FISHES OF THE TERTIARY SYSTEM

274 PROCEEDINGS OF THE NATIONAL MUSEUM

Q. FISHES OF THE TERTIARY SYSTEM

A

richassortmentoffossil fishesfromtheGreen River Eoceneof

Wyoming

andother westernlocahtiesiscontahied inthecollection, and

among

the

number

are includedseveral important type speci-

288 PROCEEDINGS OF THE XATlOyAL MUSEUM.

vol.52.

mens.

From

^theMioceneof Florissant,Colorado, quantitiesof

Amy-

zon remains, and from corresponding strata in Esmeralda County, Nevada, large

numbers

of Leuciscus skeletons have been added to the collection. Thereis also an abundantrepresentationofEocene and Miocene ^ichthyicremains from the Atlantic coast region, and fromforeignTertiary horizonsmentionshouldbe

made

of a

number

offineslabsfromthe

Upper

Eoceneof

Monte

Bolca,innorthernItaly.

In the following pages

some

ofCope's typeswhich havenot previously beenfigured receive attention, and two orthreencAV species of Ter- tiaryfishesaredescribed.

Family

OSTEOGLOSSIDAE.

This family,firstappearing in the Eocene, is represented

by

sev- eral

modern

genera, two ofwhich, Osteoglossum^* and Arapaima, ^are fomid in South American ^rivers. Heterotis ^is a tropical African genus.

The

skull in thisgrouphas a distmctly primitive appearance, thesuperficialbones

bemg

^thinlycovered

by

skinand

havmg

asculp- tured surface.

The

wide nasals, frontals, and parietalsmeetin the middlefine,andthe supraoccipitaiscarcelyreaches thesurface.

Both

the premaxillaandmaxillaaretoothed,andsharein

formmg

themar- gin of themouth; there is no supramaxiUa.

The

suboperculum is small,often hidden behindthe preoperculum,andfikewisetheinter- operculum(Goodrich).

Genus

DAPEDOGLOSSUS

Cope.

Syn. PhareodusLeidy(undefined).

The

largestandbestkno\^^l species of thisgenus isD.testisCope, fromtheGreen River Eoceneof

Wyoming,

ofwhichbeautifully pre- servedspecimensexistintheUnitedStates National

Museum,

andin theAmerican

Museum

ofNaturalHistory,

New

York.

One

nearlycom- pleteskeleton

m

^thelatter institutionisinstructive forhavingthebones ofthe skullpartly dissociated and displayedto excellentadvantage forstudy. ItiscataloguedasNo. 4587.

Through

comparison with thisspecimen ithas beenpossible to identify positively theisolated skull

shown

in plate 16,figure 1,asbelonging to this species. This well-preservedcraniumisthe property^oftheUnitedStates National

Museum

^(Cat. No.4916), andhas been carefully preparedso as to reveal the underside, freed from the matrix. Its primitive charac- ters are evident, and

among

surviving genera the resemblance is closest to Heterotis of tropical Africa.

1Foraninvestigation ofthecranialosteologyofthisgenus,seethefollowing:Bridge, T.W. Oncertain features oftheskullinOsteoglossum formosum. Proc. Zool. Soc.London,1895,pp. 302-310.—Ridewood, W.G. Onthecranialosteologyofthefishesofthefamilies Osteoglossidae,Pantodontidae,andPhracto- laemidae.Joum.Linn.Soc. Zool., vol. 19, 1905,pp.^252-282.

NO. 2177. FOSSIL

FISHES

IN

NATIONAL MUSEUM— EASTMAN. 289

Family

GONORHYNCHIDAE.

Thisfamily,representedinthe

Upper

Cretaceous

by

CJiaritosomus, andinthemiddle Eoceneb}'- Notogoneus, is

known

to have but one survivingspecies

m

^the

modern

fauna. ThisisOonorliynclius grcyi,a speciahzed form which inhabits the seas off Japan, South Africa, Austraha,and

New

Zealand. Ithas thehead and

body

coveredwith small ctenoid scales, and carries a ventral barbel on theprolonged snout.

