11. undulosus, C.^F. .... Atlant 12. *sellicauda, Gill Pacific Coast 13. *analogus, Gill Pacific Coast Plectropom-v, Cuv. U. *afrum, Bl ^t'- ^- ^«c Rhyptictjs, Cuv. 15. *decoratus, GJ// Pacific Coast ^ Mesoprion, Cuv. 16. chrysurus, Bl Atlant 17. griseus, C. i,- F. Ail. ^f Pac 18. uninotatus, C. i^ V. . . • • Atl. 8f Pac 19. *aratus, Gthr Chiapam, Panama 20. vivanus, C. i^' V. Atlant Apogon, Lacep. 21. *dovii, Gthr Pacific Coast . . B. . F. & M. . . B. . . B. . . M. . . M. . . M. . . M. . . M. . . M. . . M.? . . M. . M. . M. . M. M. & B. . M. M. VOL. VI. PART VII. 3h 386 DE. GlJNTHEE ON THE FISHES OF CENTRAL AMEEICA. Fam. PRISTIPOMATID.^. Pristipoma, Cuv. 22. *melauopterum, C. ^ V. . . Ail. ^ Pac M. 23. *vu-ginicum, L Atl. ^ Pac M. 24. *dovii, Gtkr Panama M. 25. *chalceum, Gthr Panama M. 26. *hmiiile, Kner & Steiud. . Rio Bayano F. 27. *macracantlmm, Gthr. . . Chiapam B. 28. crocro, C. Sf V. , Rio Motagua (Trop. Amer., Atlant.) . . . F., B., & M. 29. *leuciscus, Gthr San Jose, Chiapam, Panama B. CoNODON, C. ^ V. 30. *pacifici, Gthr Chiapam B. HiEMULON, Cuv. 31. chromis, Brouss Atlant M. 32. canna, C. If V. Atlant M. 33. xanthopterum, C. if V. . . . Atlant M. 34. *brevu'ostrum, Gthr. . . . Panama, Puerto Cabello M. 35. *margaritiferum, Gthr. . . Panama M. LOBOTES, Cuv. 36. auctorum, Gthr Atlant. {India) B. Fam. SQUAMIPINNES. Ch^todon, Cuv. 37. striatus, L Atlant M. 38. capistratus, L Atlant M. 39. *humeralis, Gthr Panama (Sandwich Isl.) M. POMACANTHUS, LuCCp. 40. paru, Gthr Atlant. (Colon) M. 41. *zonipectus,G'\\\ Pac M. Ephippus, Cuv. 42. faber, Bl Atlant. (Belize) M. Fam. MULLID.E. Upeneus, C. Sf V. 43. *tetraspilus, Gthr Panama M. 44. *ffrandlsquainLi, Gill . . . Panama M. Fam. SPARID.F;. Sargus, Cuv. 45. unimaculatus, Bl Atlant. (Belize) M. 46. ai-ies, C. If V. Atlant. (Belize) M. Chrysophrys, Cuv. 47. *calamus, C. if V. All. if Pac. (Panama) M. PiMELEPTERUS, CuV. 48. boscii, Lacep Atl. if Pac. (Chiapam & Panama) M. DE. G-tJNTHER ON THE FISHES OF CENTRAL AMERICA. 387 Fam. CIRRHITID^. CiRRHITICHTHYS, Blkr. 49. *rivulatus, Val Galapagos Islands, Panama M. Fam. SCORPiENID^. ScoRP^NA, Art. 50. plumieri, Bl., Schn Atl. ^ Pac. (Panama) M. Fam. POLYNEMID^. POLYNEMUS, L. 51. *melanopoma, Gthr San Jose M. 52. *approximans, Lay ^ Benn. Pacif., Chiapam, Panama M. 53. *oj)ercularis, Gill Pacif. M. Fam. SCI^NID.^. Larimus, C. ^- V. 54. *breviceps, C. Sf V, . . . . Atl. ^ Pac. (Panama) M. Micropogon, Cuv. 6f Val. 55. undulatus, L Atlant M. & F. 56. *altipiunis, Gthr Chiapam, San Jose, Panama M. & B. Umbrina, Cuv. 57. *elongata, Gthr Chiapam JJ. 58. *nasus, Gthr Panama M. 59. *analis, Gthr Panama M. CORVINA, Cuv. 60. ronchus, C. i^ V. Atlant M. 61. *chry8oleuca, Gthr Panama M. 62. *vermicularis, Gthr Panama M. 63. *armata, Gill Pacif. 64. *ophioscion, Gthr Panama M. Otolith us, Cuv. 65. *squamipinnis, Gthr. . . Panama M. 66. *albus, Gthr Chiapam B. 67. *reticulatus, Gthr San Jose, Chiapam M. & B. Fam. ACRONURID^. ACANTHURUS, Schfl. 68. chirurgus, Bl Atlant M. Fam. CARANGID^. Caranx, Gthr. 69. cnunenophthalmus, Bl. . . . Atl. §• Pac M. 70. amblyrhjTichus, C.i^V. . . . Atlant M. 71. *leucnms, Gthr Panama M. 72. *speciosus, Forsk From Panama to East Africa M. 3 11 2 388 DE. GtNTHEE ON THE FISHES OF CENTEAL AIMEEICA. 73. carangus, £/ Atlant. i^Ind. Occ. (Chiapam & Belize) M. 74. *hippos, L Tropics generally M. 75. *caballus, Gthr Panama M. 76. *caninus, Gthr Panama M. 77. *dorsalis, GiY/ San Diego (Cal.), Panama M. Argyriosus, Lac. 78. vomer, L All. 8f Pac. (Belize, Chiapam, Panama) . . . . B. & M. 79. setipinnis, Mitch Atl. ^ Pac. (Panama) M. Chorinemus, C. ^ V. 80. occidentalis, L Atlant M. 81. saliens, jB/ Atlant. ^ Pac. (Chiapam, Isabel) B. & F. 82. *altus, Gthr Panama M. 83. *inornatus, Gill Panama M. Trachynotus, C. ^ V. 84. ovatus, L Atl., Pac, &^ Ind. Oc. (Panama) M. 85. *fasciatus, Gill . . . .' . Panama, San Jose M. Fam. SCOMBRID.E. Pelamys, C. ^ V. 86. *sarda, Bl Atl. §• Pac M. Cybium, Cuv. 87. maculatum, Mitch Atlant. (Belize) M. ECHENEIS, Art. 88. remora, L Atl., Pac, ^- Ind. Oc M. 89. naucrates, L Atl, Pac, S^ Ind. Oc M. Fam. BATRACHID^. B.\TRACHus, Gthr. 90. *pacifici, Gthr Panama, West Coast of Africa M. 91. surinamensis, Bl. Schn. . . Atl. ^ Pacif. (Panama) M. *Thalassophryne, Gthr. 92. *maculosa, Gthr Puerto Cabello M. 93. ^reticulata, Gthr Panama M. Porichthys, Girard. 94. porosissimus, C.^V. ... Atl. o^o«, Stcindachner . ? Lake Peten F. 160. *melanurus, Gthr Lake Peten F. 161. *macracanthus, G PHYSOSTOMI. Fam. SILURID.E. Amiurus, Rafin. 201. *meridioualis, Gthr Rio Usumacinta !''• PlMELODUS, Gthr. 202. modestus, Gthr Rio Chagres, Esmeraklas !*'• 203. guatemalensis, Gthr Huamuclial !''• 204. godmanni, Gthr Lower Vera Paz, Rio Motagua, Mexico .... !'"■ 205. *wagneri, Griln Pacific i^ Atlantic rivers of Panania F. 206. *managiiensis, Gthr Lake of Managua F. 207. microptcrus, Gthr Rio San Geronirao F- 208. nicaraguensis, Gthr Lake of Nicaragua F. 209. petenensis, Gthr Lake Peten F. 210. motagnensis, Gthr Rio Motagua F. 211. salvini, Gthr Rio San Creronimo F. 212. polycaulus, Gthr Rio San Geronimo I''. Arius, Gthr. 213. guatemalensis, Gthr Chiapam, Panama H. 214. *assimilis, Gthr Lake of Yzabal F'. 215. platyjoogon, Gthr San Jose F. 216. seemanni, G//(r V 217. cserulesceus, Gthr Huamuclial F. 218. troschelii, Gill Pac 219. *dovii, Gill Pac 220. melanopus, Gthr Rio Motagua F\ 221. multiradiatus, Gtlu- Rio Bayano F. tElurichthys, Baird ^ Gir. 222. *nuchalis, Gthr Panama 223. *panamensis, GUI Panama PlecostomuSj Gthr. 224. *?«/?., Kner & Steindachncr . Rio Chagres F. Ch^tostomus, Heck. 225. *aspidolepis, Gthr Yeragua !'. 226. *'^cin-hosus,\&\ Rio Chagres F. LoRiCARi.\, Larep. 227. *Mr««/w//; Henle. 303. *latirostris, A. Bum Gaboon, Panama ^I- § 6. Partial Idmtity of the Fish-faunas of the Atlantic and Pacific Coasts of Central America. It will be seen that, as far as our present knowledge reaches, of these 303 species, 173 are truly marine forms, 57 being found on both sides of the Isthmus. 25 have been found in brackish water, of which 3 are found on both sides of the Isthmus. 101 are freshwater fishes, 17 being found in rivers of the Atlantic and Pacific sides. There will be but very few species which are entirely limited to brackish water, and which may not be with equal propriety added either to the marine or freshwater fauna. Thus, five of the 25 species hitherto known from lagoons with brackish water belong to freshwater genera ; and, admitting two groups only, we have 193 marine fish, 59 of which are found on both sides of Central America=30i per cent. 106 freshwater fish, 19 being found in rivers of the Atlantic and Pacific sides = 18 per cent. From the circumstance that our collectors paid more attention to the freshwater than to the marine fauna (at least of the Atlantic coast), we may assume that the pro- portion between the two groups will be increased by future researches in favour of the marine fauna, but that the proportion between species peculiar to one side and those common to both will be lessened, inasmuch as every collector will discover other Atlantic forms on the Pacific side, and vice versd. The very curious fact of the partial identity of the species of both coasts of Central America was first distinctly stated by myself in the Society's ' Proceedings ' for 1861 (p. 370), when, out of fourteen species collected by Capt. Dow on the Pacific side, five were found to be Atlantic forms. To these vai-ious others were added by me in the ' Catalogue of Fishes ; ' and Mr. Gill confirmed this observation in Proc. Ac. Nat. Sc. Philad. 1862, pp. 140, 249. Professor Wagner, in his memoir quoted above (p. 384)', has made the same observation ; but the species enumerated by him, fourteen in number, are, with one exception, freshwater forms, the geographical distribution of which must have been brought about at periods and in ways different from those of the diffusion of marine species. Knowing now that at least 30 per cent, of the marine fish are found on both sides of ' See also ' llecord, Zool. Literal." ii. p. 177. 398 DE. GtJNTHEE ON THE FISHES OF CENTEAL AMEEICA. Central America', we cannot account for this fact by resorting to such occasional means of dispersal as the accidental transmission of spawn from one shore to the other by birds or water-spouts, or even the close proximity of the sources of rivers flowing in opposite directions. If we do not adopt the view that species were created at the spot where Ave And them now, similar creations being produced under similar physical conditions, we have but one way of explaining the partial similarity of these marine fish-faunas, namely, by assuming that the Isthmus did not form a continuous barrier between the two oceans at a former period, but that one or more open channels existed. I am not aware that geology has, up to this time, furnished us with proof positive that this is really the fact ; but considering the volcanic nature of Central America, and the absence of all fossiliferous strata, it does not appear too bold an hypothesis to assume that North and South America were formerly connected by a chain of islands similar to that of the Antilles, and that subsequently an elevation (as in other parts of the globe) took place, resulting in the final continuity of dry land: the long-continued activity of the numerous volcanoes may have been another, though secondary cause in filling up the cliannels on the Pacific side. If such a bodily elevation of Central America has taken place, it is easy to show where some of the broadest channels existed, namely, where we find the greatest depressions running from one ocean to the other. The northern- most of these depressions exists between Tehuantepec and the river Coatzalco ; the second is indicated between Puerto Cabello and the Gulf of Fonseca ; the third by the T,uke of Nicaragua (the remnant and deepest part of a very broad channel) ; a fourth between Chagres and Panama. (See map, PL LXIII., where these supposed former depressions are coloured green.) As far as I have been able to ascertain, the greatest elevation of the first of these lines of depression would be 1500, of the fourth 287 feet only^ If we presume that only one of the channels was open at a period when the present marine fauna was already in existence, it will fully explain the existence of identical species on both sides of the isthmus, especially if the difierence of the tides was as great as it is now^ causing strong currents from one ocean to the other. Such an instance of a disconnexion of a marine fauna by elevation of land as I am inclined to assume in the case of Central America does not stand quite alone. We owe to the researches of Prof S. Loven and Dr. Malmgren* the knowledge of the fact that marine animals (Crustacea, Annelids, and Fishes) inhabiting the glacial ocean are found in the great freshwater lakes of Sweden and in the Bothnian Gulf, and that this is to be explained only by the former continuity of the Baltic with the Glacial Ocean. During the second half of the glacial period the greater part of Finland and of the ' Mr. Darwin (' Origin of Species,' 3rd edit. p. 378) was not acquainted with this fact, which by no means mili- tates against his argument, but merely modifies it. " M. Wagner, I. c. p. ST. ■'' At Chagres the mean elevation is 1-16 foot, while at Panama the highest flow is 22 feet. (Seemann, Voj-. of H.M.S. • Herald,' i. p. 236.) * Loven, Skand. Natui-forsk.-Siillskap. forst. oftentl. miite d. 9 Juli 1863 : Stockholm, 1864. Mabngren, ■ Kritisk Ofvorsigt af Finlands Fiskfauna,' see ' Zool. Eecord,' i. pp. 130-138. DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 399 middle of Sweden was submerged, and the Baltic was a great gulf of the Glacial Ocean, and not connected with the German Ocean. By the gradual elevation of the Scandi- navian continent, the Baltic became disconnected from the Glacial Ocean, and the great lakes separated from the Baltic. The Isthmus of Suez appears to have been a much more permanent barrier between the faunas of the Mediterranean and the Red Sea. R. A. Philippi has di-awn up a list of species of shells common to both faunas ; but it was founded on a collection made by Ehrenberg, in which the shells from both seas had been mixed' ; and P. Fischer- has lately shown that the two faunas are quite distinct. As regards the fishes, I have men- tioned (on former occasions) a few occurring in both seas (Sargus noct, Sargus rondeUtii) ; but the number is so small that one might be tempted to account for it by the tempo- rary existence of an artificial communication between the two seas. Looking at the results of the separation of the Baltic from the Glacial Ocean on the one hand, and of that of the Pacific from the Atlantic on the other, we find them very different. As soon as the continuity of the Baltic with the Glacial Ocean was inter- rupted, the amount of fresh water earned into the former by rivers exceeded the quantity lost by evaporation of its surface, and the salt water gradually changed into brackish, and in the northern parts into fresh water. By far the greater part of the animals became extinct ; but a few survived^ however, in spite of the greatly altered physical con- ditions, loithout altenng their specific characters, still agreeing loith the typical fm-ms in every point, except in size, remaining smaller, leaner, almost starved. The same thing might happen if by a rising of the chain of the West- Indian islands the Gulf of Mexico or the Caribbean Sea were at a future time converted into inland seas with narrow out- lets into the open ocean. The separation of the Atlantic and Pacific Oceans was, of course, not accompanied by a change of the water ; and any diff"erence that existed in the physical conditions of both seas, as, for instance, the formation of corals on the Atlantic side, and their total absence on the Pacific, existed already before the communication between the oceans was closed ; so that the life of species was not in any way affected by the discontinuance of this communication. Let us for argument's sake assume that the part of the isthmus between the Lake of Nicaragua and Panama was once an island, apeupres of the form of Cuba, inhabited, like Cuba, on its northern and southern coasts by a certain species of fish. The only effect of a gradual rise of the land on the life of tliis species would be to force it to retreat further and further from the original coast, and to accommodate itself to the new one — an eff"ect to which, if felt at all, the indi\aduals on the northern and southern coasts would be equally exposed. Thus there is in this case no apparent external cause for an alteration of the species ; and, indeed, the specinwns examined hy me from opposite coasts of the isthmus are absolutely identical, and there is not the slightest indication that one of them has heen modified or degenerated into a climatic or local variety. I trust that ' Martens, in ' Zoolog. Record,' ii. p. 237. " Journ. C'unchyl. xiii. 1S65, pp. 241-248. ^ Seven or eight species of the northern part of the Baltic are believed to be of Arctic origin. 