PROCEEDINGS

XXVIII, pp. September

PROCEEDINGS

orTHE

BIOLOGICAL SOCIETY OF WASHINGTON

SCALES OF PANAMA FISHES.

BY

T. D. A.

COCKERELL.

The year

^before last,

when Dr.

S.

E. Meek was working

^at

^the U.

S.

National Museum on

his collection of

Panama

fishes,

he very kindly allowed me

to

take

scales of

a

_large

number

of species.

The

present report deals

with

this material,

including

also

a few

species collected

by

^{others in the}

Republic

^of

Panama

or the

Canal Zone. Whenever the

collector's

name

is

not

_given,

it is to

be understood

thatthe material

was obtained by

^Messrs.

Meek and Hildebrand.

The

_following

key enumerates

the species studied,

and shows how _they may ^be

separated.

With few

exceptions, a single

normal (not regenerated)

^scale

from

the

middle

ofthe sidewill serve todistinguish

a

_species

from

all

the

others in

the

col- lection.

Scalescycloid ⁽obscurely ctenoidⁱⁿ Gerres) 1.

Scalesctenoid 16.

1. Circuli

and

radii entirelytransverse (Clupeidae)

Opisthonema ^{libertate Giinther.} (Tabogal.) Circuli transverse, basal only;

no

radii, butbasalmarginlobed,

and

radial folds faintlyindicated (Atherinidae)

Atherina

arxa

Jordan

&

Gilbert. (Porto^Bello.)

Circuliconcentric; radii absent 2.

Circuli concentric, or rarely (Opisthopterus) essentially transverse;

radii present, not transverse, or (Gastropelecus) only partly^{so . 3.}

2. Circulicomplete; ^scalesvery minute(Belonidse)

Tylosurus scapularisJordan

&

Gilbert. (Balboa.) Circuli absentfrom_apicalpartof scales (Characidse)

Roeboides guatemalensisGunther.

(Gatun

^R.) .3. Scales elongate^; with very

numerous

radiiallaround(Rypticidse)

Rypticusnigripinnis ^Gill. (Corazal.)

Radii notallaround 4.

32—Pboc. Biol.Soc.Wash.,Vol.XXVIII.^{1915. (151)}

4. Scalesvery broad

and

_short, with very few ^radii; circuli absent in

apical ^{field 5.}

Scales longerthan_{broad, or}ifbroader than_long,notexcessively^{so. 6.}

5. Circuliextremelydense,essentiallytransverse (Clupeidse)

Opisthopterus ^dovii Giinther.

(Panama

market.) Circuli notdense, concentric(Hemiramphidy?)

Hyporhamphus

unifasciatusRanzani. _(Balboa.)

6. Smallscales,with roundedlaterobasal corners,

and numerous

radii ^. 7.

Larger ^scales, with angular^or subangular laterobasal corners, or ^if (Gastropelecus) sometimes rounded, then ^radiiveryfew

....

9.

7. Lateral

and

apical circulicrossed

by

^{fineradial lines of} hyaline dots (Pleuronectidae) ^{. .} CitharichthysspilopterusGiinther.

lowerside (Toro^Point.)

No

suchlinesofhyaline dots ^8.

8. Radii (allbasal) ^lessthanten (Pleuronectidae)

Paralichthyswoolmani Jordan

&

Williams, upper

and

lower sides (Taboga^I.]

Radiiextremely

numerous

; circuli in apical ^fieldbroken

up

⁽Dacty- loscopidse) Dactylagnus

mundus

GiU. (Taboga^I.)

9. Radii few, mostlyapical^; or

when many

^basal,central region^ofscales withcoarseradial reticulation (Characidse) ^10.

Radii numerous,^allapical^;basal

margin

^notlobed (Characidse) ^.11. Radiiapical

and

basal; basal margin deeply lobed^; center of scale notreticulated(Characidse) ^. Curimatus magdaleneeSteindachner

(Rio Abaco.)

Radii basal ^12.

10. Circuliabsentin apical^field; centre of scalewith aminute reticula- tion derivedfromthe circuli

Gastropelecusmaculatus Steindachner. (Creek nearChorera.)

Circulibrokenup, butcoarse

and

conspicuous,in apical field;

some

scaleswitha centralreticulatepattern derivedfromtheradii Piabucina panamensis^Gill. (RioCalobre.)

