NOTES OX A COLLECTIOlSr OF FISHES FROM THECOLOKADO BASIN IN ARIZONA. By Charles Henry Gilbert,Professor of Zoology, Leland Standford Junior Unirersiiii,andNorman Bishop Scofield,Assistant, California State Fish Commission. During the latter half of April and the early part of May, 1890, tliesenior author and Mr. A. B. Alexander, of the United States FishCommission, were temj)orarily detached from the "Albatross'' andassigned to special duty iu Arizona. The principal object of theexpedition was to investigate the alleged occurrence of shad in theLower Colorado, where it had been planted several years before by theUnited States Fish Commission. It was soon ascertained that thereports of the capture of shad had been erroneous, the lish in questionbeing the German carp, tlien a stranger in the Colorado River.Repeated trials of the shad net and seine at Yuma, and later in theHorseshoe Bend of the Colorado below Lerdo, Mexico, failed to demon-strate the presence of shad. The river seejns entirely unsuited to it.Collections of the native fishes were made in the Colorado and iu themouth of the Gila River at Yuma; in the Salt River from Tempe tonear the mouth of the Rio Yerde, and in the u[)per course of the RioYerde at Chino. The commoner species are well represented in thecollection, and in addition such desiderata as Phtffoptenis arf/entis.shmiSyMtda fulgida. a new species of Pantosteus, and, most important of all,Tiaroga cobiti.s, known heretofore only from the tyi)es, discovered in1851. We secured all the species credited to the Lower Colorado andGila rivers excepting of Paniosteus elarld and Catostomus insignis.In their excellent historical account and check list of the fishes ofthe Colorado River,^ Evermann and Rutter call attention to the fact thata remarkably high percentage of its fishes are peculiar to the ColoradoBasin. This becomes still more striking when we examine the shortlist (seven species in all) supposed by them to be found in other basinsas well. For of these seven, two must be eliminated: Lepidomeda vit-iata and Cyprlnodon niacularius. L. vittata was, it is true, taken iusouthwestern Arizona by the Death Yalley Expedition, but the river in > Bull. U. S. Fish Commission, XIV, 1894, pp. 475-486.Proceedings U. S. National Museum, Vol. XX?No. 1131. 487 488 mocEEDJxas or the nattoxal mcseim. voi.xxwhich it was found is properly a tributary of the Eio Virgen, thoughnow lost in the desert sands before reaching that stream. C. macula-rliis is known from the Coh)i'ado llivei', and also from springs in thegreat Colorado Desert. iSo iar as Icnown to ns, it occurs only where tlienatural drainage is toward the Colorado River. Of the remaining livespecies, three are of very wide distribution, seemingly able to set atdefiance what are effective barriers to the dispersion of other fishes.By virtue of what special characteristics they accomplish this resultwe do not know. Their testimony must for the present be simplyignored in any discussion of faunal relations. The case is differentwith one of the remaining species, Lenciscus lineatits, which is commonto mountain tributaries of the Colorado and to the Utah Basin. Itsoccurrence in the Colorado is an unexplained anomaly, and contradictsall the other facts, which bespeak a very long period of absolute isolationfor the Colorado Basin and its fauna. The last of the list, Agosia chryso-gasler, was described from the Rio Santa Cruz, Sonora, INIexico, a tril)u-tary of the Gila. In a recent paper on the fresh-water fishes of Sonoraby Rutter ' it is listed, together with PoeciUa occidentalism from the RioSonora and the Yaqui, which flow independently into the (xulf of Cali-fornia. These Gulf streams are practicallj^ unexplored, and may or maynot prove to have had a comparatively recent connection with the LowerColorado.For the sake of completeness we include in this paper two undescribedspecies of marine fishes taken at the mouth of the Colorado. Gillich-iln/s detrusus enters the river, and lives in salt or brackish water. Likeits California congener, G. mirahilis, it sustains immersion in freshwater without apparent inconvenience.Family CATOSTOMIDJE. 1. PANTOSTEUS ARIZONiE Gilbert, new species.