400 DE. GtJNTHEE 0^ THE FISHES OF CENTEAL AMEEICA. geology will furnish us with the proof of the former partial submergence of a part of Central America, as it has done with respect to the northern part of Scandinavia. We should then be able to speak with more confidence of the permanence, or rather endu- rance, of the characters of a specific type, and arrive at a somewhat more definite idea of the age of species which must have existed before those geological changes were completed '. Sir Charles Lyell has directed my attention to collateral evidence from other classes of the animal kingdom, by which the partial identity of the faunas of the two coasts is shown, although not in an equally conclusive manner. The majority of malacologists appear to have presumed a j^iiori their distinctness, and consequently described Pacific shells generally as distinct from Atlantic species. However, Dr. Morch, in a paper in which he describes or enumerates about 360 Panama species, makes the following remarks (Pfeifi". Malakozool. Blatt. 1859, p. 107):— " The tropical [molluscan] faunse may be classed in two principal divisions, the Indian and the Atlantic. To the latter belong, 1, the Guinean (Senegalian) ; 2, the Antillian ; and 3, the Panaman, which, although belonging to the Pacific, appears to be most analogous to the Guinean. A great number of species, especially of Bivalves, have been regarded as identical with those from the eastern (Brazilian) shore. I believe I can prove that they are difi'erent. Certain irregular mollusks cannot be separated diagnostically ; but I can recognize them by their general habit. It is at all events a fact that no species stamped with definite characters (wohlausgepragt) is identical on both sides of the isthmus. The Panama species may be divided into: — 1, those analogous to West-Indian ; 2, those analogous to species from Guinea and Senegal ; ;-), those very remotely analogous to East-Indian species." ' I may on this occasion recur to a remark made by me in Proc. Zool. Soc. 1858, p. 381, witli regard to the sea-snakes obsei-ved in the Bay of Panama by M. Salle, Capt. Dow, and Mr. Salvin. There is now not the least doubt that the snakes seen were Pehmi/s bicolor, and that they are, moreover, very common there. I find' that Dr. Seemaun (Voy. ' Herald,' i. p. 205) abeady mentions them. But I am much inclined to think that this most common Indian species has migrated eastwards, and that its arrival on the West- American coast is of veiy recent date. Dampier and the other bucaniers who have left us records of their adventures, and who passed weeks and months in the Bay of Panama, could not have failed to observe them, and to mention them in their notes, just as they did on other occasions. It is also probable that these snakes would have spread into the Atlantic Ocean, had they been so numerous on the Pacific side at the time when a communication existed between the two oceans. Whilst this paper was passing through the press, I found two notices of the existence of water-snakes on the western coasts of South America, in seas considerably more southwards than the Bay of Panama. The notes are in Capt. Sharp's Voyage in " The History of the Bucaniers of America." London, 1699, 8vo, vol. ii. p. 50 : " As we sailed " [near Cape St. Francisco, which is nearly under the equator] " we saw multitiides of Grampusses every day ; as also Water-snales of divers colours."' And p. 72, when sailing in lat. 19° S., the author mentions " A huge shoal of fish, two or three Wafer-snakes, and several Seah." I find in another part of the same work a note which I believe to be the first description of Tapinis bairdi. The part has a separate title-page, "A Journal of a Voyage made into the South Sea by the Bucaniers or Freebooters of America from the year 1684 to 1689. Written by the Sieur Eaveneau de Lussan." Lond. 1698, 8vo. The Indian name of the Tapir is given as Alcmijwwye, page 16. DE. GUNTHER OX THE FISHES OF CENTRAL AIMEEICA. 401 These remarks appear to me to convey very strong testimony in accordance with my own observation on the ichthyological fauna, inasmuch as the author refers the Panama MoUusks generally to the Atlantic fauna. He, indeed, denies the perfect identity of the species, admitting merely an "analogy" between them; but then it is a question whether malacologists do not go too far in making specific distinctions, when they are not even able to express those distinctions " diasrnostically," recognizing the forms merely " by their general habit." Shells are, after all, that portion of a moUusk the formation and development of which is most influenced by the peculiarities (physical and chemical) of the surrounding medium and locality ; and only too many specific forms have been distinguished on account of slight differences in the sculpture and shape of the shells, the importance of which disappears on comparing a large series of examples. However, as I am not prepared to form an opinion with regard to the shells of Central America from my own examination, I am bound to receive the testimony of so celebrated a malacologist as Dr. Morch ; and should his observations prove to be fully correct, they will give an additional interest to this fauna, as proving that the shells of MoUusks suffer change under circumstances in which the specific charactei-s of fishes remain unaltered. With regard to fossil shells, Mr. J. C. Moore, who has examined several collections from tertiary beds in San Domingo, has made the observation that " many bear a strong resemblance to shells now living in the Indian Seas and the Pacific, and that one or two appear to be identical" (Quart. Journ. Geol. Soc. 1853, p. 131), and " that a channel or sound may have existed in the equatorial parts during some portion of the tertiary period, by which some few of the tropical shells may have migrated from the one ocean to the other" (ibid. 1850, p. 43). Of the other marine animals, the Corals have been made the object of elaborate researches, the various authors arri\dng at somewhat different conclusions. First, Mr. Duncan, in a paper " On the Fossil Corals of the West-Lidian Islands" (Quart. Journ. Geol. Soc. xix. 1863. p. 455), has shown that "in all the calcareous formations which are covalliferous, and are considerably elevated above the level of the Caribbean Sea [being probably of miocenc age], there is a very limited series of Corals witli generic relation to those now existing and characteristic of the West-Indian Coral Fauna, but a predominance of forms resembling those of the present Coral-seas of the Pacific, South Sea, and the Indian Ocean." Tiiis identity of the Corals proves an identical condition of the phj^sical circumstances, and evidently a wide contiiuiity of the West-Indian and Western seas. On the other hand, Prof. Veeeill, when speaking of the living Polyi)-fauiKV of the Atlantic and Pacific sides of Central America (Proc. Post. Soc. Nat. Hist. x. 1866, p. 323 et seq.), states that their differences of character are verj- remarkable; that at Panama none of the reef-building corals of Asi)inwall, Florida, or the West Indies occur, nor even any of the genera of the families to which they belong, with the VOL. VI. — PART VII. 3 K 402 DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. excei^tion of a small Porifes and Stephanocom ; that these and other differences do not favour tlie theory entertained by some geologists, viz. that there has been a communi- cation between the two oceans at this point, and that the Gulf-stream flowed across the isthmus into the Pacific, within comparatively recent geological times. It is not within the scope of this paper further to discuss the point on which Messrs. Duncan and Verrill are at variance, as we cannot assume that the present fish-fauna existed at so early a period. From tiie observations made on the fishes and shells Ave are obliged to conclude that down to a very recent period a connexion between the two seas has been kept open by channels and straits wide enough to allow of the passage of these animals. Why corals, or at least a part of them, should not have been dis- persed by their floating germs in a similar manner, is a circumstance which we cannot explain. The occurrence of identical species oi freshwater fishes in rivers running to the two opposite oceans is a matter of much less difficulty, and, besides, has been very generally observed in various parts of the globe. The same agencies which in other countries have effected a -wider dispersion of one species than of another must have been at work here also. Prof. M. AVagner has, in his Memoir quoted above, so fully treated of this part of our subject, with particular reference to the hydrographical peculiarities of the isthmus, that we need not dwell further on it. § 7. Definition of the Characteristics of the Fishfauna of Central America. In defining the zoological characters of Central America, expressed in its fish-fauna, I confine myself to the freshwater fishes proper. Here the nearctic types become extinct, and are represented by five generic types, four of which, although with numerous species in the north, have but a single one here — Lepidosteiis, Amiiirxs, Sclerognathus, and Haplochilus. Fwndulus, extending a little further southwards (with one species in Western Ecuador), is represented by four species in Guatemala. Not one of these species is identical with a North-American. Much greater is the aflSnity with neotropical types ; and their representatives are much more numerous : there is one species of Acara, one of Macrodon, seven of Tetragono- pterus, one of Anacyrtus, twelve of Pimelodus, one of Plecostomus, two of Chcefostormis, two of Loricaria, one of Anableps, one of Carapus, the latter being identical with a species from Guiana. Types in common with the West-Indian Islands are — Agono- stoma with three species (one of which is said to be identical with a Jamaican species), Girardinus and Gambusia with one, the two latter genera being also represented in the Southern States of North America. The Siluroid genus Arius, which extends over the tropics generally, is represented by nine species. Finally, the following genera are peculiar to Central America, or at least have attained there to the greatest development : — Heros and the allied Neetroplus and Petenia with thirty-four species, yElurichthys with two, Chalcinopsis with three, Characodon with one. DK. GUNTHEE OX THE FISHES OF CENTEAL AMEEICA. 403 Xiphoi)horns \\\t\\ one, Mollimesia with one, Pa-cilia with eight, and Belonesox with one species. The affinity of this freshwater fauna with that of Mexico, will be found to be greater than with that of any other country (I might mention about ten species common to Guatemala and Mexico) ; but until we are better acquainted with the habitats of species described as Mexican, a more detailed comparison of the two countries would be of but little advantage. The freshwater fish-founa of Central America may be shortly thus characterized:—^ part of the Chromides (Heros, &c.) and the Gijimnodontes (jenerally have attained to their greatest development; neotropical types extending northxcards prevail over nearctic extending southwards, the latter heing represented by a few extreme branches. § 8. An Attempt to Subdivide this Fauna info Provinces. We may subdivide this part of the freshwater fauna into the following provinces : — A. Hie fresh waters north of the Lakes of Managua and Nigaragua, emptying info the Pacific. — To this province belong the fishes collected at Chiapam [C'h.], Huamuchal [H.], San Jose [J.], in the Rio Guacalate (Duenas) [G.], San Salvador [S.], and Libertad [L.] ; also the fishes from the Lakes of Amatitlan [Am.] and Atitlan [At.] may be referred to the same province, [The species printed in italics in the following lists are found also in one or more other provinces, and in Atlantic rivers.] Heros macracanthus Cli. H. — — — — trimaculatus Ch. H. — — — , • iiigrofasciatus — — — — — -^™- ^^- guttulatus — — — — — ■^™- Agouostoma microps — — — G. Arius guatemalensis Ch. — — — platypogon — — — J- — csenilescens — H. — — — Pimelodus guatemalensis — H. — - — — — Anacyrius yuatemalensis — H. — — — — Tetragouopt. microj)hthalmus .... — — — — — -A-m. huniilis — — — — — ^^™' Funduhis giiatemalensis — — — G. — Am. pachycephalus — — — — -^^- punctatus Ch. — — — — Anableps dovii Cli. — - — — — Pcecilia mexicana Ch. H. — G. — Am. — dovii — — — — — Am. — Girardiniis pleurospilus — — — G. — — — Chipea libertatis — — — — L- — Lcpidosteus tropicus — H. — — — — — 3k2 M. Y M. — M. M. Y M. Y M. M. M. 404 DE. GtJNTHER ON THE FISHES OF CENTRAL AMERICA. B. The fresh waters north of the Lakes of Managua and Nicaragua, em})tying into the Atlantic. — To this province belong the fishes collected in the Rio Usumacinta [U.] (and in its tributaries Rio de Santa Isabel, Rio Chisoy, and Rio San GeronimoX in the Rio Motao-ua [M.], and in the Rio Cahabon (Yzabal) [Y.]. Eleotris dormitatrix — Ayonostoma nasutuin U. Heros parma — • spilurus — aureus — motaguensis — microphthalmus — ■ oblongus — augiilifer — — ^- salvini U. — irregularis U. — — godmanni — Arius assimilis — melanopus — Amiurus meridionalis U. Pimelodus godmanni U. motaguensis — salviui U. polycaulus U. Tvtragonopterm paiiuinetiftis — brevimanus U. Clialcinopsis dentex l^- I'unduhis labialis U- Belouesox bebzanus Honduras, Bebze. PcEciba chisoyensis U- — — Xiphophorus heUeri U. — — Selerognatbus meridionabs U. — — Carapus fasciatiis — ^i- — C. Lake Pet en. — The fish-fauna of this limited district is so pecuharly developed, that we cannot hesitate to describe it as a separate province. — Y. — Y. M. — M. M. — — Y. — Y. M. Y. Y. Heros margaritil'er. melauurus. urophthalmus. aflBnis. friedricbsthabi. salvini (in common \\ itli vinco B). intermedius. Peteuia splcndida. Pimelodus petenensis. Tctragonopterus petenensis. Belonesox (in common with province B). Poecilia petenensis. Molbenesia petenensis. Cliaitoessus petenensis. DR. GiJNTHER ON THE FISHES OF CENTRAL AlVIERICA. 