11. Circulirepresentedin apical^field

by

^abrokenzigzagpattern

^sJyanax^sc^m

Steindachner. (Creek nearChorera.) Circulinot showinga zigzag patternin apical^field

Astyanaxgrandis

Meek &

Hildebrand. (Rio Abaco.)

12. Scales longerthan_broad, withonly three or fourradii, indeepfolds;

basal

margin

^{lobed (}Synodontidse)

/Si/nodus

evermanm

Jordan

&

Bollman. (Taboga^I.)

Scalesbroader than long ^13.

13. Radii 6to 8; basalmarginstrongly crenulate (Xystsematidse) OerresrhombeusC.

&

V. (Mindi^Cut.)

Radii

more

than ten; basal

margin

^without regular crenulations

(Poeciliidse) ^14.

14. Circuliobtusely angulateⁱⁿ

median

lineabovenucleus

Pceciliasphenops^C.

&

V. _(Corazal.)

Cockerell

—

_{Scales of}

_Panama

_Fishes.

₁₅₃

[Circiili notangulateⁱⁿ

median

lineabovenucleus; scales larger.

Xiphophorus ^helleriMeckel. (Obispo, Mexico; ^{S. E.} Meek).]

15. Apicalmargin with ^asingle rowof sharp teeth,

and no

ctenoidele-

ments below

them

; nucleus_justbelow_apical

margin

^16.

Apical ^fieldwith one or

more

rows of ctenoid elements below the

marginal^{teeth 19.}

16. Scales withroundedbase (Pleuronectidse)

CitharichthysspilopterusGthr.,upper^side.

Scales with truncate _base, the laterobasal corners evident; apical

margin

usually ^like the transverse section of a roof of a house

(Gobiida?) 17.

17. Basalradiiabout 7to9 Chonophorus^nelsoni

Evermann

(RioCulebra.)

Basalradii

more numerous

18.

18. Scales 2.5-3

mm.

long Philypnus maculatus (Chepo.) Scaleslessthan2

mm.

long^{. .Eleotris}pisonisGmel. _{(PortoBello.)} Gobius_claytoni. (Mindi.) Gobius soporatorC.

&

V.

(Fox

Bay, Colon.) 19. Scales conspicuously longerthan_{broad,parallel}sided 20.

Scalesbroader thanlong,oratmostalittlelongerthan broad. . 22.

20. Scales minute,

much

lessthan 2

mm.

_long; nucleus never _elongate,

alwayssubapical (Soleidse) 21.

Scaleslarger, at leastover2

mm.

long; nucleus often elongated,

and

thenradii not reaching middle _{of scale;} apical teeth often trun- cate (Serranidee) ^. Petrometoponpanamensis^Steind. (Tabogal.⁾ BodianusacanthistiusGilbert.

(Panama

market.) Mycteroperca xenarcha Jordan. (Corazal.) Paranthias_furciferC.

&

V. (TabogaI.) 21. Scales

more

orlessnarrowed_apically; radii3to 5

Achirusfluviatilis^. bothsides (Rio_Chorera.) Scalesnot narrowed _apically; radiivery

many

Symphurus

plagusia B.

&

S. bothsides

(Fox

Bay, Colon.) 22. Basal radiifew (4 to 6),

and

widely spaced; subapical ctenoid ele-

ments

brick-like(MuUidse)

TJpeneus maculatus^\och.

(Fox

Bay.) Basal radii

many,

^{or at least closer} together, usually arranged ⁱⁿ fan-like

manner

; subapical ctenoidelementsnot brick-like. . 23.

23. Scalesminute, ^1.5

mm.

^orlesslong; onlyaboutthreerowsof distinct ctenoidelements belowthemarginal^teeth(Sciaenidse)

Menticirrhus martinicensisC.

&

V.

(Fox

Bay.)

Scales_larger 24.

24. Elements of ctenoid patch triangular, feebly developed, with

no

sharp^salientmarginal^{teeth (}Xystaematidse⁾

Eucinostomuscaliforniensis ^Gill. (MindiCut.) Elementsofctenoid patch ^well developed; marginal ^teeth salient,

narrow and

sharp, or _(Stellifer

and

Orthopristis) often _slightly

bifid 25.

25. Submarginalctenoidelements likethe marginal; ctenoid patchvery large; basal

margin

very strongly crenate, the lobules ^free from circuli (Cichlidse) ^{. .}JEguidenscasruleopunclatus

Kner&Steind.