(Plate XXXVI.)Four specimens of this species were taken at Tempe, Arizona. It isreadily distinguished from all other species of Panfostcns by the largesize of its scales.Head broad and flat, 4^ in length. Interorbital space flattened orvery slightly concave, 2^% in head. Depth 4|; D. 11 or 12; A. 7;scales 7 or 8?05 to 67?11 to 15. The scales are unusually large infront of dorsal and along back, much smaller along lateral line, andbecome minute on belly. Those along the lateral line grow gradu-ally larger posteriorly. Eye moderate, very high up, posterior, ^ insnout, 6 in head, 2i in interorbital space; orbital ring somewhat raised.Preorbital less than half as wide as long. Snout considerably longerthan rest of head. Isthmus very broad, 2J in head. Greatest depth 'Proc. Cal. Aiad. Sci., 1896, p. 260. NO. 1131. FISHES FROM COLORADO BASIX?O TLBERTJXD SCOFIELD. 489of lieatl li ill its length. Foiitanelle obliterated, the bone covering itvery thin. IMouth very broad, its width contained 3i times in length ofhead. Lips large, papillose. Lower lip broadly V-shaped behind, butslightly incised, a broad but shallow notch at junction of edge of upperand lower lips. Upper edge of dorsal straight; front of dorsal midwaybetween tip of snout and base of caudal; pectorals 1,^ in head; ven-trals li%; anal reaching base of caudal, 1^^ in head; caudal peduncle1^ in head, its least depth 2i in head or slightly more than '1 in its ownlength.A preliminary account of this s])ecies has l>een given by Jordan &Evermann.' Table of mtasuremcnts. 490 PROCEEDINGS OF THE NATIONAL MUSEUM.of the first ray; first dorsal nearer tip of snout than base of caudal;ventrals not reacliing- quite to vent, If in head. Muzzle not ijroject-ing; about six rows of short tliick papilla' on upper lip. the smallestabove; lower lip large, incised to its base, with about twelve rows ofshort thick papilhe, which are quite small posteriorly; distance fromfront of upper lip to back of lower li in snout; jaws with a slight car-tilaginous sheath. Width of preorbital a little less than half its length.Beaching a length of 2 feet. Gila Basin.This species has been recorded from the following places: Kio SauPedro, Gila Basin (type locality), by Baird & Girard ; Fort Thomas, GilaEiver, b}^ Kirsch. Table of measurements. Lengthin inches. NO. 1131. FISHES FROM COLORADO BASIX?GILBERTAND SCOFIELD. 491Specimens from Delta, Colorado, collected by Jordan & Evermaun,differ slightly from Green Eiver specimens in size of scales, as willappear from the following tables of measurements: Measiiremevta of Catostomua dlscoholus, Hreett River, Wyoming. Lengthin inches.! 1 492 PROCEEDINGS OF THE NATIONAL MUSEUM.ments and other data is given below. The scales average smaller thanthose in the type specimen, and the majjority have one more dorsalray. The anal tin is not so deej). In the small specimens \\\e, nuchalhump forms a sharj) keel from dorsal to nape, the keel being but littleelevated. Xiiruvchen itiiconip((}i(jre, described by Jordan & Evermannfrom a single small specimen, agrees with the young of A', cypho exceptin the number of dorsal rays, there being but twelve rays in the dorsalof X. uneompalujre. These two will i)robiibly be found to be the samespecies. MeannyemnitH of Xiirauchcn cjiplio. NO. 1131. FISHES FROM COLORADO BASIN?GILBERT AND SCOFIELI). 493 8. GILA ROBUSTA Baird & Girard.Gila rohusia iJAiiiD & Girard, Proc. Ac. Xsit. Sci. Pbila., 1853, i>. 368. ? Girard,Pac. K. R. Snrv., 1858, X, p. 285.?.Jordan & Gilbert, Syuopsis, 1883, p,228.?Jordan, Bull. U. S. Fish Couuinssiou, IX, 1889, ]>. 27.?Jordan &E\ERMANN, Fishes of N. and ]\1. A., 1896, p. 227.LeuciscHs rohiistiis Gintiier, Cat., 1868, VII, p. 241.Gila puUhella Baird A- Girard, Proc. Ac. Nat. Sci. Phila., 18.54, p. 29.Gila (jrahami Baird tV Girard, Proc. Ac. Nat. Sci, Pbila., 1833, j). 