This item has the following downloads:

PDF ( 6 MBs ) ( PDF )

    Title Page
        Title page
    Main
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20

STATE OF FLORIDA
STATE BOARD OF CONSERVATION
Ernest Mtts, Director

FLORIDA GEOLOGICAL SURVEY
Robert Vernon, Director





SPECIAL PUBLICATION NO. 2


CONTRIBUTIONS TO FLORIDA

VERTEBRATE PALEONTOLOGY


PAPER NO 3

DESCRIPTION OF A BERYCIFORM FISH

FROM THE OLIGOCENE OF FLORIDA




By
David H. Dunlme
U. S. National Museum

and

S. J. Olsen
Florida Geological Survey


Tallahassee, Florida
1959








DESCRIPTION OF A BERYCIFORM FISH
FROM THE OLIGOCENE OF FLORIDA

By

David H. Dunkle
Associate Curator, Vertebrte leontology
U. S. National Museum
and
Stanley J. Olsen
Vertebrate Paleontologist
Florida Geological Survey


n 1926, the Flord State Road Department engaged
m mas ig a road-cut immediately east ofthe Chpola ,iver,
near riama2, in Jackson Couty. During the course of
this operation a n-ber of fos.lized fish remains were
recovered, of which one was described as a new species of
utlanus by Dr. William K. Gregory (1930). The unde-
scribed balance of the material was placed m storage,
where it remained util 1957, when the Florida Gological
Survey moved to new and permanent quarters. Amongthese
identified specimens has proved to be an esmple of a
beryciform fish.

The n-erous foss-l and liv-g genera which are as-
signed to the Order Beryciformes (Berg, 1940), although of
widely variant structure, diverse adaptation, and consequently
complex taonomy, are quite generally treated as a natural
assemblage. Great phylogenetlc importance has been at-
tached to the group (Gregory, 1933; Reg., 1911, nd Starks,
1904; among others) because of the common possession by
its members of a combination of characteristics inediate
between primitive "ssospondyl and advced spiny-rayed
teleosts. The beryciform fshes were well established and,
it is evident, of wide marne distribution around the world
by upper Cretaceous times. Notwithstanding, the details of


I Fblishedwaththepermssion of the Secretary, Smith-
soni.n Institutlon 65369






S FLORIDA GEOLOGICAL SURVEY

theIr origin andradiation remain obscure and, m fact, their
fossil record in the Western Hemisphere as negligible. It is
on this bass that the unique specmenfromFlordawarrants
the followng description.

The authors acknowledge their appreciation of Dr.
Robert O. Vernon, Director, Florda Geological Survey,
for affording the opportatty of this study. Dr. ,onard P.
Schultz and staff, D-lv-son of Fishes, U. S. National Museum,
and Dr, P.E. Cloud, Jr., Dr. Roland W. Brow, d Miss
Ruth Todd, U S. GeologIalSurvey, have been most generous
of assistance and advice Much of the compilation of the
comparative data employed has been carried out by Miss
Danne Hubbard, and the allustratons havebeen prepared by
Mr. L. B. Isham and Mr. Jack Scott


DESCRIPTION

The dehcate specimen, Identified by the Florida Geo-
logical Survey No. V-5776, and delving from thelower prt
of the Oligocene iaranna himestone, lacks parts anteriorly
and anteroventrally of the head, and of others dorsally alg
the left flank. I, however, -s preserved without either ap-
preclable compression or distortion and comprises about
two-thirds of an individual, which, whencomparedwithholo-
centr1d beryciformes, is of relatively elongate body habit.
The dorsum and venter are not longitudinally keeled, and in
transverse sectlonthebodyis of regularly ovate outline. On
t . . .. I gl .


estimated, standard length, 250; head length from rostral
tip to posterior border ofopercular spine, 74; and mamum
body depth (not exceeding). 96

The skull is indicated to have been somewhat longer
than deep. If the above approx-mations of si.e are correct,
the head, relative tothe standardlength, is small, and in this
character the fossil compares more closely to Holocentrus
andMyr st than toother related Recentgenera. Impres-
sion on the trzx of the orbital surfaces of the parethmods
denotes breadth o snoutfully equal to the postorbatll wdth






SPECIAL PUBLICATION NO. 2 PAPER NO. 3 3

of the skull (fig. 1).