405 D. Lake of Managua. — Althougli the number of species known from this lake is small, the forms are quite peculiar ; we find here those species of Hews which are distinguished by the extraordinary development of the lips, or by incisor-like teeth, which render the separation into a distinct genus necessary. The development ot these Chromides is the more remarkable, as this lake occupies a space which is supposed to ha^'e been a portion of a marine channel. Ileros ervtlmcus. Ilcros lobocliihis. managuensis. labiatus. • multispinis. Neetroplus nematopus. E. Lake of Nicaragua. — Also the fishes of this lake are, with two exceptions, pt-culiar ; like Lake Managua, it appears to have been part of a marine channel. Heros lubiatm (Lake of ^Luiagiia). Pimelodus nicaraguensis. Gambusia nicaraguensis. Poecilia dovii (in common with Lake Amatitlan). Eleotris longiceps. Heros longimanus. citrinellus. dovii. nicaraguensis. F. The fresh waters south of the Lakes of Managua and Nicaragua to the Lsthmas of Darien. — We are obliged, at present, to unite into one province the fish-fauna of Costa Rica, Veragua, Panama, and Darien, as our knowledge of the fishes of Costa Rica and also of Veragua is too incomplete to admit of a comparison with those of the more southern part of the isthmus. This is the more to be regretted, as a former separation of these two parts and of their faunas is, as we have explained above, a matter of great probability. The fishes of the Chagres River show a decidedly South-American cha- racter. The identity of the freshwater fish-faunas of the Pacific and Atlantic sides is here easily explained by the narrowness of the isthmus. Eleotris picta Agonostoma nasutuui — moniicola — Heros parma ("liagrcs. altifrons Acara cieruleomaculata .... Chagres. Arius multiradiatus — ^Elurichthys dorsalis — panamensis — Pimelodus wagueri — modestus Chagres. Plecostonuis, sp Chagres. Ch;etostomiis aspidolepis .... — ('hcelos-lomiis 'f rirr/iosiis .... Chagres. Loricurhi limn Chagres. R. Bavano. Panama. Panama. Western ^'e^agua. R. Bavano. Panama. Panama. Panama. Veragua. 406 DE. GUNTHEE ON THE FISHES OF CENTRAL AMERICA. Loricaria uracmitha Chagres. — — Macrodon microlepis Chagres. — — Tetragonopterus jMnameiisis ... — Pauama. — aeneus — Panama. — Chalcinopsis striatulus .... — ■ Panama. — chagrensis Chagres. — — Anacyrtus guatemalensis .... Cliagres. — — Haplochilus dovii — Costa Rica. — Poecilia elongata — Panama. — — — gillii Chagres. — § 9. Descnjjtive Part. In the following descriptive part of this Memoir I have admitted full descriptions of those species only which are not described elsewhere ; secondly, descriptive diagnoses of those of which figures are given ; and, finally, notes on some knoAvn species, if they appeared to contribute to their better knowledge. For the descriptions of all the other species (the insertion of which would be a repetition of matter already published), I must refer the student to my general work on Fishes. 1. Centropomus appendiculatus. Poey, Mem. Cub. ii. p. 119. D. 8jl. A.;l L.lat. 70-72. Nine longitudinal series of scales between the origin of the second dorsal fin and the lateral line. The height of the body is contained four times in the total length (with- out caudal), the length of the head twice and two-thirds. Prseorbital indistinctly serrated ; suboperculum produced into a short flap, which extends to or nearly to the vertical from the origin of the dorsal fin. The intermaxillary extends to below the middle of the orbit. Dorsal spines of moderate strength ; the third is the longest, and about half as long as the head. The second anal spine is generally longer than the third ; but sometimes they are equal in length, and even shorter than the third dorsal spine. The length of the ventral fin is more than one-half of its distance from the anal. Air-bladder ^vith a pair of appendages anteriorly. Silvery ; dorsal fins blackish ; lateral line black. We have received this species (which was originally described from Cuban examples) from Surinam and Mexico. Mr. Salvin and Capt. Dow obtained a specimen from the Chagres River, 10 inches long. 2. Centropomus medius. Giintli. Proc. Zool. Soc. 1864., p. 144. D. Sjl. A.i L.lat. 57. Eight longitudinal series of scales between the origin of the second dorsal fin and the lateral line. The height of the body is contained thric(! and three-fourths in the total DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 407 length (without caudal), the length of the head twice and four-fifths. Prseorbital finely serrated ; suboperculum produced into a flap, which does not extend to tlie vertical from the origin of the dorsal fin. The intermaxillary extends somewhat beyond the anterior margin of the orbit. Dorsal spines strong ; the third is longer than the fourth, and half as long as the head. The second anal spine long, but a little shorter than the third, and etjual in length to the distance between the extremity of the upper jaw and the pra?opercular margin. The length of the ventral fin is much more than one-half of its distance from the anal fin. Lateral line black. Two specimens, l.S inches long, from Chiapam. .3. Centropomi'S nigrescens. (jriintli. Proc. Zool. Soc. 1864-, p. 144. D. 8 1 1 A. I L. lat. 70. Ten longitudinal series of scales between the origin of the second dorsal fin and the lateral line. The height of the body is contained four times and a half in the total length (without caudal), the length of the head twice and four-fifths. Praeorbital not serrated; suboperculum produced into a short flap, which does not extend to the vertical from the origin of the dorsal fin. The intermaxillary extends a little beyond the middle of the orbit. Dorsal spines rather feeble ; the third and fourth are equal in length, two-fifths of the length of the head. The second and third anal spines also are equal in length, and not longer than the dorsal spines mentioned. The length of the ventral fin is scarcely more than one-half of the distance of its base from the anal. Air-bladder without appendages anteriorly. Silvery ; upper parts and fins blackish ; lateral line black. One specimen, 14 inches long, from Chiapam. This species is allied to C. appendicnlatus (Poey), but diflers externally in its con- siderably more feeble and shorter fin-spines. 4. Cextropomus parallelus. Poey, Mem. Cuba, ii. p. 120. D. 8 1 1 A. % L. lat. 85-90. ' lU D Twelve longitudinal series of scales between the origin of the second dorsal fin and the lateral line. The height of the body is contained thrice and three-fourths in t]u> total length (without caudal), the length of the head twice and a half. Pntorbital distinctly serrated ; suboperculum produced into a flap, which extends to the vertical from the origin of the dorsal fin. The intermaxillary extends a little beyond the middle of the orbit. Dorsal spines rather feeble ; the third is the longest, half as long as the head. The second anal spine is exceedingly strong, longer than the third and the thii-d dorsal spine. The length of the ventral fin is considerably more than one-half of 408 DE. GUNTHEE ON THE FISHES OF CENTEAL AMEEICA. tlie distance of its base from the anal. Air-bladder without appendages anteriorly. Silvery ; upper parts and fins greenish ; lateral line not black. This species occurs in Cuba ; we have received it from San Domingo, Jamaica, and Bahia. Messrs. Dow and Salvin collected a specimen in the Chagres River. 5. Centropomus armatus. (iill, Proc. Ac. Nat. Sc. Philad. 186.3, p. 163. D. S\L A. I L. lat. 51. L. transv. 714. Six longitudinal series of scales between the origin of the second dorsal fin and the lateral line. The height of the body is contained from thrice and two-fifths to thrice and three-fourths in the total length (without caudal) ; the length of the head twice and a half. Pra^orbital serrated in its hinder half; suboperculum produced into a long flap, which extends beyond the vertical from the origin of the dorsal fin. The inter- maxillary extends scarcely to below the middle of the orbit. Dorsal spines of moderate strength ; the third is the longest, and half as long as the head. The second anal spine is exceedingly strong, much stronger than the third, and longer than the third dorsal spine. The length of the ventral fin is scarcely more than one-half of the distance of its base from the anal. Silvery ; dorsal fins, a blotch on the opercle, and the membrane between the anal spines blackish. Lateral line not black. Several specimens, 12 inches long, were collected by Mr. Salvin at Chiapam. 6. CeNTROPOMUS EXSiFERUS. Poey, Mem. Cub. ii. p. 122, pi. 12. fig. 1. D. 8 1 1. A. I L. lat. 53. Seven longitudinal series of scales between the origin of the second dorsal fin and the Interal line. The height of the body is one-fourth of the total length (without caudal), the length of the head two-fifths. Praeorbital coarsely serrated ; suboperculum pro- duced into a flap, which extends to the vertical from the origin of the dorsal fin. The intermaxillary extends scarcely to below the middle of the orbit. Dorsal spines of moderate strength ; the third and fourth are the longest, and two-fifths as long as the head. The second anal spine is exceedingly strong, much stronger than the third, and much longer than the dorsal spines. The length of the ventral fin is somewhat more than one-half of the distance of its base from the anal. Silvery ; dorsal fin, a blotch on the opercle, and the membrane between the anal spines blackish. Lateral line not black. This species occurs in Cuba ; we have received it from Jamaica and from the Guyanas. Mr. Godraan collected a specimen. 12 inches long, at Belize. DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 409 7. Centropristis jucropoma. (PI. LXV. fig. 1.) Giiuth. Proc. Zool. Soc. 1864., p. 145. D. 1^. A. ^. L. lat. 52. L. transv. 6/16. Closely allied to ('. radialis, Q. & G. ; but whilst tliat species has a notch above the spinifeious angle, the present has its prajopercular margin not interrupted, the long spines of the angle gradually passing into the finer serrature. There are six series of scales between the eye and the angle of the prajoperculum. The maxillary extends nearly to the vertical from the posterior margin of the orbit. Dorsal fin with a notch, the ninth spine being considerably shorter than the tenth. A series of rather small brownish spots above and below the lateral line. Three specimens, 4^ inches long, were collected by Messrs. Dow and Salvin on the Pacific coast of Panama. 8. Serranus creolus, C. & ^^ T have examined specimens from the Atlantic coasts only ; but Mr. Gill has found it iu a collection of fishes from Lower California, the specimens being undistinguishable from those of the "West Indies and South America (Proc. Ac. Nat. Sc. Philad. 1862, p. 249). 12. SeRRAiNUS SELLICAUDA. Ejnnephelus sellicauda, Gill, Proc. Acad. Xat. Sc. Philad. 1862, p. 250. D. ]4- A. l L. lat. 100. Caudal fin with the posterior margin convex. The height of the body is rather more than three-fourths of the length of the head, and one-fourth of the total (caudal included). The diameter of the eye is one-fourth of the length of the head. Prseoper- culum finely serrated behind, with some coarser teeth at the angle, lower limb entire ; sub- and interoperculum entu'e. Ventrals three-fourths of the length of pectorals, and reaching two-thirds of the distance between their insertion and the commencement of the anal. Brownish, with olive-coloured spots of larger and smaller size on the body and opercles. All the fins with a narrow white margin. A square black blotch across the back of the tail. iJescription. — Body not very elevated ; its greatest height is below the third spine of the dorsal fin, rather more than three-fourths of the length of the head, and one-fourth of the total. The distance between the end of the dorsal and the commencement of tiie caudal is nearly one-sixth of the length of the base of the dorsal, is contained once and two-thirds in the base of the anal, is one-fourth of the distance between the dorsal tin and the snout, and equals the least depth of the tail. The distance between the eyes is one-half of the diameter of the eye, and covered with very minute scales, which are found also on the prneorbital around the nostrils. The length of the snout is two- thirds of the diameter of the eye. The maxillary bone reaches the -vertical from the posterior margin of the eye. The mandibulary is one-half of the length of the VOL. VI. — PART VII. o L 410 DE. GtJNTHEE ON THE FISHES OF CENTE.y:. AMEEICA. head. The lips are not very thick. Posterior limb of prseoperculum very convex, minutely serrated, with three coarser teeth at the angle ; lovi^er limb toothless. Sub- and interoperculum entire. Operculum terminating in three not very strong teeth, the upper of which is somewhat more remote than the others, the middle one being the more prominent. Suprascapular concealed by the scales. The membrane of the dorsal fin is scaly for about half the height between the spines and rays ; the spinous portion scarcely lower but longer than the soft, with its upper margin convex, and a small membranaceous appendage behind the tip of each spine. The first spine is the shortest, rather more than half the length of the second, which is one-fifth shorter than the third ; from the third to the seventh the spines are equal, becoming slightly shorter at the eighth ; the last two spines are of equal length. The rays increase slightly from the first to the sixth, after which the upper margin is straight, becommg again rounded posteriorly. The first ray is one- fifth longer than the preceding spine. Caudal with posterior margin convex. The commencement of the anal is on a line with that of the soft dorsal, and it ends before the termination of the dorsal ; the first spine is short, not half the length of the second, which is long and strong, longer than any of the dorsal spines ; the third is slenderer, and equal to the third dorsal spine : the margin of the soft part of the fin is nearly straight, sharply rounded off posteriorly. The pectoral consists of eighteen rays, is rounded, and longer than the ventral, and covered with very minute scales to one-third of the length. The ventrals reach the vent ; the second ray is the longest, the spine being equal to the second of the dorsal. Canine teeth of moderate size, those of the lower jaw rather small. Coloration as described above. A single specimen, 4 inches long, was sent by Capt. Dow from the Pacific coast of Panama. The specimen in the collection of the Smithsonian Institution is from the coast of Lower California ; a statement of its size, which would have been of some importance, is omitted. 13. Serranus analogus. Epinephelus analogus, Gill, Proc. Acad. Nat. Sc. Philad. 1863, p. 163. D. }5. A. I L. lat. ca. 100. Adult. — The height is contained thrice in the total length (without caudal), the length of the head twice and two-thirds. The prajoperculum is finely serrated behind, and towards the angle armed with three or four strong teeth. The diameter of the eye equals a sixth of the head's length, and equals the interorbital space as well as the snout behind the intermaxillaries. The third, fourth, and fifth spines are equal, and contained twice and two-thirds in the length of the head ; the tenth thrice and a half The caudal fin enters five times and a half in the length, the height of the dorsal twice and three-fourths in the head. The anal is deeper; its third spine is longest, and enters four times and three-fourths in the head's length ; the pectoral is at least half as long as the head ; the ventral shorter, but coterminal with it. DE. GUNTIIEE ON THE FISHES OF CENTRAL AMERICA. 