(RioLas Sabanas, A.

H.

_Jennings.) Satanoperca crassilabrisSteind. _(Frijoles, Canal Zone.) Submarginalctenoidelements unlike themarginal

... ...

26 26. Submarginal ^ctenoid elements _long, strongly ridged; basal margin

notcrenulate(Sciaenidse) ^{. . .} Paralonchurusdumerili Bocourt.

(Panama,

C.

H.

Gilbert.)

Submarginal^ctenoidelementsshort _27.

27. Basalmargin not^crenulate (Scisenidae)

Odontoscion dentexC.

&

V. (ToroPoint.) Stellifercolonensis (

Mindi

Reef.₎ Basalmargin crenulate (Hsemulidpe)

Orthoslcechus maculicaudaGiW. (Tabogal.) Pomadasis bayanusJord.

& Everm.

(Corazal.) Pomadasis macracanthusGthr. (Balboa.) Orthopristis chalceus Gthr. (Balboa.) CHARACID^.

An

accountofthescales ofthe neotropical Characidfe

was

_publishedin Annals Carnegie

Museum,

ix,1914. AVhile recording the

Panama

_species,

Igive

some

notes

on

_others, supplementary^tothepaper mentioned.

BiviBRANCHiix.E. BivibrancMa protractila

Eigenmann.

Rockstone, British

Guiana

(Eigenmann). ^{U. S. N.}

M. A

remarkable _fish, re- sembling Albulavulpes.

Scalesabout2

mm.

long

and

over2.5 wide,broadly rounded apically, sides sloping, base verybroadlytruncate, laterobasal anglesdistinct, basal

margin

^{with a} strong

median

_lobe, nucleus _central; circuli coarse, absentfrom_apicalfield.

No

radii, butaradial fold

on

each side of basallobe.

The

scaleresembles thatof the Curimatinse, but

ispeculiar^forthe_singlebasal _lobe,

and

the thin

weak

_{apical field,} withoutradiiormarginalteeth.

ANOSTOMATiN^.

ScMzodonf

asciatus Spix. BoHvia( Gibbon). U.S.N.

M.

Large ^red scales,subquadrate, about ^9.5

mm.

long

and

broad, basal

margin

strongly bilobed, ^circuli very^fine

and

dense, nucleus approxi- mately central; apical radii few

and

weak, except one on each side, basal radiirudimentary;apical^fieldcoarsely pustular.

The Anostom-

atinegenera

known

to

me

are separable thus^:

Scaleswith _{a strong} radial line extending horizontally^to each _side,

and

two (rarelyone) extending

upward

^toapex,in lateral line scales

sometimes twoalso tobase Anostomus.

Scaleswithout suchradialpattern^; apical^fieldwith

weak

radii,usually

rather

numerous

1.

1. Nucleuswellbelow middle _Leporinus.

Nucleiis aboutmiddle _Schizodon.

BRYCONiN^.

Brycon

^{dentex Gthr.} Nicaragua (L. F.

H.

Birt.). U. S.

N.

M.

Scalesabout 11

mm.

_long

and

13broad.

The

three before

me

Cockerell

—

_{Scales of}

_Panama

_Fishes.

₁₅₅

are all latini;cleate, but they agree with those of B. _falcatus in all essential characters; the apical radii are

many ^and

parallel,

and

the apicalmargin ^isstrongly crenulate or subdentate, though not ctenoid in

any

proper^sense.

TETRAGONOPTERiN,E.

Two

_species of

Astyanax

^are given ⁱⁿ the table above. A._fischeri isrelated in the characters

shown

_bythe scales to

^4.bimaculatus,thoughthescalesare

much

smaller (5-5.5

mm.

_broad),

thecirculiarelessdense,

and

the

weak

subparallelapical radiiarevery different.

The

relationship ^is perhaps actually ^closer with the still

smaller scales of the A. _{polylepis group.} A. _grandis isvery ^{close to} fischeri.

Thus

thetwo

Panama

species formalittle groupintermediate between the South

American

_groups of A. bimaculatus _(large scales with very coarsespreadingapical radii)

and

ofA._polylepis,abramoides,

etc. (small scales with very

weak

apical radii).

The

J., mucronatus group (with V-likeapical radial pattern) stands apart fromall these;

to it

must

be added A. fasciatus Cuvier, Rio _Primero, Argentine (J.