The

supraoccipital separates the parietals, the premaxilla

Fig.9.—NotogoneusosculusCope. GreenRivee Eocene;TwinCreek,WTOJimG. Doesal(A) ANDLATEK.VL(B) ASPECTSOF CRANIUM,f. (AFTERL. HUSSAKOF). AlSfG,ANGULAR;AR,ARTICU- LAR;D, dentakt; £,ethmoid; FR,front.vl; HM, HYOMANDiBxn.AR; lOP, interoperculum;

J/«P,metapterygoid; l/A", maxilla;OP, operculum;P^,pariet.\x;PPP,prefrontal;PMX,

premaxilla;POP,pkeoperculum;PSP,paraspuenoid;PTER,pterotic; Q,quadrate;SANG, surangular; SO,supraoccipital;SOP,subopeeculum;SPH,sphenotic;SY,symplectic; VO, vomer;X,cjeek-plate.

articulateswith the maxilla and excludesitfrom themarginof the smallmouth. According toSmith

Woodward

^the

members

of this familyare related tothe Scopehdac.

Genus

NOTOGONEUS

^Cope.

Syn.Protocatostomus Whitfield.

The

typespecies of thisgenus is N. osculusCope,from theGreen RiverHmestoneof

Wyoming, m

^{sizeattainmga lengthof about60}

cm.

The

general structureoftheheadin this speciesisindicatedin theaccompanyingtext-figure9,takenfromL. Hussakof,^

who

com- binedinthediagramdetails

shown by

threespecimenspreservedin theAmerican

Museum

^ofNaturalHistory.

In plate 15, figure 2, is

shown

a

young

individual, which is the smallest

known

belonging to this species. It adds nothing to our

IBull.Amer.Mus. Nat.Hist., vol. 25, 1908, p. 83.

65008°—Proc.N.M.vol.52—17—

—

¹⁹

290 PROCEEDINGS OF THE XATIOXAL MUSEUM.

vol.52,

knowledge of thespecies in displaying charticters already observed, butit isinteresting forthe

same

reason that theyoungofotherspecies are interesting towhich

we

havecalledattentioninthe presentarticle

and figured in plate 14

—

^namely, immature examples ^of Belonosto- mus, Notagogus, and Acanthurus (pi. 14, fig. 1) (the last from the upper Eocene of

Monte

Bolca, Italy). These young individuals acquaint us withearlygrowthstages oftheseveral species in question, andenableusto comparetherelative proportionsof differentparts at differentperiodsinthelifehistoryforthespecies in question.

The

original of ourplate15,figure 2,^iscataloguedasNo. 6037.

Formationaridlocality.

—

Green RiverEocene,

Wyoming.

Family

CYPRINODONTIDAE.

In this family, which includes forms ofextremely small size, the

mouth

isprotractile,teeth arepresentonthejawsandpharyngeals, but rarely on the palate; the palatoquadrate arch is

more

or less reduced; and the supraoccipital extends forward to the frontals, separating the parietals in the median line. Only two or three extmctgeneraare

known

withcertainty.

Modern

forms aremostly confinedto fresh waters,butafewarefoundinbrackishwaterand on thesoacoast.

Genus

GEPHYRURA

Cope.

InJuly, 1891,E.D.Copepublisheddescriptions offive

new

species offossilfishes from asupposed

Lower

Tertiary horizon in

Roe

Hills,

South Dakota. Tliree oftheformswereregarded astypical of

new

genera,which were

named by

the authorGepJiyrura,Prohallostomus, andOligoplarcTius; thetaxonomicrelations ofthefu'sttwobeingwith the cyprinodonts, and of the last

named

with the percoids.