389. Girard,U. S. aud Mex. Bound. Surv., Zool., 1859, p. 61. ? Jordan ?fc Gilrert, Syn-o])sis, 1883, p. 228.Li'iicisciis (/rahami GC'NTHER, Cat., 1868, VII, p. 242.Gila (jriicilis Jordan & Gilbert, Synopsis, 1883, p. 229.Very abundant in Salt Eiver at Tempt^', where it exceeds in iiinnbersall other species. It was not taken in the Colorado, and is probablymore abundant iu smaller streams thaii in the main river channels.Our specimens show a great range of variation in the size of thescales, as can be seen from tlie following table. The two specimenshaving eighty-three and one hundred and ten scales in the lateral lineare probably abnoruml. The larger specimens bear a striking resem-blance to Gila eJegans^ and the younger ones are difficult to distinguishfrom LenciscHS intermedins. Compared with the latter, Gila robustahiisslightly smaller scales and a slenderer caudal peduncle. The body isnot so deep and its head is more slender. The scales below the lateralline are not specked with black.This species is abundant throughout the entire Colorado River Basin.Owing to the close resemblance which this species bears to Leuciscusititermedius, the synonymy of the two has been greatly confused.Measurements of Gila rolmsta, Tempe, Arizona. D. A . 494 rnocKEJUxas of the national museum.Jtf/oma f/ibhom (iiiiARD, Proc. Ac. Nat. Sci. Phila,, 1856, p. 207.?GiRAun, U. S.Mox. Bound. Surv., Zool., 18.59, p. 61.Tifiomn intermedin Girahd, Proc. Ac. Nat. Sci. rbihi., 1856, p. 206.SquaVms iniermedius .Tokdan & Gilbkrt, Synopsis, 1883, p. 238.Leiieiscns iiiiermedhiH Jokdax & Evkhmaxx, Fi.sbcs of N. andM. A., 1896, p 235.Gila vifira Coi'K, Zool. Wht^eler's Expl. W. 100th Mer., V, 1875 (1876), p. 663.Squaliiis nifira Joudan & Gilbert, Synopsis, 1883, p. 239.LexciscKs niger .Jordan &. Evermann, Fishes of N. and M. A., 1896, i>. 235.Squaliiis lemmoni Rosa Smith, Proc. Cal. Ac. Sci., 1884, p. 3.Leiicisciis zunnensis GC'NTHKH, Cat., VII, 1868, p. 241. Substitute for /.. t/rdriUs,preoccupied.Numerous specimens about ;> iuclies in length were obtained atTempe, and at Chiiio, Arizona. It is more robust than the younj>- ofGila robusta, and has the scales a little larger, those below the lateralline specked with black. The specimens taken at Chino diifer from theTempe specimens in the slightly deeper caudal peduncle. This speciesvaries greatly in its scale formula, as can be seen from the accompany-ing table. Like G. elegans and G. rohnsta it is found throughout theColorado River Basin. It has been commonly known as L. nif/e)-, butthere is no reason to consider the two nominal species distinct. iSqua-lius lemmoni is described as having the scales 68, but in one of the typeswe find them 21-75-10. It may therefore well belong here.Tahle of measurements. NO. irn. FISfTES FROM COLORADO BASIN?GILBERT AND SCOFIFLD. 495 111 five specimens examined, four liave the teeth 1, 1-4, 1, and one 2^4_4, 1^ Avitbout grinding surface. Tlie lateral line is median and aboutstraight, with seventy pores to base of caudal fin. The l)elly and theback ill front of the dorsal are destitute of scales. Least depth of cau-dal peduncle 2=^ in head. Mouth very small, terminal, oblique; the lipstlcshy. The maxillary is without barbels and is contained 1.^ insnout. The mandible is contained a little less thau three times in head;premaxillary not protractile. The pectoral tins reach two-thirds dis-tance to veutrals. The ventrals reach the front of the anal. The thirdray of tjie anal is the longest, 2^ in head. The front of the dorsal isslightly behind the origin of the ventrals, and considerably nearer thebase of the caudal than the tip of the snout. The edge of the dorsalfin is straight, its second ray longest, If in head, its rudimentary raysnot enlarged.Cohn- (in alcohol), pale gray or yellowish, mottled with reddish-brownon sides and back: a dark elongate black spot on base of middle cau-dal rays broadening posteriorly into a vertical bar, which follows theposterior outline of the caudal fin; tiiis followed by a second and insome by a third fainter bar witli lighter interspaces; a conspicuouswhite patch above and below caudal spot; there is a small but con-spicuous white spot under the first rays of the dorsal and one underposterior end of dorsal, the two encroaching slightly on the fin; dorsalwith two broad but faint dark bars parallel to its free edge. Lengthaverages 2^ inches. 