W.

Titcomb).

The

mucronaius-fasciatusgroup^istypicalAstyanax.

Poecilurichthys, which

Eigenmann

^has recently treated ^as a distinct genus, ^istypified

by

A. bimaculatus_;but_accordingto scalecharacters

we

should be inclined to include in it

some

of the _species which

Eigenmann,

^{in his}catalogue, has leftin typical Astyanax.

GASTEROPELECIN.E. Gasteropclecus maculatus has scales which do not differ in

any

tangible

way

^{fromthoseofG. sterniclaL.}

piABUCiNiE. Piabucina panamensis has ^{scales of the}

same

general type as those of Chalceus macrolepidotus^, but

much

smaller (length 7,

breadth8

mm.). The

intermediate,

minor

apical radii arenotsoevi- dentinthePiabucina as inthe Chalceus. Inall respects the^{scales of} Piabucina areessentially as in the African genus^Alestes.

Both show

aradialpolygonal patternⁱⁿthe middle, at least in

some

ofthescales.

Arethe Piabucinae separable^asa subfamilyfromtheAlestinae?

HYDRocYNiN^. Luciochavax _{insculptus Steind.,} Rio Abaco,

Panama (Meek &

Hildebrand). ^Scales subquadrate, about 6

mm.

_long

and

broad; basal

margin

strongly undulate, or emarginate ⁱⁿ middle;

nucleus alittle abovethemiddle; circuli fine; afewbasal, apical

and

lateral radii (1or_{2 apical,} 1 to3_basal, 1 to 5 lateral, but

when more

than onelateral,they are only partly developed)^,inAlestiform fashion_; apical ^field without circuli, but

thrown

into strongparallelpleatsor folds(notradii),

which end

asstrongsharp^teeth

on

themargin.

The

circuli are microscopically beaded. These are ctenoid _scales, using that

word

in a purely descriptive ^sense. There is

no

resemblanceto Hydrocynus, ^{at least} judging from ^{thescales of the latter}genus ^seen

by

me,

which

are, however, apparently quite immature.

Eigenmann

states thatadultscales of

Hydrocynus

^havedenticulate_apical margins.

The

scale ofLuciocharaxissingularly^likethatof Phractolxmus.

CHARAciN^.

The

_generaof this subfamily

now

before

me _may

_{be sepa-}

ratedthus^:

Apical^fieldwith

numerous

sharp^{isolated teeth in}about_{four rows,}one ofwhich^ismarginal ^; base broadly rounded, with rather coarsecir- culi,

and no

radii; nucleus central.

The

arrangement^{of the teeth} reminds oneofthe Macruridse.

Cynopolamusargenteus Val.,Paraguay (Page)^;U.S.N.

M.

No

suchteeth in apical field; scalecycloid ^1.

1. Apical ^field with

many

radii, between which the circuliare very coarse

and

widely spaced, ⁱⁿ complete contrast with thoseof the base and sides of the _scale; scale about 4.75

mm.

_long

and

6 broad; laterobasal corners rounded; no basal radii. System ^of circuli like thatoftheAfricangenusSarcodaces

SalminusmaxillosusC.

&

V. Paraguay(Page);U.S.

N.M.

Apical^fieldwithoutradiiorcirculi 2.

2. Circulivery dense, largely transverse, wavy, broken and branching;

weakly developed

and

very variable undulating ^{transverse radii;} scalestransversely oval or nearly^circular. Theselooklikeclupeid

scales

Bramocharax

bransfordii^Gill.

Nicaragua (Bransford); U. S. N.

M.

Circuli not very dense; apical ^field with evident growth-lines;

no

trace of radii

anywhere

^3.

?). Scalessmall, transversely short-oval ^. Rceboides guatemalensis Gthr.

Scalesextremely broad, butofthe

same

_type

Charax and

Acanthocharax.

Thereare_clearlyindicatedseveral tribes; Characini(Charax, Acanth- ocharax

and

Rvehoides), Salminini(Salminus),

and

Cynopotamini (Cynopotamus).

The

Bramocharax, ^ifcorrectly determined, rep- resentsanother_group.

CURIMATIN^. Curimatus _magdalensehas scales like those of C. spilurus Gthr.

CLUPEIDiE.