The

holotypesofthese

now

genera andspecies are

now

preservedin the American

Museum

ofNatural Historyand haverecentlybeenstudied

by

thepresentWTiter. AlthoughthetypeofGepJiyrura

was

referred

by

Cope with

some

hesitation tothe cyprinodonts, there doesnotappear tobetheslightestdoubtas tothecorrectness of this association; and the unique specimen available for study evidently stands in close relations with the

new

cyprinodont genus immediately to be de- scribedfrom theLahontan bedsnearHazen, Nevada. Forconven- ience of rcferrence, however,

we may

fij'st restatethespecific char- acters of G. concentrica, asdefined

by

Cope.

GEPYRURA CONCENTRICACope.

GephyruraconcentricaCope,Amer.Naturalist,vol. 25, 1891, p. 654.

Char,specif.

—

Theonlyspecimenisbrokenvertically acrossthemiddle, andthe posteriorhalf shiftedsoasto lieimmediately belowitsproperposition. Itappears thatlittleornopaitofthefishhasbeenlost. Eadii, P.9;D.9;C.6-16-8; A. II11.

NO. 2177. FOSSTL

FISHES

IN

NATIONAL

lIUSEUil—

EASTMAN. 291

V. 1-6; vertebrae, 10-18. Scales in twelvelongitudinal rowsbetweendorsal and ventralfins, andequal innumbertothe vertebraeonthe longitudinalline, or 28.

Headcovered withscales;fiveinavertical lineontheoperculum. Thedorsal,pec- toral,andventralfinsare rather small. Thecaudalfinisprobablynotmuchforked, ifatall. The^orbitislarge,butitsoutlinesare not well preserved. Theheadenters thetotallength fouranda quarters timestothebaseofthecaudalfin-rays,andslightly exceedsthedepthatthe ventralfins. Total length, 61mm.;lengthofhead, 15mm.

;

lengthtobaseofventralfin,24.5mm.;lengthtobaseofanalfin, 30mm.; depthat ventrals,14mm.;depthatcaudal peduncle, 6mm.

Formationandlocality.

—

^{Oligocene ( ?)}

^Ree

^Hills, South Dakota.

PARAFUNDULUS,

newgenus.

A

genus closely related to existingkillifislies, and alsoto the ex- tinct Gepliyrura, but distinguished from the latter cliiefly

by

its smallerandlessconspicuously

marked

scales,1^'ger

number

ofdorsal finrays, and presence of ahypural bone. Caudal fingethyrocercal.

Type

of^thegenus.

—

Parafundulusnevadensis,

new

species.

PARAFUNDULUSNEVADENSIS, newspecies.

Plate16,fig.2;plate17;plate18,fig.3.

A

^smallform attaining a total lengthof about 5.5 cm., inwhich the length of the head and opercular apparatus is contained three and one-half times. Dorsal comprising 11 rays, supported

by

an equal

number

of interspinous bones, and inserted opposite a point

midway

between thepelvicsand anal. Scalessmall and thin, with fine concentric markings, crossed

by

afew inconspicuous radiating proximalstriae.

Finformula: D.11; C.23; R. 10; V.9; P. 11-12.

The

specimen (Cat. No. 8120) selected as type of this speciesis

photographedof the natural sizein plate16,figm-e2,and a drawing of itis reproduced in plate 18, figure 3. Itis themost perfect of severalthatwereobtainedin 1905

by

Mi\ N.H.Darton,

m

strata of very white clay nearHazen, Nevada, which havereceived the

name

ofLahontonbeds.

From

the

same

localityasingle species of stickle- back,

known

as Gasterosteus doryssus Jordan,

was

described almost simultaneouslyin 1907

by

Drs. D.S.Jordan^{^}andO. P. Hay.^ Be- sidesthetype severalotherexamples of this species,

shown

inplate 17, werecollected

by

Mr. Darton atthe

same

locahty, andare

now

preserved inthe collection of the UnitedStates National

Museum.

The

writerisindebtedto his colleague, Mr. JohnTreadwellNichols, oftheAmerican

Museum

ofNatural History,forhelpfulsuggestions in regard to comparing this form and its scale characterswith the existingFundulus.

Formationandlocality.

—

Lahontanbeds; nearHazen,Nevada.

1Pub. Univ.Cal.,vol.5,1907,No.5,p. 131,figs.25,26; Smitlis. Misc. Coll., vol.52,1910, p. 117.