Large frontals cover practically the entire, trs-
versely convex skull roof The elements are joined m. a
gmeda Longttudialsuture whlchis depressed Ito a groove,
Strong r dgeabelarg alewposterlorly radatg anddentate
ribs and moderately deep mucous channels trk the dorsal
surfaces of the bones, he pIaretals are small, posteriorly
restricted, and widely separated from each other. nly

tsupra-occiptal could have been exposed laterally and pos-
teriorly in My dorsal view of thae skull

The occipital face of e neurocraniu extends down-
ward and backward from the posterior edge of the skull roof
an angle of approximately45 from the horontal (fig. 2).
The skulls thus prolonged in a spino-occipltal development
greater than that displayed by the skulls of hvng relatives
with which comparison has been made. The supra-occlptal
crestls produced backwardlyands neither elevateddorsally
nor eendedanteriorly Inthe mdhne of the roof. This fact,
coupled with evidences of deep and completely roofed post-
temporalfossae, suggests stronglythat the cranal insert-ons
of the epaalbodymusculature were confinedto the posterior
surface of the skull. Insofar as canbe observed, the foramen
agnum is bounded ventrally by a vertebra whxch s firmly
nkylosed to the more anteriorly lying bas.-occpital. The
exocclpitals, which do not contribute to the formation of the
occipital condyle, form the lateral margins of the same
endocranlal opening and if, in fact, they do not meet each
other dorsally m the mldhm, only the narrow inferor ex-
treaty of the supra-occpital spine can intervene. The
ventrolateral extent of the exoccaptals Is unknown. How-
ever, ndorsalaspect, because of the spno-occipital devel-
opment of the skull, the posteroventral parts of each 12e
horizontally on the underlying bones and appear narrow and
attenuated. Anteriorly the exocc.pitals gradually -den and
flare dorsally in a concave posterolateralydlrected face.
Remanns of both eplotic elements are preserved Those on
the r-ght side comprise the broad, trangularly exp-ded,
post-temporal process and the more anter-or dorsolateral
amna, which meeting the dorsal wing of the pterot-c, form
the posterior port-on of the roof of the post-temporalfossa.
The fragment of the left eplotlc Is that anterovertical part
const.tutng the meslalwall of the left post-temporal fossa.






FLORIDA GEOLOGICAL SURVEY














Figure 1. Holocentrtes ovals Conrad. (Referred spec-
imen. Flor1da Geologcal Srvey No. V-5776.)
Drawingofskull, aspreserved, dorsalvyew.
Reproduction appror tely X2.

Explanation ot bbrev2atLons. Ex, exoccpital PFr, frontal;
Ifr. infra-orbltals; Pa, parietal; Socc, supra-
occipital, Sphot, aphenotic; Sov, spo-occipital


the supra-orbital and infra-orbltal sensory
hnes; andptf, post-tempralfossa.






SPECIAL PUBLICATION NO. 2


PAPER NO. 3 5






6 FLORIDA GEOLOGICAL SURVEY
















Figure 2. .. .. red spec-
V-5776. )
Attempted restoration of occpital face of skull.
Reproduction approx- -tely X2




pterotlc, sc, supra ocspltal. S5, splno-
occptalvertebra, fm, foremen mgnum, f. son,
foramma for spino-occipial nerves; and ptf,
post-temporal fossa.






SPECIAL PUBLICATION NO 2 PAPER NO


fmN' Soy






S FLORIDA GEOLOGICAL SURVEY

Thi conhtion of this remnant and inclination of adjacent
exoccupItal part lndscate that the fossae openedbackwardly
-,th sharp meal and ventral hps I all the hlng holo-
centrlds examined. the losses are posteriorly deepened and
teir floors are contnuedas grooves, convergent toward the
foremen magnum, across the posterodorsal faces of the
exoccpltals (fig 3).

The exposure of the ntercalar (op-sth.otl of Nelson,
1957)on the poster ace of the skull is Lmited to the proc-
.,es serving the ventral point of attachment for the post-
temporal bone. This s apparently situated belowthelateral
third of the width of the post-temporal fossa. The element
does not possess a mesally directed spihnt apphed to the
exocclptal. Laterallythe Intercalar extendsforwardacross
fully one-half the length of the otto region, foorng a deep
concavity immediately below the artcular facet of the
hyomandibular. Similar butless welldeveloped depressions
SHolocent rus ando display in their ante arrts
the posteror open of the jugular. c l andthe foremen for
the glossopharyngea nerve, and their anteror and dorsal
walls serve as areas of orIgln for at least the adductor
hyomandlbularis and adductoropercul muscles.