Opisthonema ^{libertate scales} (mine _evidently immature) do not differ fromthose of0. oglinurn,except ⁱⁿ the absence of pitting.

The

first transverse radius ^iscomplete, thesecond alwaysinterrupted.

Opisthopterus ^dovii belongs ^to the Pristigasterinae.

The

five transverse circuli

and

the hyalineapical ^fieldare as in the Clupeinse, but the few radii in the basal field are directed obliquely basad, instead^ofbeing transverse. Therearefine irregular apical radii,poorly developed.

The

apical margin^isnot dentate.

SYNODONTID^.

Synodus evermanni^{scales are like those of} S. foetens, butsmaller, with the nucleus nearer the ^centre. Probably

most

of thisdifferenceisdue toimmaturity.

POECILIIDjE.

Poeciliasphenops ^scalesdonotdiflerappreciablyfromthoseof P.butleri.

Xiphophorus ^{helleri scales} resemble in structure those of Pseudoxipho-

Cockerell

—

_{Scales of}

_Panama

_Fishes.

₁₅₇

phorus himamlatus Heckel, from _Quirigua,

Guatemala (W.

^P.

Cockerell), butdiffer iu detailthus:

Scales subquadrate,about5

mm.

_long

and

broad; apical circulidenser Xiphophorus^. Scales broaderthan _long,about_{3.6 long}

and

4.5broad; apical circuli

notdense Pseudoxiphophorus.

BELONID.E.

Tylosurus scapularis^scaleslook likevery

young

^{scales of T. acus.}

The

structureisentirelythe same.

My

^{material of scapularis is} probably immature.

HEMIRHAMPHID.E.

Hyporhamphus

unifasciatus from _Balboa, Canal _Zone, differs from the

same

_speciesfrom

Woods

Hole, Mass., ⁱⁿtheabsenceofthedense fine circuli inthe _apicalfield,butitisimmature.

At

thesides ofthe apical field intheBalboa fishthereare variable_feeblydeveloped ^{circuli run-} ning

upward

^insteadof transversely, abouttwo-thirds as farapart ^as theuppermost ones^{ofthe basal} field,

which

_theyalmostmeet. Inthe narrow zone between these

two

sets of circuli arerudimentsofthefine transverse circuli which are so conspicuous ^{in the}

Woods Hole

fish.

The

Balboafishispresumably ^true

H.

unifasciatus; thatfrom

Woods

Holeshouldperhaps ^bereferred to

H.

roberti. It remains to be seen whether _fullymature scales from the

Panama

_region will be readily distinguishablefromthoseofthenorthernfish.

ATHERINID^.

Atherina araeascalesagreeentirely in type withthoseofA._pinguisfrom Australia, even to the frequent presence ^{of a} curious nuclear _pit.

A. araeascalesare

however much

smaller,about1.5

mm.

long

and

3.5 broad. In

Mem.

_Queenst. Mus.,

Dec,

1913,^Ihavediscussed thescales ofAtherina,

and

_givenan accountofanother

Panama

_species,A. _stipes.

The

lobules inthemiddleof the basalmargin ^of A. araeaare usually two, but sometimesthree.

MULLID^.

Upeneus maculatus scales

do

not differ appreciably from those of U.

dentatus.

RYPTICID^.

The

scales of Rypticus ^{are so} entirely different from those of the Serranidee that adistinctfamilyappears^tobeindicated.

The

absence ofthe anal_spines,

and

other characters, further distinguish the group.

The

_groupiscalledGrammistinte byBoulenger,

and

_appearsto include three genera. Jordan

and Evermann

recognize

two

subfamilies,

Gram-

mistinae

and

Rypticinae.

Scales of Rypticus nigripinnis are minute, about ^1.3

mm.

_long

and

.4

broad; greatly elongated, with rounded ends; nucleus _central, elon- gated; ^radii numerous, ^all around; ^circuli coarse; margin ^{of course} whollycycloid.

Only

a very small area^iscoveredby^skin.

The

_apical

end

^isdistinctly larger than the _basal,

and

the apicolateral ^radiiare curved, with the convexityupward. Thesescales are extraordinarily similar in general type ^to those of

some

Ophidiidse, ^as

Lepophidium

microlepis.

SERRANID^.

Petrometopon, Bodianus

and

Mycteroperca are Epinephelinse,

and

have the characteristic elongated, parallel-sided scales of that subfamily.