»Proc.U.S.Nat. Mus.,voj. 32, 1907,pp.271-273,figs.1-3.

292 PROCEEDINGS OF THE NATIONAL MUSEUM.

^{tol. 52.}

Family

CYPRINIDAE.

Genus

AMYZON

Cope.

TMs

isanextinctgenusrelated to

modern

suckers,butvnth.a

more

extendeddorsalfin. Itisstated

by Woodward

to be "scarcelydis- tinguishablefromSclerognathus,butwith pharyngealbones expanded behind." Mr. J. T. Nichols has pointed out to the\\Titerthat the Canadian species A. hrevipinne approaches very closely to existing buffalo fishes of the genus ^Ictiohus; and Cope, in his descriptionof Amyzon,hasremarked

upon

itsnearrelationswith BubalicMiys.

AMYZONBREVIPINNECope.

Plate19,figs.1, 2.

AmyzonbrevipinneCope, Proc.Acad. Nat. Sci. Phila.,1893, p.402.—Lambe, Trans.Roy.Soc.Canada,vol. 12,1906, pp. 151-155,^pi.1.

The

typeoftliisspecies

was

obtainedfrombedsin BritishColumbia supposed to beof lateEocene orearly Miocene age. Ithas never been figured, but one small specimen from Horsefly River, British Colmnbia,andanother belongingtoadifferent species (not ofA.com- mune, however), have been described and illustrated

by

Dr. L.

M.

Lambe

mtliin recentyears.

To

this speciesshouldprobably bereferred two specimensfigured

illplate19,which werecollectedin1910

by

;Mr. J.B.

Umpleby

from beds ofsupposed

Lower

Miocene age, near Republic, Washington.

These are

now

the property of the National

Museum,

and arecata- loguedasNos.81and8117. Tlieyarethelargestand best preserved examplesof this speciesyetbroughttofight.

Genus LEUCISCUSCuvier.

LEUCISCUSTURNERILucas.

LeuciscusturneriLucas,21st Ann. Rept. U.S.Geol.Surv.,1901,pt.2,pp. 223- 224, pi. 31.

The

holotype of this species (Cat. 4302a), and alarge

number

of weU-preserved specimens from the Esmeralda formationinwestern Nevada, ^recontainedintheUnitedStates National

Museum

^collec-

tion.

The

age of these beds is discussed

by

H.

W.

^{Turnerin the}

Twenty-first

Annual

ReportoftheUnitedStates Geological Survey, 1899-1900 (part2,pp. 203-205), andinthe

same

volmne (pp. 209- 220) the fossilplants occurring in this formation are described

by

F.H. Knowlton.

Nothing can be addedtoourknowledgeofthespeciesbeyond the information ah'eady contributed

by

Dr. F. A. Lucas.^ There

may

be compared with it, however, a specimen that apparently belongs to

121atAnn.Rept.U.S. Geol. Surv., 1901,pp.223-224,andProc.U.S.Nat. Mus.,vol. 23, 1900,pp. 333- 334,pi. 7.

NO. 2177. FOSSIL

FISHES

IN

NATIONAL MUSEUM— EASTMAN. 293

thisor aclosely related species,fromtheMadisonValley,Montana, whichis

shown

of slightly larger than the natural size on plate 18, figure4. 'Mi. Earl Douglass,

who

collected a small

number

offish remainsfromthislocality,isofthe opinion that thestratacontaining

them

^isofOhgocene or

Lower

Mioceneage.

An

undeterminedspe- cies ofOsmerus is apparently indicated

by

theoriginal of plate18, figure5, from the

same

locahtyas thelast. Both arepreserved in the Carnegie

Museum

inPittsburgh. Inplate 19, figure 3,is

shown

an exampleofan undeterminedcyprinodontspeciesfromthe Tertiary ofMexico.

Family

SILURIDAE.

Fossil representatives of this family are.

uncommon, and

for the most part imperfectly preserved. Fragmentary remains from ^the

Lower

Eocene have been assigned to Arius, and others from the

Lower

Miocene to Ameiurus.