11. AGOSIA OSCULA Girard.About thirty specimens were obtained at Chino. This species is atonce distinguished from the other sjiecies of A(/osio in the ColoradoBasin by its very small scales and its definite lateral band. We givehere a description based on our specimens, as there is no good currentdescription.Head 4 in length; depth 4 to 4^; eye 4; snout 3J; scales 17 to 19-80to 80-15 to 17; D. 8; A. 7; least depth of caudal peduncle 2^ in head;teeth 1, 4-4, 1, hooked and with grinding surface. Body terete andrather elongate, the caudal peduncle not much compressed. The headta])ers to an elongate but obtuse snout. The mouth is slightly ob-lique; the lii)S not fleshy; maxillary with a small barbel at its tip.Free margin of dorsal and anal straight. The front of the dorsal isslightly behind the origin of the ventrals and midway between thecenter of the orbit and the base of the middle caudal rays. Lengthof dorsal 2 in head; its longest ray 1^ in head. The pectorals areshort, reaching two-thirds distance to ventrals. The veutrals do not(juite reach the front of the anal. J3usky olive above, silvery below; adefinite dark lateral band about width of eye, expanding at the base ofthe caudal and narrowing abruptly to a faint caudal spot. In somespecimens the sides are slightly mottled with darker. Our specimensare about 2 inches in length. 4 DC rnoLEEDixas of the national mcseim.AGOSIA COUESII Yarrow. ^Niiie specnueiis were taken at Tempe, and do not agree perfectlywith current descriptions. The tbllowing description is based on oursiDecimens: Head 3| to 4; depth 4 to o; eye 4 to 4.^ in head, 3 insnout; J). 8; A. 7; scales 14 to 17-70 to 77-10 or 11; teeth 1 or 2,4-4, 2 or ]. r>ody stout, the head long and conical, the snout pointed.Mouth broad, interior, horizontal, the lii)s Heshy. Width of istlinuis10 to V.) times in length of tish. loye small ami high up. Candalpeduncle deep and compressed, its least depth contained 2^ m the head.Fins all large; the pectorals reaching front origin of ventrals; tlie ven-trals reaching past front of anal; free edge of dorsal and anal slightlyconcave; front of the dorsal in advance of origin of ventrals andhalf way between base of middle caudal rays and nostril. The lengthof the dorsal is contained 1 /^o in head ; its longest ray 1 ^ i" head.The longest ray of the anal is contained li in the head. Color darkabove and on sides, mottled slightly with black; pale below; tins allplain; no lateral band or caudal spot. Our largest specimen, 2^ incheslong. Measurements of Ayosia coiiesii. Head. NO. 1131. FlSHKS FROM COLORADO BASIN?GILBERT AND SCOFIELD. 407as head, curved and slightly longer than the second spine, which isreceived into a longitudinal groove in the first; back of these the raysare thickened and ossified for a little over half their length, their tipsarticulated and issuing from the tips of the spines; length of anal1.] ill head, with one rudimentary and ten developed rays; pectoralreaching ventrals, their rays slightly ossified at base; veiitrals reach-ing vent, the first ray thickened and ossified for half its length, theremaining rays developed as six shar}) fiat spines which fokl togetherlike a fan when the fin is closed. From the posterior side of each si)ineand from just below its tip an articulated ray issues, the first extend-ing beyond its spine for one-fourth length of latter, the others success-ively shorter; the last scarcelj^ projecting; osseous portion of last rayjoiiu d for its whole length by a membrane to the abdomen; caudalforked for half its length. .ALouth moderate, horizontal, lower jaw included; maxillary extend-ing to front of orbit, with a small barbel at its tip; length of mandibleequal to distance from tip of snout to center of orbit, the spacebetween them papillose and spongy. Nasals elevated, the muzzleslightly depressed. Lateral line deflected opposite the dorsal, not quitecomplete, about thirty-five pores to opposite front of anal; rudimentsof scales can be seen above lateral line, more numerous in front ofdorsal Color pure silvery, yellowish beneath; dorsal region veryfinely punctulate; peritoneum and gill cavity light silvery. 