The dorsally and anteriorly succeeding pterotic,
sphenot0c, pro-otzc, and alsphenotd bones in the specimen
are either damaged or poorly exposed. However, the doro-
laeral flange ofthe pterotc andthe hghanterodorsal amina
of the'sphenot-c encompass a relatively large, lateral tem-
poral fossa which provides or0gn for the dplator operculi
and levator palate muscles. As the region is exposedonthe
left side, the lateral wall ofthe jugular canal and trgemgno-
facahls chamber Is missing, but the pro-otc is mnlated
posteroventrally and forms a port-on of the anterior wall of
the audtorybull The out-t urnedp of ths part ndcates
a flat, laterally facing, membraneous area external to the
saccular region.

Two elements of the infra-orbital ser-es of bones are
preserved -ntact on the right side (fg. 4). These, situated
below the orbt, arerelativelyarge and doivdeddnto shallow
extern- l and deep internal lama by the anteroposterior
passage of the broad nfra-orbltal mucous chapel across






SPECIAL PUBLICATION NO PAPER NO 3 9

their surfaces The margins of these laminate are only
weakly dentate

Of the visceralskeleton, the possession of two supra-
m llary bones s show. Although lying In approximately
correct position on the left side of the specimen, t must be
presumed that these elements are displaced from tne right
and as thus interpreted display the same configuration and
relationships as the corresponding supra-maxillaries of
. The hyomadlbula has double articular heads
ad the th m anteriorlamella ofthebone is thus farlybroad.
The characterstc external rib of the bone s well developed,
and situated near the posterior border of the bone. The

vertical porton of this rb below the level o the operula
process pro3ets lateral n a practically true transverse
plae. Canals for the passage of branches of the hyomandi-
bular nerve emergebothdorsallyandanteriorlytothis ridge.

The preoperculum Is a short, exceedingly deep ere scent
of bone.whch displays neither a sharpilorward angulation of
Its mter1orpart nor a posteror spine. The thlckness of the
anterIor border of the bone complments the transverse in-
lination of the vertical hyomandabular rib andthe posteror
margins of the lam1nae bounding the vertical preopercular
mucous channel are finely dentate throughout their extent.
A smgle vertical series of scales underlies the posternor
arg-n of the preoperculum andoverlies the anteror portion
of the operculum me longitudinally ornamented expanse
of the operculum behind these scales is deeper than long
with generally concave, dentate margins above, and below
posterior spinous process.

The remains of at least eight spinous rays of the dorsal
fn are to be observed. Of these the foremost one preserved
is the largest. These facts denote a long-based sp.nous
dorsalfin composed of I0 or more rays orig-nating aboveor
shghtly beh-nd the posterior arg-n of the operculm. As
n the Recent the individual spines are balater-


and then the other

Scalaton was apparently complete over the body and
cheek. The scales, although somewhat variable In size






10 FLORIDA GEOLOGICAL SURVEY














Figure 3. Holocentrtes ovalis nrad. (Referred spec-
imen. Florida Geologzcal Srvey No. V-5776.)
Drawing of skull aa preserved from the left
de. Reproduction approximately X l


Alp. p fractured edge of ali-
exoccipltal; c, mtercalar;
phb, pharyngobranchials; rot, pro-otic;
Pter, pterotic; Snx, supra-Mxillary; Sphot,
sphenotlc, ab, auditory bulla; f. j, posterior
opening of jugular canal, L son. forIama for
,spno-occipital nerves, f. IX, 1oramen for
glossopharyngealnerve; se, scales; and s.socc,
spurous process of supra-occzpatal.





SPECIAL PUBLICATION NO 2 PAPER NO 3 11











Asocc
Sphot 'f
AlPr




af,.^ : *, -






IS FLORIDA GEOLOGICAL SURVEY











Figure 4 -_"-- --- -- Iteferred spec-
-, - I, No. V-5776.)
Drawg f the skull as preserved, from the
r-ght side. Reproduction approx lately X 11.

r, plot-c- Ft. f-ontal,
fra-obatal; Op,
opercular, a, parietal, Pop, preopercular;
1Pt, 'post-temporal, Per, pterotic, Sphot,
sphenottc, f. hyo VII, foramen for hymandi-
bular branch of facial nerve, Itf, lateral-
temporal fossa; me mucous cha nnel
associatedith resp 1i'i thV e supra-orbital,
nfra-orbital, andpreopercula sensory canals;
and ptf, post-temporal fossa.