Scales of P. panamensis ^are broader than those of the Australian Epinephelusmegachir,butotherwiselittledifferent. Paranthias_furcifer (subf.Anthiinse) hasscalesabout 5

mm.

_long

and

3.5broad,entirely of thePetrometopon type.

Thus

itappears ^that the Anthiiufecan not be separated from the Epinephelinse on the scales. Bodianus acan- thistius scales, only about 2.5

mm.

long, are structurally ^like the others,butwith rather conspicuously bulging orconvexsides. Myctero- perca^islikeBodianus,^{but thescales are}larger

and

thesides straighter.

H^MULID^.

The

scales in this family, so far as seen, are veryuniform, subquadrate, usually broader than long, always with ^{the lower}

margin

^{crenate or}

scalloped, a characterwhich distinguishes

them

from the Sciaenidse, InPomadasisthescales are aboutaslong^asbroad,

and

the marginal teethare pointed. In Orthostoechus thescalesare broader than _long, lengthabout_3,breadtli3.75

mm.

; themarginal^teethare pointed. In Orthoprististhemarginalteeth are subtruncate or obscurelybifid,

and

thescale^isvery^littlebroader than_long.

XYSTAEMATID^.

The

scale of Gerresrhombeus appearscycloid, but under the

compound

microscope ^the ctenoid patch, ^of

weak

transversely

diamond

_shaped dentiform structures, can be seen. Theseareevencrossedby broken

and

rudimentary^{transversecirculi.} IntheEucinostomusthesectenoid elements are narrower

and

rather

more

_{distinctly dentiform,} without rudiments of circuli; but the general type ^of structure is the same.

The

Australian

Xystaema

darnleyensishas lost the ctenoid patch^alto- gether.

Both

the

Panama

_species have transversecirculireaching the middleofthelateralmargin at rightangles withit.

SCIAENID^.

The

lateral circuli are directed vertically, instead of transversely ^{as in} thelastfamily.

The

scalesarebroaderthan _long,

and

thelower margin

iswithoutdistinctcrenulations. Leavingout Cynoscion, the^scalesare ofa veryuniform_type,but Paralonchurusdumeriliisreadilyseparated fromthe others

by

^thecomparativelylong

and narrow

elementsof the ctenoid patch. ^Ithas themarginalspinesstout, obtusely pointed.

CICHLID^.

Cichlidscalesarebroaderthan_long(sometimes notgreatlyso), withthe lateral circuli vertical,

and

the basal

margin

strongly crenate. In

my

Cockerell

—

_{Scales of}

_Panama

_Fishes.

₁₅₉

table in Bull. Bureau Fisheries,

XXXII,

p. 160, both the

Panama

species runtoGeophagusbrasiliensis, tliough the Satanoperca scale ^is not _quite so broad as the others.

The

Brazilian Chietobranchopsis ocellarisdoes nothavebasal lobulesfreefrom circuli; butsucli lobules are well developed ⁱⁿ Chaetohranchus flavescens, as in the

Panama

species. Tilapia^niloticahas shortbroadlobulesfreefromcirculi.

GOBIID^.

The Panama

scalesconfirm_{the very}distinct and _nearlyuniform Gobiid type, already discussed elsewhere.

DACTYLOSCOPID^.

Dactylagnus

mundus

has_alreadybeendiscussedinProc.Biol. Soc.

Wash.

_,

XXVI,

p. 91.

PLEURONECTID^,

Paralichthys woolmani_{scales, as} represented by

my

material, are

much

smaller thanthose ofP. oblongus, butthe structureisthesame.

Citharichthysspilopterusare also small, but structurally closelyresemble

Syacium

papillosum, the^lateralfieldbeingessentially afterthe

manner

of this species, notas inCitharichthysarctifrons.

SOLEIDiE.

Achirusfluviatilis scalesagreestructurally withA.lineatus.

Symphurus

plagusiascales essentially agreewith thoseof S. piger.

XoTE.

—

_{InProc.Biol.Soc.}

_Wash.

,

XXVI,

p.77,thescale ofBregmaceros

isdiscussed,

and

said to apparently lend

some

_supportto the

Tims

_theory

of theorigin of teleostean scales. Furtherinvestigationshowsthat this ideawas based on an^errorof interpretation,

and

otherstudiesappear^to

show

conclusively that the

Tims

_theory^isincorrect.