411 The colour is purplish grey, with numerous dark spots about as large as the pupil ; those of the pectoral and caudal fins are smaller and more crowded, of the dorsal, anal, and ventral more like those of the body. The specimens are from 11 to 15 inches long, and were found by Capt. Dow at Panama. We have received also a s»i«//^>' example, 5 inches long, from the same gentleman. It differs from those described above in being provided with five cross bands, paler in colour than the spots, which are one-thu'd the size of the eye. The dorsal fin is scarcely notched, the tenth spine being but little shorter than the third or fourth, the length of which is contained twice and two-thirds in that of the head. The example being young, its eye is comparatively lai'ger. 14. Plectropoma afrum. (PI. LXVII. fig. 3.) Ejnnephelus afer, Bloch, Taf. 327 (fide Peters, Monatsber. Ak. AViss. Berlin, 1865, p. 105). Alphestes afer, Bl. Schn. p. 236. Plectropoma chloropterum, Cuv. & Val. ii. p. 398. Poey, Mem. Cub. i. p. 7B, lam. 9. fig. 3. monacanthus, Miill. & Trosch. in Schomb. Hist. Bai-b. p. 605. Giinth. Fish. i. p. 164. muUiguttatum, Giinth. Proc. Zool. Soc. 1866, p. 600. D. jiL.. A. I L. lat. 75. Caudal rounded. The height of the body is equal to the length of the head, and contained twice and three-fourths in the total (without caudal). The diameter of the eye is one-fifth of the length of the head, and a little less than that of the snout. Prseoperculum with a strong spinous tooth below the angle, pointing forwards. Olive- brown, head and body with numerous spots. Description. — Body somewhat elevated ; its greatest height is below the fourth spine of the dorsal, and equal to the length of the head, which is contained thrice and one- third in the total (the caudal included). The distance between the dorsal and the caudal is contained seven times and one-third in the length of the base of the dorsal fin, twice in that of tlie base of the anal, four times in the distance between the dorsal fin and the snout, and is considerably less than the least depth of the tail. The distance between the eyes is about two-thirds of the diameter of the eye, and covered with scales which extend forward beyond the nostrils on the praiorbital, and in a narrow band on the upper maxillary. The length of the snout equals the diameter of the eye, which is one-fifth of the length of the head. The maxillary reaches a little beyond the level of the posterior margin of the eye. The mandibulary is covered with minute scales, and is equal to one-half the length of the head. The lips are thick and fleshy. The posterior limb of the prseoperculum slants obliquely backwards, and is minutely serrated, the denticulations becoming coarser at the angle ; and beneath on the lower limb at some distance from the other teeth there is a single strong tooth pointing downwards, and nearly concealed by the skin ; sub- and intcroperculum not serrated. 3l2 412 DK. GUjSTHEE on THE FISHES OF CENTEAL AMEEICA. The operculum terminates in three, flat, triangular teeth, the upper of Avhich is the more distant and somewhat more obtuse than the others, the middle one being the longest, but not very prominent, and the lower one the shortest and weakest. The suprascapula is concealed by the scales. Base of dorsal fin covered with very small scales, a tapering band of scales runs up between each pair of the spines and rays to about half the height of the fin. Spinous portion rather lower but longer than the soft, its upper margin convex; the membrane between each spine is deeply notched, and there is a small membranaceous appendage behind the top of each spine. The fii'st spine is the shortest, half the size of the second ; the second is five-sixths of the length of the third ; the third, fourth, and fifth are the longest, and of nearly equal length ; the spines then become gradually shorter to the last one, which is scarcely longer than the preceding. The soft portion exhibits an entii'ely rounded upper margin, the rays becoming longer from the first to the sixtli or seventh, and shorter from the fourteenth to the last ; tlie first ray is one-fourtli longer than the preceding spine. Caudal with the posterior margin convex. Anal commencing a little behind the commencement of the soft dorsal, and terminating in advance of the end of the same ; the first spine is not very strong, and short ; the second long, thick, and strong ; the third more slender and shorter, being but little longer than the second dorsal spine ; the margin of the fin is rounded throughout, the third ray being the longest, and the subsequent ones becoming progressively shorter. The pectoral is composed of eighteen rays, rounded, one-fourth longer than the ventral, and covered with minute scales for about oue-third of its length. The ventral reaches to the vertical from the origin of the eighth spine of the dorsal, but not to the vent ; the spine is a little less than two-thirds the length of the first ray ; the first and second rays are the longest, the others diminisliing gradually in length ; the length of the spine is somewhat less than that of the second dorsal spine. Canine teeth small in both jaws. This species varies somewhat in coloration, as most of its congeners ; the spots are numerous and small, either of a uniform dark-brown colour, or of a light colour and mixed with large brown spots. Pectoral fins with narrow blackish cross bands. One example, 10 inches long, and three smaller ones have been collected by Capt. Dow on the Pacific coast of Panama. The latter have the spots somewhat larger and less conspicuous than the adult. This species occurs also in the West Indies and at the Falkland Islands. 15. Rhypticus decoratus. Rhijpticus nigripinnis, Gill, Proc. Ac. Nat. Sc. Philad. 1861, p. 53. Promicropterus dtcoratu.i, (Jill, /. c. 1863, p. 164. D. 4 A. 16. The two dorsal spines are continuous with the soft portion. Body generally with more or less numerous round whitish spots, many of which have a brown centre. DK. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 413 Messrs. Dow and Salvin have collected several examples, from 3 to 8 inches long, on the Pacific coast of Panama. The species described by llolbrook as 11. iriaculatus, and said to have the ilorsul spines separated from the soft portion, may eventually prove to be identical with the Pacific fish. 19. Mesopriox aratus. (jiinth. Proc. Zool. Soc. 18G4, p. 145. D. ^^. A. |. L. lat. 45. L. transv. 4i/12, The height of the body equals the length of the head, and is contained thrice and two-fifths, or thrice and one-fifth in the total (without caudal). The maxillary does not extend backwards to the vertical from the centre of the eye. Praioperculum finely serrated, with scarcely a trace of a posterior notch. Dorsal spines of moderate strength ; the third and fourth ai'e the longest, two-fifths of the length of the head ; the eleventh is scarcely longer than the tenth, which is rather more than half as long as the fourth. Caudal fin emarginate, two-thirds scaly ; anal spines short, rather feeble, the tliird longer than the second, and equal in length to the last dorsal spine. Upper and lateral parts brownish-olive, each scale with a peaii-coloured spot, the spots forming together very distinct longitudinal stripes ; no black lateral spot ; hind part of the root of the pectoral brown. Lower parts salmon-coloured. We have six examples : two, 15 inches long, were collected by jNIr. Salvin at Cliiapiun ; and four others were sent by Capt. Dow from the Pacific coast of Panama. 21. Apogox dovii. (math. Proc. Zool. Soc. 1861, p. 371. D. 6 I J. A. '=. L. lat. 25. L. transv. 3/9. A roundish black spot on each side of the root of the caudal ; the spinous dorsal colourless, ti'ansparent ; uniform olive (in spirits). Head densely punctulated witli brown. Only the hind margin of the posterior proeopcrcular ridge is serrated. Dorsal fins nearly equal in height. The height of the body is one-third of the total length (without caudal) ; the hnigth of the head two-fifths; eye large, its diameter being more than one-third of the length of the head. Palatine and vomerine teeth present. The upper jaw overlaps slightly the lower ; maxillary extending backwards to below the posterior third of the orbit. Operculum with an up])er flexible point, and with a lower stiff spine. The third dorsal spine is a little longer than the second, one-half the length of the head. Caudal fin slightly emarginate, with the angles rounded. Total length 26 lines. This species is so closely allied to Jl. iiier/nis from the Mediterranean, that perhaps 414 DE. GUNTHEE ON THE FISHES OF CEXTEAL AMEEICA. it would be better not to separate it ; the only difference which I can find is the form of the soft dorsal fin, which is considerably higher than the spinous in the MediteiTanean species. 22. Pristipoma melanopteeum. Pnstipoma melanopterum, Cuv. & Val. v. 1830, p. 273. bUineatum, Cuv. & Val. v. 1830, p. 271, pi. 122. Hamulon melanopterum, Ranzani, Comm. Bonon. v. 1842, p. 343, tab. 30. Pristipoma melanopterum, Giintli. Fish. i. 1859, p. 287. Var. Genytremus interruptus, GiU, Proc. Acad. Nat. Sc. Philad. 1862, p. 256. Pristipoma melanopterum, Giinth. Proc. Zool. Soc. 1864, pp. 23 & 27. This species occurs on both sides of Central America, Capt. Dow having collected specimens at Panama and Colon. Mr. Gill has found it also in a collection of fishes from Lower California. He describes his Pacific specimen as a distinct species ; but the distinctive characters are, according to my views, not of specific value. He mentions it in the following terms : — " The species is so closely allied to hilineatus, that it might be even considered as a variety, but it appears to differ by the steel-blue colour of the back, and the discon- tinuance of the lateral band a short distance before the spot on the taiP ; at its end the liand is bounded below by the lateral line. In other respects, the two species are so similar, that a detailed description would be only a repetition of that of hilineatus.'' 23. Pristipoma virginicum. We have examined specimens of this species from the West Indies, from the Atlantic coasts of Central America, and from Bahia. Mr. Gill has described an example from Panama under the denomination of Anisotremus tceniafus, Proc. Ac. Nat. Sc. 1861, p. 107. Although six or seven is the normal number of longitudinal bands, it is some- times increased by a more or less complete division of one or several bands. It appears to be more natural to consider the golden colour the ground-colour than the blue, as after death it fades into the same colour as that of the space between the black vertical bands. In e^Z specimens, I have found the bluish bands edged with purplish. Mr. Gill, in describing his A. tceniatiis, has taken the blue colour as ornamental, whilst in his description of A. virgiiiicus the character assigned to the colours is reversed, and the blue colour regarded as ground-colour. There is no specific difference between these fishes. 24. Pristipoma dovii. Guuth. Proc. Zool. Soc. 18G4, p. 23, pi. 3. fig. 1. D. }=. A. J. L. lat. 48. L. transv. 8/15. The height of the body is one-half of the total length (without caudal) ; the length of the head one-third. Snout obtuse, not much longer than the eye ; cleft of the ' This is also the case in some Atlantic specimens. DR. GUNTHER ON TIIE FISHES OP CENTRAL AMERICA. 415 mouth small, the maxillary extending to the vertical from the anterior margm of the orbit. Lips thick ; a pair of pores on the symphysis of the lower jaw, a central groove behind it. Snout naked, the remainder of the head being scaly. The width of the interorbital space is much less than that of the orbit. Dorsal and anal spines exceed- ingly strong ; the third of the dorsal fin is the longest, and nearly two-thirds as long as the head. The second anal spine is much longer than the third, and a little shorter (but stronger) than the third of the dorsal fin. Each ray of the soft fins is accompanieil by a series of minute scales, but only on the caudal fin are these scales dense enough to cover the rays. Caudal fin slightly emarginate. Silvery, with four black cross bands ; the first runs from the occiput, through the eye, to behind the angle of the mouth, the second from before the dorsal fin to below the base of the pectoral, the third from the base of the sixth, seventh, and eighth dorsal spines to the vent ; the fourth descends from the origin of the soft dorsal to that of the soft anal. Fins blackish. The cross bands appear to become fainter in old age. Two specimens, 8^ and 9 inches long, in the collection from Panama. 25. Pkistipoma chalceum. Giinth. Proc. Zool. Soc. 1864, p. 146. D. {j. A. ^,. L. lat. 5G. L. transv. 11/19. The height of the body is contained twice and two-thirds in the total length (without caudal), the length of the head thrice and a third. The diameter of the eye is nearly equal to the width of the interorbital space, and two-thirds of the extent of the snout. The maxillary does not extend backwards to the vertical from the anterior margin of the orbit. Prseoperculum minutely serrated behind, with the angle rounded, but not produced. There is no notch between the spinous and soft portions of the dorsal fin, the hinder spines being only a little shorter than the anterior rays ; dorsal spines of moderate strength, the fourth being the longest, not quite half as long as the head ; anal spines short, the second being only a little longer than the thii-d, two-sevenths of the length of the head. Caudal fin subtruncated, scarcely emarginate. Dorsal and anal perfectly scaleless. The pectoral fin extends to the vertical from the vent. Bronze-coloured, shining silvery, perfectly immaculate; vertical fins blackish, with an indistinct light band along the base. One specimen, 8 inches long, was discovered by Messrs. Dow and Salvin on the Pacific coast of Panama. 2G. Pristipoma humile. Kner & Steindachner, Sitzgsber. Ak. Wiss. Miinch. 1863, p. 222; and Abhandl. bayer. Ak. Wiss. x. p. 3, tab. 1. fig. 1. D. % A. I L. lat. 56. L. transv. n.^,.. Cjec. pyl. 3. The height of the body is contained thrice and two-tliirds in tlie total length (without 4]fi DE. GUNTHER ON THE FISHES OF t'ENTRAL AIMEEICA. caudal), the length of the head thrice. The diameter of the eye equals the widtli of the interorbital space, is one-fifth of the length of the head, and two-thirds of the extent of the snout. Snout produced ; cleft of the mouth wide ; the maxillary extend- ing beyond the front margin of the eye. Praeoperculum with the hind margin vertical and finely serrated. The spinous and soft portions of the dorsal fin are separated by a notch ; dorsal spines modevately strong, the fourth being the longest, its length being contained twice and three-fourths in that of the head. Second anal spine exceedingly strong, more than half as long as the head. Caudal fin slightly emarginate ; pectorals terminating at some distance before the vent. Scales ctenoid. Coloration uniform. This species is known from a single example (size not stated) from the Rio Bayano (Panama). 