P-.a


"II


SPECIAL PUBLICATION NO. 2 PAPER NO. 3 13






14 ILORIDA GEOLOGICAL SURVEY

regionally, are generally large Eight occur in the vertical
rows between the venter and longitudmal lateral lne sequence
below the spnous dorsal fin. divduald examples from the
anteror flank measure approximately 17. 5 m-llhmeters deep
by 12 0 mllimeters long b outline, dorsal, posterior, and
ventral borders are confluentandperformabroadly rounded
curve. The basal margin is of low forward convexlty, and
In consequence the dorsal and ventral basal angles are well
defined The nucleus is a small, vertically ovate locus sit-
uated slightly nearer the dorsal and apical margns of the
scales than to the ventral and basal ones n anterior body
scales the nuclear fields are unornamented. Posterlorly
they appear larger i proportion to totalscale area and may
be marked by mode rately coarse and entirely vertical cxrcuh.
Surroundng the nuclei, fine. crescent4cally arranged, ad
weakly incised circus cover the dorsal, basal, and ventral
quadrants of the scale. All appear to parallel the scale mar-
gbns No basal radii are present but suggestive ofthese are
five or six regular series ofph2cations 2n the courses of the
crculh, wh1ch d1vergefromthe nuclear area acrossthe em-
bedded anterior scale part. Posteriorly most of the circul
stop along the apcal diagonal. However, one of every third
or fourth crosses these dagonals onto the apical field where
they curve obliquely away from the longitudinal axis of the
scale toward the apcal perphery in radial fashion The
posterorly exposed apcal field displays a nuclear angle of
nearly 180'. Parallel and longtudhnally directed grooves
(numbering about four in one m.llmeter) traverse the ex-
ternal surface of the part. The ribs between the grooves
project as teeth along the posterior border of the scale and
their sides may be obliquely striated.


RE-RKS

The presently described fish must be considered, at
this tame, of the same identity as the single holocentrd from
the Ocala limestone of Jackson County, Florida, described
by Conrad (1941) -der the name Holocentrites ova1s. To


Howeverno attempt to determine the validity and significance
of these seems feasible It is only with the greatest dlff -
culty that the exact knd and degree of variation between






SPECIAL PUB CATION NO 2 PAPER NO 3 15

completely and 0 differently preserved specimens can be
estabhahed. In my event, onthebasl of current knowledge,
Holocentrites Is dlstmngusshable from all other fossil and
-.00 ro. 0 f,' g-, h, f .11 t0 ', ,


Romer, 1945, Starks, 1904; Woods, 1955, Woodward, 1901
and 1902, etc. ) Further, the new information afforded by
the Olgocene specimen permits revision of the diagnoss of
th0s genus, as follows: A relatively fusiform holocentr- d
with a m-axmum body length of about one-fourth the total
length a vertebral formula of 11 + 14 + 1 = 6. The fish
resembles Mypti the strength of the principal and
fewness of the radiating frontal ridges, the correspondlgly
good defAnition of the mucous channels, the expanded post-
temporal process of the epiotic, the protubernt lips of the
laterally directed membranous face of the au.dtory bulla,
the deeply channeled and weakly ornamented infra-orbital
bones, and the simple crescentic and nonspnous preoper-
culum. It differs from f and is slmllar to Holo-
centrus in possessmgabroad, double-headed hyomandlbula;
but is distinct from the two noted huvng genera byvlrtueof
a relatively greater spino-occpltal development and fusion
of a vertebra to the bas0-o1cpital, which prevents the two
exocclpltals meeting each other In the midhne below the
foramen magnum.

The Holocentr.dae date from the early upper Creta-
ceous when such referred genera as Homonotxchthys,
Traohlohthyodeos, and C from Turonian horloons
0m gland, first appear In the fossil record Occurrence
of fimlyrepresentatlvesduring succeeding geologic periods
to the present time aren't frequent but indicate a continuous
ex-stenee. A more complete representation of the group,
the detailed study of these and available specimens, ad
recognition of either morphologically or temporally older
antecedents wouldfacilhtate reconstruction of the phlogenetlc
h-story of the family, In the absence of such object1,e data
however, recourse must be taken in the analysis of certain
obvious osteological characteristices selected because of
avallbihlty to examnatxon on paleontological mnterals.
Two prncipaladaptive conditions are apparent amonglivng
holocentrid. Thoe first ofthese, exemplifiedby My ,pr-.
combines well developed mucous channels; longtudinal