15. MEDA FULGIDA Girard.This species was found extremely abundant in the upper course ofthe Rio Verde, near Chino, Arizona, and was taken also in the SaltRiver at Tempe. It had previously been taken only in the Rio SanPedro. Following is a description of our specimens:Head 4 in lengtli ; depth oi; eye 3i in head, e(jual to snout and tointerorbital width. Least dei)th of caudal peduncle 3^ in head, equal-ing diameter of eye. D. II, C, counting last divided ray as one; A. 8,9, or 10, usually 9. Front of dorsal behind origin of ventral consider-ably nearer base of caudal than tip of snout. Tlie character of dorsalrays is the same as iu PliKjopterus (oy/entisshnii,s, the first spine curvednearer its tip than in the latter, the second spine shorter than thefirst; first spine IJ m head, longer than base of fin, which is containedtwice in head; anal l-f in head; pectorals reach two thirds distanceto vent; the rays osseous at base; ventrals reaching almost to vent,and structurally the same as in /*. arrolonged into an intromittent organ. Length offemales, 2i inches; males, IJ inches; agreeing in all respects withspecimens from the type locality (Tucson, Arizona). Family GOBTID.E. 18. GILLICHTHYS DETRUSUS Gilbert & Scofield, new species.(Plate XXX VIII.)Allied to (rUlichthys mirabilifi Cooper, (littering in the broader andmore depressed head, the larger anal fin, and the greater distancebetween the two dorsals.Head3i; eye 7; snout 4; depth 5; interorbital 5i; D.VI, 13; A. 11developed rays (10 in G. onirabiU.s). Scales very fine anteriorly, becom-ing much larger posteriorly; about seventy-five oblique rows of scalesfrom base of pectoral to caudal, and about twenty-five longitudinalrwos between front of anal and front of second dorsal.The head is depressed, the frontals broad, tlie least frontal width K0.1131. FISHES FROM COLORADO BASIN?GILBERT AND SCOFIELD. 499being contained in the head eight times (eleven times in 0. mirabilis).The post frontals are small and project but very little, diitering from(r. mirabilis, where the post frontals project into an elevated vring-likeprocess. The width of the isthmns is contained three times in thehead; length of maxillary ]^ times in head; mandible 1?. Least depthof caudal peduncle 2? in head. Distance between dorsals half lengthof base of first dorsal; base of first dorsal 2^ in head; second dorsalli; anal 2 in head; length of longest i)ectoral ray If in head. Colorvery pale olive, some of the sj)ecimeus with dark x)nnctulations aboutthe head and fins. The ]>ale coloration is probably due to their life inshallow water on bottom of pale sand.Several specimens, the longest 5 incljes long, were taken at Horse-shoe Bend, near the mouth of the Colorado Kiver, in Mexico, wherethey are quite abundant. The species inhabits muddy creeks andchannels communicating with the river, and is caught and eaten by theIndians.Type.?:So. 48127, U.S.N.M. Family PLEURONECTID.E. 19. PARALICHTHYS ^STUARIUS Gilbert & Scofield, new species.(Plate XXXIX.)Head 3|; depth 2^; eye oi; interorbital space tlat, 12 in head, halfthe diameter of the eye: maxillary 2 in head, equal to the pectoral tin ; gill rakers 0+20, the longest two-thirds length of eye; dorsal 72 to82; anal 58 to 64. (In the seven specimens the rays are: Dorsal 72,79,81,81,82,83,83; anal 58, GO, GO, 02, 63. 03, 04.) Vertebra 10+28;scales weakly ciliated, with small acces.sorj' scales, 105 in the lateral line.Length of the arch contained four times in straight part of lateral line,2 in head; height of arch 4i in head. Four of the seven specimensare sinistral. Color pale chocolate brown. Specimens small, to9 inches in length. Taken at Shoal Point, at mouth of the ColoradoEiver, Mexico, by the United States Fish Commission steamer Alba-tross. This species is distinguished from, the other members of thegenus by its numerous fin rays and its many gill rakers. It is nearestrelated to ParaUchthys californicKs, which is abundant along the entirecoast of California, extending as far south as Magdalena Bay, in LowerCalifornia. Two specimens in the museum of Stanford University fromthe latter locality are entirely tyjiical of P. californicus.Type.?1^0. 48128, U.S.KM. U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. XX PL. XXXVI U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. XX PL XXXVII U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. XX PL. XXXVIII o U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. XX PL. XXXIX \;- % ^:- .A '?'^4(||||||F"