The

undermentioned specimen^is the only nearly complete exampleof thelattergenus that hasbeen discoveredinthefossilstate.

Genus

AMEIURUS

Rafinesque.

AMEIURUSPRIMAEVUS, newspecies.

Founded upon

aunique specimen withoutlocalitylabel,butas far as

may

be judged fromtheappearanceof the matrixitwouldseem to have been derived from the Green River Eocene of

Wyoming.

Clearlyrelated to the

more

generalized andrepresentativegenus of

modern

North AmericancatfishesAmeiurus,itdiffersfromallliving species in its shorter anal,oonsistmg of only 12 rays,and in this respect resembles the

more

specialized genus Leptops,

known by

a single species. Itis

shown

inplate 20.

The

holotypeof this speciesisa nearlycompleteskeletonhavinga totallength to the base of the caudalfinof 21.5 cm., inwhich the lengthofthehead including the supraocoipital process is contained three times.

Head

broad behind, not

much

^contracted foi-wards;

surface sculpture consistmg ^of anastomosing rugae and pittmgs as in the recentA. catus; orbitssmall; fontanclle situated justbefore theorigin ofthe supraocoipital process.

Maximum

breadthofhead initsflattenedcondition equaltoitslength. Vertebrae 29innumber, withstrong neuralspines. Dorsalfinbetween the pectoralsandver- trals, with about six branched rays of moderate length (its spine brokenaway). Pectoral spinesstrong,smooth,lessthanhalfaslong asthe head. Ventralswitheight,analwithapparentlynot

more

than twelverays.

Rays

ofcaudalfinmostly broken away, butthecurved

lineformed

by

their articulationwith hypuralfin-supportsindioatmg that thefin

was

rounded. (Cat.No.8122,U.S.N.M.)

Thisisanextremelyinteresting

and

w^ell-preservedspecimen, save that the caudalfin isforthe

most

part lackmg, andthe dorsalhas beenfolded under the neuralspmes andpartly concealed

hj

them.

294 PROCEEDINGS OF THE NATIONAL MUSEUM.

vol.52.

The

supraoccipital process hasbeenshifted slightly tooneside ofthe anteriorvertebrae, and

was

apparently notindirectconnectionwith them.

The

relationsare evidently very close ^vith theexisting A.

catus, excepting asregards the smaller

number

of analfinraysand nonserrate character of thepungent pectoral spines.

With

respect tothe short-based analfin,it

may

besaidthat although only adozen rays are

now

^{visible in}the specimen, afew

more may

have beenpre- sentinadvanceofthose

now

tobeseen,but

became

lostorwerecut

away by

carelesstrimmingofthespecimenalong the ventral margin.

Infact,one canalmostcertainly distinguish, althoughfaintly,traces of afewinterspinous bonesin advance ofthe foremost analfinray

now

appearing in the specimen.

At

^the most, however,

we must

admittliatthisfin

was

shorterthaninexisting species ofAmeiu7^s, butothenvise thedifferencesare ofbut minorcharacter.

The

latterobservation appears the

more

remarkableif

we

accept theviewas correctthat thefossilbeforeusisofMiddle Eoceneage.

How

closely Rliineastes, fromtheGreen RiverEocene, agreesstruc- turallywith

modern

species of Silurids,cannot bedetermined,asit is

known

only

by

fragmentaryremains.

But

in the nearly complete skeleton which

we

are

now

^considermg

we

find evidence that the typicalexpressionofthegenusAmeiurus,as

we know

itto-day,

was

already attainedin theearly Tertiary, andhaspersisted unchanged oversince.

Formation andlocality.

—

Supposedly fromtheGreen River Eocene of

Wyoming.

Family

PERCIDAE.

This family and the small one

known

as Aphredoderidae,

now

nearlyextinct,areincluded in DoctorGill's superfamily Percoidea.