16 FLORIDA GEOLOGICAL SURVEY

crests on the frontal of moderate width and bearing few
rad.at.ng ribs, a nonspInous, crescentic preoperculum
short, high, and single headed hyound bula; and hormal"
premaxillary. The second condition, characterized by
Holocentrus, displays much reduced mucous channels;
prncpairontal crests very broad th numerous radiating
ribs, a spnous, sharply angulated preopeiculum, broad,
double-headed hyo..ndibula; and prema illary w1th highly
developed ascending process. While, as impled n the
literature, Holocentrus may be more advanced toward the
"true percolds" than M by reason of the splnous
preoperculum and more protrusxle premaxilary, the two
adaptations can be traced without question back through the
Tertiary nto the Eocene The upper. Cretaceous holocentrd


ma4llary with low ascending processes. The phylogeny of
the Holocentridae consequently has been pictured (Conrad,
1941) as an early Tertiary divergence of the myriprstme
and holocentrne 1hneages from a single holocentr1 d stem of
Cretaceous origin.

Homonot-ehth s f s too poorly -o4 for comment, but
in degree of development of mucous channels and frontal
crests Trahchthodes and Holocentrates are not radically
d 4fferent from M s and C114 4 would seem to
approach olocentrus. Developmentof thefrontalbones may
be far too111 adequate a crteonbut,n contrast tothe above
phylojenet.c concept, a basis is suggested for recogniing
the dlstlnctveness of the two "subfamhal"' groups for the
entire known history of the family By this latter view, the
sngle-headed hyomandlbuae of M and the more
protrusile premaxll11 of Holocentrus would be regarded as
adaptive morphologlc dlastinctlons of the visceral skeleton
whose selective values were established only during post-
ocene times. Some degree of support would seem given to
this opinon(oral commun-cat-on from W.A. Goslne) by the
andHolocentrus. The --
v-sceral complex resultOng
from a single-headed hyomandlbula would be mechancally


g ,l rakers Conversely, Holoeentrus ut4lhzes protrus4 l






SPECIAL PUBLICATION NO 2 -PAPER NO. 3 1

premaxilh for browsingg" on benthonc animals.

Fossil fishes n general are of rare occurrence m
nature. Corroborative experimental datahave notbeenfond
but t has been tdelyassumed that therate of decompoaston
of the dead bodes of these animals Is comparatively rapid
der aerobic conataons. Conversely, while not and.cative
of life habitats, the best preserved skeletons are generally
attributed to reduction environments (cf. Rayner, 1958).

The Marianna limestone, of lower Oligocene age, has
been described (Moore, 1955) as a 15-foot thick layer of
soft, cream to white, masesve limestone. It has a lmated
dastributxon, outcropping over most of Jackson County,
Florida, and occurring inthe subsurface only as far west as
Walton Cunty. l-randVernon(1956)refer to the presence
m the formation ofabtdant .l as index fossils.
From the matrix removed from the presently described
specimen during preparation, iass Ruth Todd, of the U S.
Geological Survey, has packed and identified the followmg
l ast of m acrofossals:

al Caushman?
Bohvna spp.



Cad baaae p (Cushman)
Globigerina spp.
Nodosarla? sp.
Robulus sp.
N aada sp.
Sparoplecstamina sp.
aaaaava a s Cushman and Ellsor?
Ni..-a sp.

About these Mss Todd has stated (in httoris). "Four
features of this foramiferal assemblage point to its having
been deposited moderately deep water(suchas 10 fathoms
ormore):(a)presene of globigerds, (b)type of benthonac
genera, Recent representatives of which are more hkely to
occur in deep than in shallow water (Rab Nodosaria,
UvPgerina, Agulogerina, pada), (c)absenc of malihlds






18 FLORIDA GEOLOGICAL SURVEY

usually considered to ndcate shallow deposition; and (d)thin-
walednaureof the test n contrast tothe heavy-walled tests
commonly found in shallow water -der the effect of wave


To be pointed out also Is the fact that the foramnf-
eral assemblage undoubtedly indicates welloxygenatedbottom
water. Under these circumstances the geological environ-
meIt of the originally wel1 preserved specimens of Holo-
entrltes and associated fishes Is of onSderable inter .