27. Pristipoma macracanthum. (PI. LXIV. fig. 1.) (xunth. Proc. Zool. Soc. 1864, p. 146. D. ll|i. A. 3/8. L. lat. 47. L. transv. 0/13. The height of the body equals the length of the head, and is one-third of the total (without caudal). The diameter of the eye equals the width of the interorbital space, iind is two-thirds, or somewhat less than two-thirds, of the extent of the snout. Hind margin of the anterior nostril with a broad flap. Snout somewhat produced ; the maxillary does not extend to below the anterior margin of the eye. Prgeoperculum with the hind margin rather concave, and with stronger teeth at the angle, which is rounded. The spinous and soft portions of the dorsal fin are separated by a deep notch, the spine of the soft portion being much longer than the preceding, which is somewhat longer than the second. Dorsal and anal spines exceedingly strong; the fourth dorsal spine is the longest, its length being contained tmce and a third in that of the head. The second anal spine much longer and stronger than the third, and even than the fourth dorsal spine. Caudal fin truncated. Each soft ray of the vertical fins is accompanied by a series of minute scales. The pectoral fin extends to the vent. Scales smooth. Silverv. with several very indistinct dark cross bands on the back, which appear to be arranged as in P. leuciscus. Two specimens, 11 and 14 inches long, were collected by Mr. Salvin at Chiapam. 29. Pristipoma leuciscus. (PI. LXVI. fig. 3.) (ruuth. Proc. Zool. Soc. 18G+, p. 117. D. nil. A. 3,7-8. L. lat. 51. L. transv. ^\ i 1.J lU The height of the body is contained thrice or thrice and a third in the total length (without caudal), the length of the head thrice and a fourth. The diameter of the eye is equal to, or more than, the width of the interorbital space, but is less than the extent of the snout. The maxillary does not quite extend backwards to the vertical from the anterior margin of the orbit. Prieoperculum finely serrated behind, with the angle DR. GUNTHER ON THE FlSHE.S OF CENTRAL AMERICA. 417 rounded, and with the hind margin slightly concave. Tlie spinous and soft portions of the dorsal fin are separated by a deep notch, the spine of the soft portion being nearly twice as long as the preceding. Dorsal spines long, of moderate strength : the third is the longest, and one-half, or more than one-half, as long as the head. Anal spines rather strong : the third is a little longer than the second, equal to the seventh dorsal spine, and more than one-third of the length of the head. Caudal fin emarginate. Each soft ray of the vertical fins is accompanied by a series of minute scales. The pectoral fin extends to the vertical from the origin of the anal in the younger example, but is shorter in adult ones. Scales smooth, bright silvery ; young specimens with several very indistinct dark cross bands on tlie back, the first from the nape of the neck to the gill-opening, the second below the seventh dorsal spine, the third below the last dorsal spine ; old specimens with the marginal membrane of the operculum black. One specimen, 7^ inches long, was found by Mr. Salvin at San Jose. Three others, from 11 to 12 inches long, are from Chiapam ; and C'apt. Dow found it also at Panama, where it does not appear to be rare. 30. COXODON PACIFICI. (PI. LXIV. fig. ?).) Gunth. Proc. Zool. S(X'. 1861, p. It". D. 11 l^. A. ^. L. lat. 17. L. transv. 7/13. Diagnosis. — The spinous teeth at the angle of the praeoperculum are not much stronger than the others. The height of the body is contained twice and two-fifths in the total length (without caudal). One specimen, 12^ inches long, was collected by Mr. Salvin at Chiapam. Descrqjfioii . — The body is compressed, and considerably elevated; its greatest height, which is below the fifth dorsal spine, is contained twice and three-fourths in the total length. Upper profile lounded from the first dorsal spine to the nape, concave over the eyes, whence it descends abruptly over the snout. The upper surface of the head is very broad, the space between the eyes bemg nearly twice the width of the orbit. The snout is thick and obtuse ; the lips thick and fleshy. Teeth in a villiform band in both jaws, with an outer series of conical teeth. Chin with a median groove and a pair of pores. Posterior limb of praeoperculum straight, regularly and distinctly serrated, the teeth becoming gradually a little larger at the angle, and continued on the lower limb ; the entire surface of the pra-operculum is covered with scales, which are smaller than those of the operculum, and reach to the margin of the bone. The operculum has a notch behind, between two obtuse and feeble points. Suprascapular margin indistinctly toothed or roughened. The origin of the dorsal is in the vertical from the root of the pectoral, and its termination is vertically opposite to that of the anal ; the base of the spinous portion is nearly twice as long as that of the soft. The spines are strong, broader alternately on one side than on the other; the first is small, not quite one-half the length of the second, whicli is rather more than half that of the VOL. VI. — PART VII. 3 M 418 DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. third ; the third spine is three-fourths the length of the fourth ; the fifth is the longest, its length being contained twice and two-thirds in the height of the body ; the sixth and fourth spines are equal in height, and the subsequent spines decrease gradually iu length ; the twelfth, which must be considered part of the soft dorsal, is slightly longer than the preceding spine, and equal to the tenth. The soft portion has a rounded margin ; the third or highest ray is not quite equal to the fifth spine, and is twice as long as the last. The spinous portion as Avell as the soft can be received into a scaly sheath. The caudal fin is very slightly emarginate, scaly to within a short distance from its tip, and one of its longest rays is nearly one-fifth of the total length. The distance between the caudal and anal fins is less than the base of the latter ; the first anal spine is opposite to the third ray of the dorsal, it is strong, broader on the right side, and e.xcavated posteriorly, and is one-half the length of the second, which is very long and strong, equal in length to the fifth dorsal spine, and broader on the left side ; the third anal spine is equal to the third of the dorsal, and little more than half the height of the first ray ; the first and second rays are the longest, and the margin of the soft portion is vertical. The pectoral is moderately long, its length being contained four times and a half in the total. Root of ventral immediately behind that of pectoral ; the spine is of moderate size and strength, a little more than half the length of the first ray, which is produced about one-eighth of an inch at its tip ; the other rays decrease gradually in height. The scales are of moderate size, very finely crenated, with the margin convex. The lateral line is parallel with the curve of the back. Scales silvery, with purple reflexions ; membrane between the scales brown ; fins blackish. 34. H.EMULON BREVmOSTRUM. D. ^- 8^- L, lat. 50. L. transv. 5 14. This species is closely allied to //. chromis and H. canna, differing from both by its much shorter and more convex snout. The height of the body is contained twice and two-thirds in the total length (without caudal), the length of the head thrice and one-fourth. The snout is short, not much longer than the diameter of the eye, which is more than one- fourth of the length of the head. Cleft of the mouth rather wide, the maxillary extending beyond the vertical from the front margin of the eye. Hind margin of the praeoperculum slightly emarginate, its angle with more con- spicuous denticulations. Dorsal fin notched, ^^ ^ with strong spines ; the fourth is the longest, ((w ^ ^ half as long as the head. Caudal fin forked. DR. GtJNTHBR ON THE FISHES OF CENTRAL AMERICA. 419 The second anal spine is strong, and somewhat longer than the tliird, hut not quite as long as the fourth of the dorsal fin. Scales above the pectoral fin not conspicuously larger than the others. More or less conspicuous oblique brown streaks run along the series of scales, and are broken up into series of spots in larger examples. A vertical black spot covered by the angle of tlie procoperculum. We possess four examples of this species : three were collected by ('apt. Dow at Panama; and the fourth is from Puerto CabcUo. The largest is 8 inches long. 35. H^MULOX MARGARITIFERUM. (PI. LXV. fig. 2.) Ouiith. Proc. Zool. Soc. 1864, p. 147. D. jl A. l\. L. lat. 55. I-. transv. 6 15. T])e height of the body is one-third of the total length (witliout caudal), the length of tlic head two-sevenths. The diameter of the eye is two-sevenths of the latter, and equal to the extent of the snout and to the width of the interorbital space, which is very convex. The maxillary extends beycmd the vertical from the anterior margin of the eye. Prseojierculum emarginate behind. Dorsal fin scarcely notched, with the soft portion very low ; its spines are moderately strong, the fourth is the longest, not quite half as long as the head. Anal spines strong ; the second is longer and stronger than the third, and equal to the eighth of the dorsal. The soft vertical fins enveloped in scales ; caudal forked, with the upper lobe longest. The pectoral fin does not extend to the vent. Greenish olive above, each scale with a pearl-coloured centre ; sides silvery ; a blackish spot above the axil. One specimen, 12 inches long, was obtained by Messrs. Dow and Sahin on the Pacific coast of Panama. 39. Ch.ETODOX HUMERALIS. (PI. LXV. fig. o.) (jiiiith. Fish. ii. p. 19. I have given a full description of this species (/. c). The Pacific coast of Central America appears to be its true home. Messrs. Salvin and Dow collected tliree speci- mens at Panama ; and our other specimens, which we received from the Haslar Collec- ti(m, are probably from Guatemala, from wliich country Sir J. Richardson, as we know, obtained a collection of fishes. I have no doubt that the statement of this species extending to the Sandwich Islands is correct. The Panama examples differ from the typical specimens only in ha\ing an additional black cross band near the hind margin of the caudal fin. 41. POMACANTIirs ZOU'KCTUS. Pomaranthodes zonipectus, Ciill, Proc. Ac. Nat. Sc. Pliilad. IHO.'J, p. '211. ■'The form much resembles that of Fomaconthus. The greatest height equals three- 3 M 2 420 DE. GiJNTHER ON THE FISHES OF CENTEAL AMEEICA. fifths of the length. The head forms about a quarter, and the caudal fin about a sixth of the total length. . . . The dorsal is considerably produced at the sixth ray, which passes behind the rounded posterior margin. . . . Brownish, margined with light on each scale. A very dark brown band girdles the breast behind the pectoral and ventral fins." . . . Obtained by Capt. Dow at San Salvador. 43. Upeneus tetraspilus. (PI. LXVI. fig. 1.) Giinth. Proc. Zool. Soc. 1864, p. 148. D. 8|9. A. 7. L. lat. 33. L. transv. 2 6. The height of the body equals the length of the head, and is contained thrice and two-fifths in the total (without caudal) ; the width of the interorbital space is two- thirds of tlie length of the snout. Teeth in both jaws in two series, the outer series of the upper jaw being formed by very obtuse and partly confluent teeth. The maxillary is dilated and rounded behind, and bent upwards into a sort of hook ; the barbels extend to the vertical from the root of the pectoral. The third and fourth dorsal spines are subequal in length, longer than the second, and nearly three-fourths of the length of the liead. Greenish olive above, each scale above and below the lateral line with a large pearl-coloured spot ; sides yellow ; a rose-coloured band on each side of the belly. A large blackish blotch on the lateral line, behind the hind part of the spinous dorsal fin. A second smaller blackish spot behind the orbit ; the latter is sometimes very indistinct. Two specimens, 8^ inches long, were collected by Messrs. Dow and Salvin on the Pacific coast of Panama. This species would belong to the division which has been called Mulloides. 44. Upexeus graxdisquamis. Gill, Proc. Acad. Nat. Sc. Pliilad. 18G3, p. 168. This species, which belongs to Bleeker's division Upeneiis, is described thus : — D. S|g. A. 7. L. lat. 30. L. transv. 2i/5. The greatest height is contained four times in the length to the end of the median caudal rays, and four times and a half in the total. The head equals the height, and is itself longer than high, the profile in front of tlie eyes rapidly declines downwards, and is nearly rectilinear. The diameter of the eye enters tlirice and a half in the head's length, and the height of the preorbitar twice and three-fourths. The supramaxillar ends at the vertical from the front of the eye. Tlie teeth in front of the upper jaw are biserial ; below uniserial. The first dorsal fin is highest at the third spine, and there equals the head in front of the prseopercular margin ; the first is exceedingly short, and the second and fourth nearly equal, little shorter than the third ; all the spines are very slender towards the ends. The distance of the second from the first dorsal enters once DR. GiJJNTHER ON THE FISHKS OF CENTRAL AMERICA. 421 and three-fourths in the base of tlie former, and in that interval are three scales ; its length is less than that of the first. The ventral equals the distance of the hinder margin of the orbit from the snout. The tubes of the lateral line have slender branches diverging from them, generally directed obliquely upwards. The larger scales have six radiating strise. The colour is light greenish brown above, with an indistinct silvery spot at the centre of each scale. Below the lateral line, especially between it and the anal tin, the colour is rose. The dorsal fins covered with spots of the colour of the back. The others are immaculate. Two specimens, the longest of which is 7^ inches long, were collected by C'apt. Dow on the Pacific coast of Central America. 47. Chrvsophrys calamus. A fine example, 16 inches long, has lately been sent by C'apt. Dow from Panama. 49. ClRRHITlCHTHYS RIVULATUS. (Plate I.XXXVL fig. 4.) Cirrhitts rivulatiw, Valeiic. Voy. Venus, Poiss. p. 309, pi. 3. fig. 1 (bad) . D. [j. A. g. L. lat. 47. L. transv. 