The

typesof

most

^ofCope'sspecies of

AmpMphga,

Asinepos, Erisma- topterus,andTriclioplianes,belongingto the Apliredodcridae, arepre- servedinthecollection of theUnitedStates National

Museum. The

nearlycomplete example of TncJioplianes

foliamm

Cope, which has recentlybeenfigured

by

T. D. A.Cockerell,^ is the propertyofthe American

Museum

ofNaturalHistory,andthe typesofCope'sspecies of Mioplosus are dividedbetween this institution and the National Museum.^^

Fossilperchesintherestrictedsenseofthetermarerepresentedin the Tertiary rocks of this country

by

the genera Mioplosus, Plio- plarclius, andOligoplarchus.

The

firstofthesediffersfrom Percain havingfewer vertebrae

and

aspineless operculum. Itis

known by

several species intheGreen RiverEocene,ofwhichthe genotype,

M.

lahracoides Cope, is the

most common.

^{Probably theso-called M.}

1Amer.Naturalist, vol.42,1908, p. 571.

2Forlistsofspecimensseethe published cataloguesoftypeandfiguredspecimens belongingtothese museums,ah-eadyreferredto. Thepartsdealing withfisheswere publishedin 1907and1908, respectively.

NO. 2177. FOSSIL

FISHES

TN

NATIONAL MUSEUM— EASTMAN. 295

longusisonly a

young

exampleof thisfomi, theindividuals ofwhich

show

a considerable

amount

of variation. Inplate 21 is

shown

a large (44cm.long)andwell-preservedspecimenv/hich

may

bereferred to M. lahracoides, notwithstanding the fact that it displays one abdominal

and

one caudal vertebrainexcessof thenormal

number

occun-inginthetype-species.

The

typeof M.multidentatusCopehas not beenfigured. ItispreservedintheAmerican

Museum

^ofNatural History.

Formation and ^locality

.—Gveen

River Eocene;

Wyoming.

Genus

PLIOPLARCHUS

Cope.

This extinct genus, closely related to Miophsus,isrepresented

by

three

Lower

Tertiaryspecies, ofwhichonlyone, P.wliiteiCope,has beeniUustrated.

The

differencesbetween this species, whichisthe type,andP.sexspinosus, arestatedto consist inthe

more

numerous spinous,andlessnumeroussoftrays, ofthedorsalandanalfins. In P.whiteitheradialformulais:

D. IX-12; C.-17-; A.V-14; V. ?; P. 13.

And

ⁱⁿP. sexpinosusitis:

D. X-13; C. -17-; A. VI-9.

InP. septemspinosus theformulaisgivenas:

D. XI-?; A. VII (?)-12.

PLIOPLARCHUSSEXSPINOSUSCope.

Plate15,fig.1.

Plioplarchus sexspinosusCope, Amer. Joiirn. Sci., vol. 25, 1883, p.416; Rept.

U.S.Geol. Surv.Terr., vol.3,1884, p. 729.

Intheoriginal descriptionit isstated that "thisspecies is repre- sented

by

twospecimens,bothofwhichlack thehead and

body

an- terior to the dorsalfin." These specimens are

now

in theUnited States National

Museum

collection,and oneofthem,

marked

"type,"

iscatalogued asNo.4236. Itisfrom the

Lower

Tertiary,perhaps Miocene, near Sentinel Butte,ⁱⁿBillmgsCounty, North Dakota.

A

much more

completeexample, alsofrom the typelocality, is

shown

inourplate 15, figure^1. Initthe analfinisseentohavesixspinous andtwelvesoft rays.

The

specimenbears the catalogue

number

8118.

Formation and^locality.—Miocene ( ?)^; topofSentinel Butte,North Dakota.

PLIOPLARCHUS SEPTEMSPINOSUS^Cope.

Plate22.

Plioplarchusseptemspinoms Cope, Amer.Naturalist, vol. 23,1889, p.^625.

This species

was

established

by

Cope

upon

the evidence of four distorted and mutilated specimens from shales near

Van

Horn's ranch,on theJohn

Day

River, Oregon, thestratawhencetheywere