One plausible explanation was suggested by Dr, Preston E.
Cloud, Jr. (oral cou icat.on), who called the attention
of these writers to the relatxonsh1 p of sedlmentparticle size
to density of anaerobitb.ctemlalpopulations below the sed-
ment water interface and the amount of gaseous interchange
between water and the substratum Entombment of these
fishes n a segment as fine-grained as the Mriaa hme-
tone might well insure anaerobc preservation. Re ent

secrete themselves in practically any available crevice or
Berg, (LI4g14p ..d 1 1 ,. 4 4, 1941) t.
deprelon by dy and toforage forood by night. The snap-
pers also are reported(.b6dem)of somewhat similar habits,
dlng i schools in protected situations during the day and
feeding sngly at .nght. I similar behavioral patterns .re
displayed by the fossl representatives of these two groups,
the chances of burial, although perhaps accidental in every
case, would be much enhanced.


REFERENCES

Arambourg, C. (also see Bertn, L.)
1917 Ls po ss.on a foselle d'Oran: hat. P. Crte
4ol. 1'Alg ne, Ser 1, Palont. no. 6, p. 1-
298, 48 flgs., 44 pls., 7 tables.


1940 Clas sificationof flashes, bothRecentand fossl:
Trav. 4's1t. oo1. V'Acad. Sci., U. S S. R.
v. V, no. 2, 517 p., 188 figs.


Bert958 (and Arambourg, C.) Super-ordre de T-14o-
st4ens. Tralte de Zoologie, v. VII, Fas. III,
p. 2204-2500, figs. 1561-1788,






SPECIAL PUBLICATION NO. 2 PAPER NO. 3 19

Conrad, G. Mile
1941 A fossil squirrel-fish from the upper Eocene
of Florida. Florida Gol. Survey Bull. 22,
12 p., 3 pls. I fig.

Gregory, W.K.
1930 A fossl teleost fish of the snapper family
(Ltianidae) from the lower Oligocene of Flor-
ida Florida Geo. Survey Bll. 5, p. 7-17.

1933 Fish skulls. A study of the evolution ofnatural
mechanisms: Am. Phxlos. Sc. Trans., new
ser., v. XXIII, pt. 2, p. vi,, 481, 302 figs.

Hildebrad, S.F. (see Longley, W.N.)

Longley, W.N.
1941 (and Hldebrand, S. F.) Systematc catalog
of the fishes of Tortugas, Florida: Carnegie
nst. Washington Pub. 535, p. x-, 317, pls. 1-
34.

Moore, W.E.
1955 Geology of Jackson Couty, Florida: Florida
Geol. Survey Bull. 37, p. 1-101.

Nelson, E. M.
1955 The morphology of the wim bladder and audi-
tory bulla in the Holocentndae: Fzeldiana:
Zoology, v. 37, p. 121-130, 3 pls.

Pauca, M.
1931 Neue Fische aus dem OligozA von Patra-
Neamt. : Acad. Roumane, Bull. Sec. Scent.,
v. XIV, no. 1,2, p. 29-34, 4 figs.

Puri, H.S.
1956 (nd Vernon, R. ) A su ary of the geology
of Florda with emphasis on the Miocene de-
posts and a guidebook to the Mooene exposures:
Florda Geol Survey, G. S. A. Feld Trip, p. I-
SS.






0 FLORIDA GEOLOGICAL SURVEY

Raner, D.H.
1958

rented to D. M. S. Watson, Lndon, p. 129-
156.

Regan, C. T.
1911 The anatomy and classficatonof the teleostean
fishes of the orders Berycomorphl ad Xeno-
beryces Annals and Mg. Nat. History (8),
VII, 1-9. pt. 1, 2 figs.

1929 dishes: Encyclopedia Britannica (14), IX,
p. 305-328.

Romer, AS 5.
1945 Vertebrate Paleontology, Ed. 2, p. vni, 687,
377 figs.

Starks, E.D.
1904 The osteology of some berycoid fshes: U. S.
Nat. Mus. Poc., XXVII, p. 601-619, 10 figs.

Vernon, R.0. (see Puri, H.S.)

Woods, L.P
1955 Western Atlant-i species of the genus Holo-
centrus: FIeidlan: Zoology, v 37, p. 91-
119, 18 figs.

Woodward, A.S.
1901 Catalogue of the fossil fshes in the Brtlsh
Museum (Natural History). pt. IV. Printed by
order of the Trustees. London, XXXVI1,
636 p., 19 p1s.

1902 The fossil fshes ofthe EnglishChlk: Paleont.
Soc. Pub pt. p. 1-56, 13 pls.




21938 '