6 14. The height of the body is contained thrice in the total length (without caudal), the length of the head twice and two-thirds. The snout is of moderate extent, compressed and rather elevated ; the maxillary extends beyond the front margin of the eye. Inter- orbital space deeply concave, and half as wide as the orbit ; a low longitudinal median crest ou the crown of the head. Pra;operculum finely serrated behind. The fourtli, fifth, and sixth dorsal spines are the longest, two-sevenths of tlie length of the head, all are of moderate strength. Seven simple pectoral rays, none of which extend so far backwards as the ventral fin. The second anal spine is longer, but scarcely stronger, than the third. Brownish, with transverse dark brown bands and spots, all of which are edged Avith light blue. There are two of these bands on the head crossing the praioperculum ; five on the body and tail, composed of large, more or less confluent, round spots ; especially the third and fourth terminate above each in a pair of large spots, the first pair occupying the end of the spinous and commencement of the soft dorsal, the second the basal portion of the end of the soft dorsal. Caudal and anal fins with similar ocellated spots ; a brown band across the inner side of the root of the pectoral. A single example of this beautiful species, 6 inches hmg, was obtained by Capt. Dow at Panama. The typical specimen was obtained at the Galapagos Islands. 51. POLYNE-MUS 3IEL.1X0P0MA. (iiiiitli. I'roc. Zool. Soe. 1864, p. 148. D. 7|y2- -'^- js- L. lat. 73. Nine free pectoral appendages, the longest of which extends to the vent. Praopcr- 422 DR. GUNTHBR ON THE FISHES OF CENTRAL AMERICA. culuni finely serrated, with a small spine above the angle. The vomerine teeth form a rounded patch ; the band of the palatine teeth is as broad anteriorly as the front part of the intermaxillary band. Operculum black. A single specimen, 15 inches long, was obtained by Mr. Salvin at San Jose. Description. — This fish is elongated in form, its greatest height being contained five times and a half in the total length, with the caudal, and four times and one-sixth without it. The tail is compressed, its height above the end of the anal fin being half the length of the head. The head is much longer than high, and is containii'd four times and two-thirds in the total length with the caudal, and thrice and one-third with- out it ; its width between the eyes is two-ninths of its length. Snout produced beyond the mouth, obtusely conical, and shorter than the diameter of the eye, which is con- tained five times and a half in the length of the head. The cleft of the mouth is situated on the inferior side of the head, it is extremely wide, the maxillary being more than half the length of the head. The posterior margin of the pra?operculum is finely seirated ; the ang|e is produced, forming a rounded membranaceous lobe. The posterior margin of the opercular apparatus is membranaceous, rounded, and formed by the oper- culum and suboperculum. The origin of the fii-st dorsal is in the vertical from the ninth scale of the lateral line, or from a point about midway between the pectoral and ventral fins. The first spine is minute, the second is the strongest, all the others being flexible ; the third is the longest, contained once and two-thirds in the length of the head ; the fourth is longer than the second, and the following rapidly decrease in length. A series of scales ascends behind the second, third, and fourth spines, but disappears at the fifth ; the distance between the two dorsals equals the length of the base of the second, which is entirely covered with scales and has the ujjper edge strongly emarginate ; the second ray is the longest, nearly as high as the spinous dorsal, and twice the height of the last ray. The distance betAveen this fin and the caudal is one-fourth of the total length (without caudal). The caudal fin is completely covered with scales, deeply forked, with the lobes pointed, the upper one being slightly the longer, and one-fourth of the total length. The distance between the anal and caudal fins is less than that between the caudal and dorsal, as the termination of the anal falls behind that of the dorsal, and in the vertical from the 52nd scale of the lateral line. It is entirely covered with scales ; and its origin corresponds to that of the seventh ray of the dorsal ; its lower edge is emarginate ; the first spine is very small, the second being only one-third the length of the first ray ; the first and second rays are the longest, and about thrice the length of the thirteenth or final ray, which, however, is rather longer than the one which pre- cedes it. The pectoral is nearly one-sixth of the total length ; its root is covered with minute transparent scales. The free pectoral appendages are long, the third and fourth being the longest, considerably longer than the pectoral fin, and reaching to the vent ; the fourth is one-eighth of an inch longer than the head. The root of the ventral fin falls behind that of the pectoral, and in a vertical from the twelfth scale of the lateral DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 423 line ; it is short, one-eighth of the total lengtli, and does not quite reach the vent ; its spine is ahout one-half the length of the adjacent ray. The scales are of moderate size, longer than high, and have the posterior margin minutely crenulated. Lateral line straight, very slightly bifurcated between the lobes of the caudal. The teeth are minute and villiform, those of the vomer form a rounded or nearly square patch ; the band on the palatines cuneiform and elongated, broadest anteriorly. The body is uniform silvery, greenish grey, darker on the back ; the fins are minutely dotted with black, the dorsals becoming blackisli at their margins. Operculum black. •32. TOLY-NEMUS APPROXI.\IANS. Pulyneiiius upproa^imans, Lay & Bciiu. in Beechey's Voy. Zool. Fisli. p. 57. Trichidion approximans, (iill, Proc. Ac. Nat. Sc. Philad. 18(;.'5, p. 1G9. I). 7 I ^3. A. jl. T. lat. GO. Six pectoral appendages, the longest of which reaches to the commencement of the anal fin. The length of the caudal lobes is rather more tlian one-fourth of the total length. Pectoral fins blackisli. Description. — This fish is moderately elongate in form; its greatest height, whicli is between the root of the second dorsal and anal fins, is contained four times and one-third in the total length with the caudal, and thrice and one-fourth in the same without caudal. The tail is compressed, its height above the end of the anal being contained seven times and one-third in the total length. The head is much longer than high ; its length is about four times and a half in the total with, and thrice and a half without caudal; its width between the eyes is nearly one fourth of its lengtli. Tlie snout is produced, obtusely conical, and sliorter than the diameter of the eye, whicli is one-fifth of the length of the head. The cleft of the mouth is situated at the inferior side of the head, as usual ; it is wide ; the maxillary reaching considerably beliind the orbit, but the length of the bone is only two-fifths of that of the head. The posterior margin of the praeoperculum is armed with a distinct serrature, and one or two more distinct teeth above the projecting membranaceous lobe of the angle. 'l"he posterior extremity of the opercular apparatus is angular, membranaceous, and formed by the operculum and suboperculum. The origin of the first dorsal is opposite to the eighth scale of the lateral line, and in the vertical between the roots of the pectoral and ventral fins. The first spine is minute, the second shorter than the third, whicli is the longest, and con- tained about once and one-third in the length of the head ; the fourth is longer tlian tlie second ; and the subsequent spines rapidly decrease in length, rendering the upper margin almost vertical. There is a series of scales behind each spine almost to the top. The distance between the two dorsals is more than the length of the base of the second, which is entirely covered with scales and has the upper margin emarginate ; the first and second rays are the longest, not so higli as the spinous dorsal, more tlian twice as long as the hindmost ravs. The distance between this fin and the caudal is 424 DR. GiJNTHEE ON THE FISHES OF CENTEAL AMERICA. one-fifth of the total length. The caudal tin is completely covered with scales, deeply forked, with the lobes pointed, the upper one being rather the longer. The distance between the caudal and anal fins is less than that between the dorsal and caudal, as the extremity of the anal falls behind that of the dorsal, or in the vertical from the forty- third scale of the lateral line. Its origin corresponds to that of the dorsal ; and it has the lower edge straight or very slightly emarginate ; it is entirely covered with scales. The first two spines are very small, and the third not half the length of the first ray ; the first and second rays are the longest, but not twice as long as the fifteeenth or terminal ray. The length of the pectoral is not one-fourth of the total ; it has minute scales towards the base. The free pectoral appendages are six in number ; the upper one is the longest, reaching to the anal fin, and is not quite one-third of the total length. The root of the ventral falls a little behind the middle of the pectoral, and in the vertical from the eleventh scale of the lateral line; it is short, one-eighth of the total length, reaching to the vent ; its spine is more than half the length of the adjacent ray. The scales are of moderate size, scai'cely higher than long, and minutely ciliated on the posterior margin. The lateral line is straight, bifurcating between the lobes of the caudal. Teeth on the vomer in a narrow transverse patch. Two specimens, 12 inches long, are in the Collection, one found by Mr. .Salvin at C'hiapam, the other by Capt. Dow at Panama. Mr. Gill first recognized this species, which is not identical with P. aant/tonemus, as suggested in the ' Catal. of Fishes.' 5-3. POLYNEMUS OPERCULARIS. Trichidion opercularis, Gill, Proc. Acad. Nat. Sc. Pliilad. 1803, p. 169. This fish is described thus : — D. 8j ^. A. ^. L. lat. 69-70. L. transv. 8 14. The greatest height equals a fourth of the length to the fork of the caudal fin, and more than a fifth of the extreme, while the head enters four times and a half in the latter. The outline from the dorsal to the snout is nearly rectilinear and little declined. The distance of the anal from the outer axil of the ventral equals that of the posterior nostril from the margin of operculum. The first dorsal, when bent backwards, rests on the fourth scale, in front of the second. The second commences nearly above the twentieth scale of the lateral line. The pectoral is as long as the head behind the pupil. There are eight pectoral filaments, the longest of which extends rather beyond the front of the second dorsal. T'he colour is greenish brown above and yellowish green below. The operculum is blackish. The first dorsal and the pectorals, except below, are also blackish, as is likewise the margin of the caudal. The anal is tinged with orange. A single specimen, 1 1 inches long, was collected by Capt. Dow at Panama. DE. GTTNTHEE ON THE FL'^HES OF CENTRAL AMERICA. 425 54. Larimus breviceps. Larimus hreviceps, Ciiv. & Val. v. p. 116, pi. 11 J. Giinth. l'"i.sli. ii. p. 268. Amhiyscion argenteus. Gill, Proc. Ac. Nat. Sc. Piiilad. 18()1, p. Ki"). Ha^^ng recently received a fine example of this fish from Panama through Capt. Dow, I have convinced myself that the Pacific examples are not specifically, much less generically, distinct from West-Indian ones. 56. MiCROPOGOiV ALTIPINNIS. Gunth. Proc. Zool. Soc. 18fi4, p. 149. D. lOi.j^i. A. 2/7. L. lat. 48-50. L. transv. T'ln. The height of the body is contained thrice and two-thirds in the total length (without caudal), the length of the head thrice and a half The maxillary extends scarcely beyond the vertical from the anterior margin of the eye. A series of five minute barbels along each side of the mental groove. Two .short, strong, divergent spines at the angle of the prEeoperculum. The third and fourth dorsal spines are long, their length being three-fifths of that of the head ; anal spine of moderate strength, not quite one-fourth of the length of the head. Nearly uniform silvery. Two specimens were procured by Mr. Sahin — one, 17 inches long, at Chiapam. and another, 14 inches long, at San Jose; a third specimen, A\ inches long, was found by Capt. Dow at Panama : this agrees in every other respect with the older examples, but of the minute barbels only a trace of the anterior (longest) pair is visible ; so that it appears that this generic (!) character is developed with age. 57. Umbrika elongata. (PI. LXIV. fig. 2.) Giinth. Proc. Zool. Soc. 18(51, p. 148. D. 1()|^. A. 1/7. L. lat. 70. L. transv. 7/22. The height of the body is contained four times and a third in the total length (witliout caudal), and five times if the caudal is included ; the length of the head is two-sevenths of the total, or one-fourth if the caudal is included. The depth of the head is ccmtainod once and three-fourths in its length. Snout long; the diameter of the eye is two-fiftlis of the length of the snout, and one-fourth of the postorbital part of the head. Sym physial barbel very short, as long as the posterior nostril. Pncoperculum without distinct serrature. The length of the second dorsal spine is one-half of that of the head. Posterior margin of the caudal /-shaped, the upper lobe being pointed, the lower rounded ; anal spine ^ ery feeble. The maxillary extends to the vertical from the anterior margin of the orbit. Upper parts blackish, shining silvery, the lower white. One specimen, 17 inches long, was found by Mr. Salvin at Chiapam. -PART VII. 3 N 426 DK. GUNTHER ON THE FI.SHES OF CENTRAL AMERICA. 58. Umbrixa x.\sus. D lOl^L. A. 18. L. lat. 54. L. transv. G U. The height of the body is contained four times in the total length (without caudal), the length of the head thrice and one-fourth. Snout much produced beyond the mouth, which is quite at the lower side of the snout. The diameter of the eye is two- thirds of the length of the snout, and two-fifths of that of the postorbital portion of the head. Symphysial barbel very short, scarcely as long as the posterior nostril. Prseoperculum distinctly serrated. The second and third dorsal spines are as long as the head, without snout. Posterior margin of the caudal finy-shaped, the upper lobe being pointed, the lower rounded; anal spine very feeble. The maxillary extends to below the centre of the orbit. Silvery, fins blackish. One specimen, 10 inches long, was found by Capt. Dow at Panama. D. 10 59. Umbrina analis. A. 2/6. L. lat. 46-48. L. transv. 615. The height of the body is one-third of the total length (without caudal), the length of the head two-sevenths. Snout compressed, rather deep, of moderate extent, longer than the eye, which is two-ninths of the length of the head, and equal to the width of the interorbital space. Snout overlapping the mouth, but not much protruding beyond it. Barbel very short, scarcely as long as the posterior nostril. Prseoperculum distinctly serrated. The second and third dorsal spines are not quite as long as the head without snout. Caudal fin subtruncate. Anal spine very strong, more than half as long as the head. The maxillary extends beyond the front margin of the eye. An oblique dark streak runs along each series of scales. The spinous dorsal fin blackish. One specimen, 11 inches long, was found by Capt. Dow at Panama. DR. GUNTHER ON THE FISHES OF CENTRAL AMERICA. 427 1 thought it possible for sonic time that this fish might be identical with Umhrino iindulata of Girard ; however, as this writer states that the anal spines of U. undulata are feeble, and gives 1/9 for the number of anat rays, we are not justified in identifying these two species. 61. CORVLXA f;HRYSOLKUC.\. (PI. LXVII. fig. 1.) Allied to C. ronchus. 1). 10:^3. A. ^^. L. lat. 55-50. L. transv. ■^. The height of the body is contained thrice in the total length (without caudal), the length of the head thrice and one-third. Head thick ; snout obtuse, with the upper jaw sliglitly overlapping the lower, as long as the diameter of the eye, which is con- tained four times and two-thirds in the length of the head. The maxillary is nearly entirely hidden by the prseorbital, and extends beyond the vertical from the centre of the orbit. Teeth of the outer sei-ies of the upper jaw rather stronger than the others. Interorbital space slightly convex, only one-third wider than the orbit, its width being two-sevenths of the length of the head. Pi-aeoperculum with spinous teeth round its margin, three on and below the angle being much stronger than the others. Supra- scapular denticulated. The second dorsal spine is the strongest, and the third the longest, being as long as the postorbital portion of the head. The second anal spine is very strong, as long as the longest of the spinous dorsal, and not much shorter than the first anal ray. Caudal fin irregularly rounded. Silvery, irregularly mottled with large brownish patches shining golden. A young specimen (5 inches long) is more uniform silvery. Two specimens, 9 inches long, were collected by Capt. Dow at Panama. I have observed in this species a most extraordinary variation in the size of the scales above the lateral line, such as I do not recollect having seen in other Acanthopterygian fishes. The two larger specimens are of nearly the same size ; yet the dorsal scales of one arc only lialf the size of those of the other. The lateral line is composed of nearly the same number of scales in both, and also the scales below the lateral are of nearly the same size. 62. CORVI.NA VKRMICULARIS. (PI. LXVII. fig. 2.) D. 10 1 1. A. l L. transv. ^. The height of the body is a little more than one-third of the total length (without caudal) ; the length of the head two-sevenths. Head moderately compressed, snout obtuse, with the upper jaw overlapping the lower, a little longer than the diameter of the eye, which is one-fifth of the length of the head. The maxillary is entirely hidden by the pneorbital, and extends somewhat beyond the vertical from the centre of the orbit. Teeth of the outer series of the upper jaw rather stronger than the others. Interorbital space convex, only one-fourth wider than the orbit, its width being one- 3 X 2 42S DR. aUNTHER ON THE FISHES OF CENTRAL AMERICA. fourtli of the length of the head. PrEeopercukim with spinous teeth round its margin ; they are rather widely set and of equally small size. Suprascapular scarcely denticu- lated. The second dorsal spine is scarcely stronger than, and but half as long as, the third, the length of which exceeds somewhat that of the postorbital portion of the head. The second anal spine is very strong, rather shorter than the succeeding ray. and equal in length to the postorbital portion of the head. Caudal fin rounded, with the upper lobe slightly produced. Scales irregularly arranged. Purplish shining silvery ; a purplish brown streak, obliquely ascending backwards, follows the middle of each series of scales. Fins bnnvn. A single specimen. 8 inches long, was found by ('apt. Dow at Panama. 63. CORVIN.V ARMATA. Bairdiella nnmita, Gill, Proc. Acad. Nat. Se. Philad. 186;3. p. I(i4. This species, which is evidently allied to C. ronchus, is described thus : — D. 10 1 1 A. =. L. lat. 51. L. transv. 7 TO. The height equals a fourth of the total length, of wliicli tlie liead forms a fourth. The caudal fin equals the head behind the front margin of the eye. The diameter of the eye enters four times and a half in the head's length, somewhat exceeds the inter- orbital area, which is scarcely convex, and equals the snout. The fourth dorsal spine is longest, and nearly equals half the head's length ; all are stout and robust. The second dorsal commences nearly above the twentieth scale of the lateral line, or tip of pectoral. The second anal spine is very strong, longer than the first ray, and nearly equals the interval between the front of orbit and opercular fiap ; the soft fin behind is incurved. The pectoral equals the interval between the middle of the pupil and the opercular flap, and the ventral that between the front of the pupil and the same. The colour is hoary above, silvery below ; the fins yellowish ; the vertical, especially the first dorsal, clouded with darker. P'ound by Capt. Dow at Panama. 64. CORVINA OPHIOSCION. Ophioscion typicujf, Gill, Proc. Acad. Nat. Sc. Philad. 1863, p. 1(51. D. 10 j 22123- -•^- f I^- Ifit- ■^9- L. transv. j^. The height of the body is nearly equal to the length of the head, and two-sevenths of the total (without caudal). Head rather low, snout obtuse, but prominent, with the upper jaw projecting beyond the lower, the cleft of the mouth being quite at the lower side of the snout. The diameter of the eye equals the extent of the snout, and is two- ninths of the length of the head. The maxillary is entirely hidden by the prieorbital, and extends to below the middle of the orbit. Teeth of the outer series of the upper jaw rather stronger than the others. Interorbital space scarcely convex, only one-third DE, GUNTHEK ON THE FISHES OF CENTRAL AMERICA. 420 wider than the orbit, its width being two-sovcntlis of the length of the liead. Vrx- operculum with straight, widely-set, spinous teeth round its margin, those on or mar the angle being slightly the strongest. The second dorsal spine is the strongest, the third the longest, its length being more than that of the postorbital portion of the head. The seccjnd anal spine is exceedingly strong, about as long as the third dorsal spine, or as the first anal ray. (.'audal fin irregularly rounded. Uniform silvery ; top of the spinous dorsal blackish. This species appears to be scarce at Panama. C'apt. Dow having collected only two examples, of 8 and 6 inches in length. 65. Otolithus squamipinnis. D. 8|2,\3.,. A. ^j. L. lat. 85. Scales rather irregularly arranged ; there are nine longitudinal series between the origin of the first dorsal fin and tlie lateral line, and five or six between the end of the second dorsal fin and the lateral line. The height of the body is contained four times and one-sixth in the total length (without caudal), the length of the head thrice and one-fourth. T/Ower jaw very prominent, the extent of the snout being contained thrice and one-tliird in the length of the head. The width of the interorbital space is more than the diameter of the eye, and equals the extent of the upper jaw from the orbit. The maxillary extends to the vertical from the hind margin of the orbit. Prttopercular angle slightly produced, dilated into a membranaceous margin which is faintly striated. The spinous dorsal is longer than high ; the spines are feeble, the length of the third being two-fifths of that of the head. Caudal fin rounded, the middle rays being the longest. The membrane of the soft dorsal and anal fins is covered with small, transpa- rent scales, which form a thickish cover on the base of these fins. The length of the pectoral is three-fifths of that of tlie head. Body uniformly coloured, scales on the sides minutely punctulated with brf)wn ; liinder side of the axil of the pectoral brown. Inner membrane of the gill-cover black. Ventral yellowisli. Two specimens, 10 & 11 inches long, were collected by Capt. Dow at Panama. 6G. OtOUTUUS ALliUS. Giiuth. Proc. Zuol. Soc. I SCI, p. U9. D. 10 .;^. A. 2 9. Scales rather irregularly arranged ; there are seven series between the origin of tlie dorsal fin and the lateral line. The height of the body is one-fourth of the total length (without caudal), the length of the head two-sevenths. The extent of the snout is one- fourth of the length of the head ; the maxillary extends somewhat beyond the vertical from the posterior margin of the eye. Proeopercular angle not produced behind. The spinous dorsal is much longer than high ; its spines are feeble, the length of the fourth 430 DE. GUNTHER OX THE FISHES OF CEMKAL AMEWCA. being two-fifths of that of the head. Caudal fin rounded, witli the middle rays pro- duced. The second anal spine is truly spinous, not flexible, two-fifths of the length of the first soft ray. INIembrane of the dorsal and anal fins not scaly. Tlie pectoral fin extends as far backwards as the ventral, being more tlian half as long as the head. Immaculate, silvery, back greenish. (Pseudobranchiiie present.) One specimen, 14^ inches long, was obtained by Mr. Salvin at C'hiapam. 67. OtOLITHUS RETICULATUS. (nuith. Froc. Zool. Soc. 1864, p. U9. 1). IOI^tjjVt- a. 11(2.9). Closely allied to 0. caroUnensis. Scales rather irregularly arranged ; there are nine series between the origin of the dorsal fin and the lateral line. The height of the body is contained four times and a third in the total length (without caudal) ; the length of the head thrice and a third. The extent of the snout is two-sevenths of the length of the head ; the maxillary does not extend backwards to the vertical from the posterior margin of the eye ; prseopercular angle somewhat produced behind, membranaceous, striated ; the posterior margin of the prscoperculum obliquely descending backwards. The spinous dorsal is much longer than high ; its spines are feeble, tlie fourth being the longest, two-fifths of the length of the head. Caudal fin subtruncated, the middle rays somewhat produced. The first anal ray is quite rudimentary ; the second as long as the eye, flexible, scarcely spinous. The pectoral fin extends as far backwards as the ventral, being more than half as long as the head. Back and sides with an irregular network of brown undulated streaks ; fins immaculate. Two specimens were collected by Mr. Salvin — one, 15 inches long, at San Jose, the other, 13 inches long, at Chiapam. 71. Caraxx leucurus. Giiuth. Proc. Zool. Soc. 1864, p. 24. ^'ery closely allied to C. bicolor. The first dorsal fin is composed of short, stoutish spines, the fourth of which is the longest, but scarcely longer than the eye. The soft dorsal and anal are rather elevated ; the caudal is eraarginate, and has the lobes rounded. Teeth very small, forming a single series in both jaws ; palate smooth. The height of the body is one-half of the total length (without caudal), the length of the head one-third. Snout rather obtuse, the jaws being equal in front when the mouth is closed ; the maxillary extends to below the anterior margin of the orbit. The lateral line makes anteriorly a subsemi- circular curve, the width of which is contained from once and two-thirds to once and four-fifths in the length of the straight portion ; it becomes straight behind the vertical DR. GUNTHER OX THE FISHES OF CEXTKAI, AMKRKA. 431 from the orijifin of the second dorsal, and is armed witli about fifty small and low shields, only a few of which terminate in a depressed spine. The i)ectoral fin extends to the anal spines. Brownish grey, body with six dark brown vertical bands ; the first crosses the body behind the base of the pectoral, and the fourth descends from the middle of the soft dorsal fin. Operculum with a largo black spot. Dorsal, anal, and ventral black ; pectoral and caudal whitish. Only two examples, three inches long, were found by Capt. Dow at Panama. 72. Caranx si'Eciosu.'; (Forsk.). Having examined specimens from Panama, collected by Mr. Salvin, and compared them with otliers from Borneo, Madras, Zanzibar, &c., I have convinced myself that (J. panamensis, Gill, Proc. Acad. Nat. Sc. Philad. 1S63, p. 1G6, is identical with C. speciosns. 74. C.vR.wx iiiri'os. L. We have received two examples frcnn the Pacific coast of Panama from Capt. Dow. The younger one, vvhich is 5 inches long, agrees in every point, especially in the height of the body, with Atlantic examples of this species, whilst the older, 10 inches, is identical with that remarkable form described by Mr. Gill as Varan fjiis )iiar(/inafi(fi (Proc. Acad. Nat. 8c. Philad. 1863, p. 1G6). This is considerably lower in form than the type, the height of the body being only two-sevenths of the total length ; but ha\iug had an o})portunity of comparing the example first mentioned, I do not think it entitled to specific rank, but regard it merely as a variety. The formula of fin rays in our example is D. 7 U,. A. 2 I — . 75. Caraxx caballus. Trachunts boops, Girard, U. S. Pac. R. R. Expetl. Fish. p. 108; (ninth. Fish. ii. p. 422 (not C. boops, C. & v.). D. 7|5^,. A.2|,^. L.lat. 37. The teeth of the upper jaw form a villiform band, those of the outer series being a little the larger ; tho.se of the lower are in a single series ; teeth on the vomer, the palatines, and the tongue. The heiglit of the body is two-sevenths of the total length (without caudal), the length of the head rather more than one-fourth. Eye with a broad adipose membrane in front and behind. Breast scaly. The lateral line is curved, the width of the arch being one-half of the length of the straight porticm ; the latter commences in the vertical from the third dorsal ray ; the plates commence from the beginning of the straight portion of the lateral line, and are well developed. Lower jaw projecting beyond the upper ; maxillary extending to below the anterior rim of the pupil. Pectoral reaching beyond the anterior anal rays. A black ojjcrcular spot. Two specimens were collected by Capt. Dow at Panama : the species extends north- wards to the coast of California. 432 DE. GtJNTHEE ON THE FISHES OF CENTEAL AMEEICA. 76. Caranx caninus. D. 8|^. A. 2|1. L. lat. 24. The teeth in the upper jaw form a villiform band, those of the outer series being much the stronger, and widely set. Lower jaw with a single series of rather strong, closely set teeth, and with the two anterior ones somewhat enlarged, canine-like ; teeth on the vomer, the palatines, and the tongue. The height of the body is a little more than the length of the head, and one-third of the total (without caudal). Snout obtuse, as long as the diameter of the eye ; eye with an adipose eyelid in front and behind. Prseorbital much narrower than the orbit. The maxillary extends beyond the vertical from the centre of the eye. Breast naked ; lateral line slightly bent, the width of the arch being contained once and one-third in the length of the straight portion ; the latter commences in the vertical from tlie fifth dorsal ray ; the plates do not reach forward to the end of the arched portion, and are well developed. Lower jaw scarcely projecting beyond the upper. Dorsal spines rather stout and short ; the fourth is the longest, and one-tliird of the length of the head. Tlie pectoral extends to the fifth anal ray. A black opercular spot. Membrane of the soft dorsal and anal blackish. One specimen, 7^ inches long, was discovered by Capt. Dow at Panama. 77. C'aranx dorsalis. Carangoides dorsalis, Gill, Proc. Acad. Nat. Sc. Philad. 1863, p. 166. D. 4-5 1 4 A. 2 14 L. lat. 25*. The teeth in both jaws form villiform bands ; teeth on the vomer, the palatine bones, and on the tongue. The height of the body is contained once and four-fifths in the total length (without caudal), the length of the head thrice and one-fourth. The first dorsal fin is but little developed, the spines being short, feeble, and flexible. Anterior rays of the dorsal and anal fins prolonged into a very long filament, sometimes longer than the whole body. Jaws equal in length, the maxillary extends to the vertical from the front margin of the orbit, l^ateral line bent, the width of its arch being as long as the straight portion ; the latter commences below the middle of the second dorsal fin. The plates are moderately developed, and commence at some distance from the bend of the lateral line. Gill-membrane above the pectoral blackish ; posterior half of the ventrals black. Panama. We have received two examples from Capt. Dow, one 19 inches long. * Mr. Gill counted 44 ; this is either a mistake, or he has counted small scales not deserving the name of plate-s. DE. GtJNTHEE ON THE FISHES OF CENTEAL AMEEICA. D. 5j ClIORINEMUS ALTUS. The height of the body is contained thrice in tlie total length (without caudal), the length of the head four times and one-fourth. Eye rather large, its diameter being equal to the length of the snout, and one-fourth of that of the head. Lower jaw projecting be- yond the upper. Maxillary very narrow posteriorly, scarcely extending to the ver- tical from the hind margin of the eye ; the length of the intermaxillary is contained once and three-fourths in the length of the head. The infraorbital, situated above the maxillary, is as broad as the bone next above it ; none of these bones reach to the anterior pi-seopercular ridge. Pectoral fin longer than the ventral, nearly as long as the head (with- out snout). Coloration uniform. One example, 11 inches long, has been recently sent by Capt. Dow from Panama. 83. Chokinemus inorxatu.s. Olic/oplites inornatus, Gill, Proc. Acad. Nat. Sc. Philad. 1863, p. 1(5(), D. 5 I ^. A. 2 "The height of the body enters four times and two-thirds in the total length; the length of the head five times and two-thirds. The upper maxillary reaches nearly to the vertical from the hinder margin of the orbit ; the intermaxillary enters t\\ice and one-thu-d in the head's length. The snout is a little longer than the diameter of the eye ; the latter equals a quarter of the head's length. The infraorbital bones do not extend to the prajoperculum ; the one above the maxillary bones is wider than the one above itself, and as wide as that behind the eye. The opercular apparatus is vertical in front of the pectoral, and for an equal space above. The prsoperculum is nearly vertical, and its angle obliquely rounded. The width of the operculum and suboper- culum in front of the lower axilla of the pectoral equals the diameter of the eye and the interval between suboperculum and axil. The pectoral equals the interval between its axis and the hinder border of the pupil ; the ventral is rather shorter, but its end almost or quite reaches to the anus. The colour is uniform, tinged with blue above." One adult specimen was collected by Capt. Dow on the Pacific coast of Central America. VOL. VI. — PART VII. 3o 484 DR. GUNTHER ON THE FISHES OF CFJ^fTRAL AMERICA. 85. Trachy.votus f.\.scutu.s. (Plate LXIX. fig. 4.) Trachynotiisfaiidutus, Gill, Proc. Ac. Nat. Sc. Philad. 1863, p. 86. glaucoUes, Giiiith. Proc. Zoo!. Soc. 1864, p. 150. D. 6|,4. A.2\^. Closely allied to T. glaucus, but witli tlie body more elevated. The height of tiie body is one-lialf of the total lengtli (without caudal) ; the length of the head two- sevenths. The maxillary extends to below the middle of the eye. Anterior dorsal and anterior anal rays, and the caudal lobes, much prolonged, the length of the latter being two-sevenths of the total. The ventral fin does not extend to the vent. Five narrow blackish vertical bars across the lateral line. One specimen, 7 inches long, was obtained by Mr. Salvin at San Jose ; two others, 11 inches long, were obtained by Capt. Dow at Panama. Bescription. — This species has the body (without caudal) of a rhomboidal form, its greatest height Ixnng between the last spine of the dorsal and the \c-nt, and one-half of the total length (without caudal) ; the upper profile between the dorsal and the snout is oblique, feebly convex over the eye. The length of the head is contained thrice and one- half in the total (without caudal). The diameter of the eye is rather more than the length of the snout, and contained thrice and two-tliirds in that of the head. The cleft of the mouth is narrow ; the maxillary reaches nearly to the level of the centre of the diameter of the eye ; its length is a little more than one-third of that of the head. The width of the space between the eyes is more than one-third of the length of the head, or equal to the distance from the tip of the snout to the centre of the eye. Prseoperculum with the hinder margin straight, and at a right angle with the lower border, which is also straight and parallel with the axis of the body. Operculum small, narrow, about two- thirds as long as high ; the hinder border of the opercular apparatus is formed almost entirely of the sub- and interoperculum ; it is rounded and membranaceous : the line of the separation between the operculum and suboperculum is at right angles with that between the sub- and interoperculum. There is a recumbent spine before the com- mencement of the first dorsal, and in a line with the posterior part of the axil of the pectoral ; the dorsal spines, seven in number, are short ; the first is minute, but erect, and not attached by any apparent membrane to the second ; the others show a slight progression in dimensions, and are united by a low membrane. The base of the soft dorsal is not twice as long as tliat of the spinous ; the first two rays, which are the longest, project considerably beyond any of the others, and are equal to half the length of the body (without the caudal) ; the following rays diminish very rapidly in lengtli, and from the eighth ray to the last the fin is scarcely higher than the spinous dorsal, and its upper edge almost straight. The distance between the dorsal fin and the caudal is equal to that between the anal and caudal. The anal fin is preceded by three short spines about equal to the fourth, fifth, and sixth of the dorsal. The base of the DR. GUNTIIER ON THE FISHES OF CENTRAL AMERICA. 4.55 soft portion of the anal is about as long as that of the soft dorsal : it consists of eigliteen rays, and perfectly resembles the soft dorsal in shape, having the first two rays mu(:h longer than the others, equal to the corresponding rays in the dorsal, and the following rays rapidly decreasing in length to the sixth, from which the margin of tlic fin is almost straight. The pectoral fin is pointed, of moderate size, its length being three-fourths that of the head. The ventrals are short, more tlian half the length of the pectoral, and not reaching to the vent. The tail behind the dorsal and anal is compressed and narrow. The caudal is deeply forked ; the lobes are equal, and con- tained thrice and a half in tlie total length ; it is covered with small scales. The body is covered with very minute scales ; those at tlie base of the vertical fins and near the lateral line are a little larger. The head and opercular bones are entirely naked. The lateral line shows a somewhat irregular sinuosity slightly above the median axis of the body for the first half of its length, after which it is perfectly straight, termi- nating between the two lobes of the caudal. Teeth small, villiform ; a small central patch on the vomer, and a narrow one on each of the palatines. Bluish green above, silvery beneath. Five vertical brown stripes down the sides of the body across the lateral line, the first two being nearer together than the others, which are at almost equal distances: the first behind the axil of the pectoral, the second below the third dorsal spine, the third below the sixth, the fourth below the seventh dorsal ray, and the fifth below the seventeenth. However, the second and third of these bands are placed sometimes more backwards, which is evidently the case in the example described by Mr. Gill, and named by him T. fasciatiis. Having recently obtained two examples from Capt. Dow, one of which shows the arrangement of the bands as in T. glaucoides, on one side, and that of T.fasciatus on the other, I cainiot entertain any doubt as to the specific identity of these fishes. 86. Pelajiys sarda, B1. AVe may mention this species here, although it is not contained in any of the collec- tions forming the material for this Memoir, because Dr. Ayres alludes to it in the following manner : — " A species oiPelamys brought to the markets of San Francisco is without ques- tion the P. sarda. The closest examination fails to distinguish it from the Atlantic form. Previous to this time we had no positive knowledge of any fisli in the low latitudes wliich inhabits Californian waters and those of the Atlantic." — Proc. Calif. Acad. 1855, p. 74. 90. B.\TR.\ciius PACinci (Gthr.). In other specimens recently collected by Capt. Dow at Panama, I find tlie membrane at the bottom of the pouch of the axil of the pectoral fin (described in Fish. iii. p. 173) folded and wrinkled, with a great quantity of coagulated mucus between the folds. The same species appears to occur also on the coast of West Africa, a specimen having been lately obtained by Dr. Steindachner, who describes it as B. libenensis (Sitzgsber. Ak. Wiss. Wien, 1867, Iv. p. 525, Taf. 1. figs. 2 & 3). o 2 436 DE. GUNTHER ON THE FISHES OF CENTRAL AMERICA. Th ALASSOPHRYNE '. Thalassophryne, Giintli. Fish. iii. p. 174. Head broad, depressed ; body subcylindrical anteriorly, and compressed posteriorly ; skin naked. Canine teeth none. Operculum with a single spine. The spinous dorsal formed by two spines of moderate length. The opercular and dorsal spines with a canal conducting a poisonous fluid from a sac situated at their base. Gill-opening not very naiTow, not extending to the isthmus. Atlantic coasts of Tropical America. 92. Thalassophryxe MAa^LOSA. (Plate LXVIII. fig. 1.) (Tiiiith. Fish. iii. p. 175. D. 2|in. A. 18. V. ]/2. Brown, marbled with darker ; some round black spots on the pectoral and the side of the body. The general habitus is that of a Batrachus. The head is somew-hat longer than broad, its length being contained thrice and one-third in the total ; it is moderately depressed. The snout is short, obtuse, with the cleft of the mouth ascending obliquely upwards, and with the chin prominent. The maxillary extends to the vertical from the posterior margin of the orbit. The teeth are obtusely conical, standing in single series, except anteriorly in the lower jaw, where they form two series, and in the upper, where they are cardiform, in a narrow band. The eyes are directed upwards and very small, their width being one-half of that of the bony bridge between the orbits. Gill-covers with a single spine ; it is long, slender, cylindrical, like one of the dorsal spines, and has the operculum for its base. Gill-opening not very narrow; it extends from the upper base of the pectoral obliquely downwards and forwards to the level of the inferior base of the pectoral. The two dorsal spines are slender, pungent, about one-third of the length of the head. Dorsal and anal fins terminate immediately before the root of the caudal, the length of which is one-seventh of the total. Pectoral obliquely rounded, extending to the origin of the anal ; ventral rather short, not quite one-half the length of the head, extending to the base of the pectoral. Skin perfectly smooth, with some very short tentacles at the lower jaw. Two short horizontal muciferous channels on the cheek and the lateral line are very distinct ; they are not, as usually, composed of a series of distant pores, but the pores are confluent, forming one continuous groove of a white colour. Other muciferous channels, as for instance along the base of the anal, are composed of separate indistinct pores. Colour brown, marbled with darker ; pectoral fins and sides of the body with some round black spots ; chin and ventrals brownish ; belly white. A single specimen from Puerto Cabello is known. ' Greek denomination for Sea-toad. DR. GtJNTHER ON THE FISHES OF CFJMTRAL AMERICA. 4:57 lines. Total length 54 I-ength of the head 10 Width of the head 14 Depth of the head 10 Diameter of the eye 1 Length of the caudal fin iS of the ventral fin 7 93. THALASSOrHRYNE RETICULATA. (Plate LXVIII. fig. 2.) (iuuth. Proc. Zool. Soc. 1864, pp. 150, 155. D. 2|24. A. 24. V. 1 '2. P. 16. The length of the head is two-sevenths of the total length (without caudal). The teeth on the palate are in a single series, very short, obtuse, incisor-like. Pectoral very large, extending backwards to the sixth anal ray. Head, body, and fins brown, with a network of yellowish lines ; vertical and pectoral fins with a white margin. In other respects this species agrees with T. maculosa ; so that we may refer to the description of tliat species given above. Three specimens were found by Messrs. Dow and Salvin on the Pacific coast of Panama; the largest is 13 inches long. In this species I fii-st observed and closely examined the poison-organ with wliich the fishes of this genus are provided. Its structure is as follows: — 1. Tlie opercular ])aTt- — The operculum is very narrow, vertically styliform, and ver\ mobile; it is armed behind with a spine, 8 lines long in a specimen of 10| inches, and of the same form as the venom-fang of a snake ; it is, however, somewhat less curved, being only slightly bent upwards ; it has a longish slit at the outer side of its extremity, wliich leads into a canal perfectly closed, and running along the wliole length of its Fig. 1. Hindt'i- hall' of the hoad, with the venom-sac of the opercular apparatus in situ.. * Place where the small openin