Occasional papers of the Museum of Natural History, the University of Kansas

UNI

HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

OCCASIONAL PAPERS

of the

MUSEUM OF NATURAL HISTORY

LIBRARY

The University of Kansas

T _ JAN 6 1975

Lawrence, Kansas

HARVARn

NUMBER 33, PAGES 1-19 DE^VffiSlTTe, 1974

POSSIBLE FUNCTIONS OF ORNAMENT
IN LABYRINTHODONT AMPHIBIANS

By

RONN W. COLDIRON1

Potentially both endochondral and dermal bone can have sculp-
tured surfaces. Such ornamentation (which for simplicity will be
called ornament) occurs on the dermal skull roofing bones, dermal
parts of the pectoral girdle, and neural arches. The rhachitome,
Cacops, has ornamented dermal plates capping the neural arches.
Ornament appears throughout the Labyrinthodontia. Within this
group the dermal skull roofing bones invariably are sculptured.
Ornament frequently covers the dermal girdle but only rarely the
neural arch. In the more advanced cotylosaurian reptiles, ornament
becomes less evident. Among modern vertebrates, dermal sculptur-
ing is common in the crocodilians and some frogs. Some turtles and
fishes possess ornamented dermal bone, but is rare.

Nearly all types of ornament can be put in two main categories —
random, closely packed pits and interconnecting ridges, and longer
ridges and furrows oriented radially or longitudinally to a center
of ossification. Other forms occur but are rare: isolated pits sunk
into a uniform surface; long, random ridges; rugose, horny projec-
tions. Within a single species nearly any combination is seen so
that, for example, Eugyrinus wildi has individual pits and radial,
asymmetric, and longitudinal ridges and furrows.

There are two opinions on the function of this ornament. Bystrow
(1944, 1947) suggested that ornament is associated with vascular
canals in the dermal skull roofing bones and hence aided in cutane-
ous respiration. Romer ( 1947 ) and others have proposed that orna-

1 Department of Systematica & Ecology and Museum of Natural History, The
University of Kansas, Lawrence, 66045.

2 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

ment provided a cranial structure to which the skin may be tightly
bound.

Romer's view very well may be true. However, the different
types of ornament, its ubiquitous presence on the skull, and the fact
that the skin is just as tightly bound to smooth bone as to the orna-
mented portions in modern crocodilians requires a further ex-
planation. Bystrow explained the function of ornament on the basis
of its internal structure. In the "hydrophilous" labyrinthodonts he
divided the dermal bone into three layers — lower, middle, and
upper. Each of these layers is penetrated by the normal Haversian
canals, which anastomose with one another. Thus, the canals of
the upper layer communicate with the larger Haversian canals in
the middle layer and open as foramina to the bone surface at the
bottom of each pit. He concludes that the Haversian system carried
a small artery, larger vein, thin branches of nerves, and lymphatic
vessels.

Bystrow termed this Haversian system a "Rete Vasculosum"
which appears only in association with the sculpture. Therefore he
correlated the sculpture with a vascular system supplying the skin
and hence with cutaneous respiration.

However, the correlation between "round-celled" sculpture ( ran-
dom pits and ridges) and cutaneous respiration does not hold for
what he called the "xerophilous" labyrinthodonts. In places these
have ridges and furrows instead of the "round-celled" sculpture of
the "hydrophilous" forms and no "Rete Vasculosum" (and thus no
cutaneous respiration). There is no neat division of sculpture re-
lated to a given species. As mentioned earlier, an individual can
have two or three different types of sculpture. Therefore, although
a correlation exists in that wherever cutaneous respiration is im-
plied by bone microstructure there also exists "round-celled" sculp-
ture, the reverse does not hold. It does not hold because both the
"round-celled" and ridge and furrow types of ornament exist on
the same specimen and yet no "Rete Vasculosum" exists. Hence,
the presence of sculpture, regardless of type, does not imply the
existence of cutaneous respiration. In Benthosuchus, a "hydroph-
ilous" form, Brystow associated a respiratory function to the
clavicle ornament, yet this ornament is the ridge and furrow type.
In the "xerophilous" forms, however, the ridge and furrow sculpture
has no "Rete Vasculosum" and thus has no association with cutane-
ous respiration. Clearly, the correlation between ornament and
cutaneous respiration does not hold.

Other difficulties exist with his interpretation of the bone micro-
structure (canals). The irregularity and complexity of the canals
within the bone seems to provide a circuitous and inefficient route
to the ornamented surface and to the skin. Furthermore, the
canals are arranged horizontally to the bone lamellae. If blood were

ORNAMENT IN LABYRINTHODONT AMPHIBIANS 3

to be directed to the surface, a more adaptive route would be via a
less complex canal pathway. It is unlikely that random anastomoses
between canals is related to a systematic connection to the outside
epidermis. Also, if the main function of the canals is to carry blood
vessels, those of the lower bone layer feeding into the larger middle
layer canals would cause a drastic decrease in blood pressure. This
would be very inefficient for rapid blood flow and efficient oxygen
exchange at the skin surface. Finally, the large Haversian canals
of the middle layer are found in the ornamented dermal bone of the
"xerophilous" forms which do not have an upper layer "Rete
Vasculosum."

Romer ( 1972 ) leveled the most striking criticism of Bystrow's
theory. Romer suggested that cutaneous respiration is a degenera-
tive characteristic of modern amphibians. Within this group, blood
is supplied from a pulmocutaneous artery (frogs) or from smaller
arteries distributed more evenly over the body surface (salaman-
ders). It is generally concluded that the cutaneous blood supply
does not pass through bone. The ancient amphibians probably re-
lied on lungs for respiration. The fossil record shows lungs in
Bothriolepis, a placoderm. Thus, lungs were not a new and unim-
portant development at the labyrinthodont stage. Furthermore, the
presence of a well-developed thoracic basket makes aspiration
breathing probable in the labyrinthodonts. Finally, the labryintho-
donts most likely had a full covering of dermal scales. In various
labyrinthodont groups, as better specimens become available, der-
mal armor is being discovered on the dorsal surface (Baird, 1964;
Bystrow, 1944; Carrol, 1969; Colbert, 1955). The scalation in Sey-
mouria, well-developed dorsally, suggests that scalation was not a
specialization of isolated amphibian groups but more probably a
primitive character which persisted well toward the reptilian level
of evolution.

If ornament strengthens bone by resistance and diffusion of
stress, then one can explain its presence wherever it occurs. Changes
in ornament relative to different skull sizes, dimensions, and jaw
musculature support ornament as a strengthening adaptation. Fox
(1964) has proposed that ornament may strengthen bone by rein-
forcement. In studying the cheek region of Captorhinus, he noticed
that ornament is alligned parallel to presumed directions of stress.
Where less stress was assumed to exist, bone is thinner and the
ornament is random. Fox's functional interpretation of the orna-
ment is supported by its orientation not being merely a result of
differential bone growth. Different ornament orientations are laid
down in specific and predictable areas.

Random sculpturing (pits and ridges), by diffusing stress, serves
an active rather than a passive role. The rest of this study will
analyze and hopefully support a strengthening function for orna-

4 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

mented bone. Four experimental approaches are taken. The first
is called a "split-line" technique in which the collagen fiber orien-
tation in a modern crocodilian is exposed. This shows the distribu-
tion of stress throughout the skull. The second experimental ap-
proach is a histological study of bone and its surrounding tissues,
again in a modern crocodilian. This study confirms criticism of
Bystrow's hypotheses. The third is a series of thin sections of
Eryops dermal skull roofing bones at various depths. This reveals
the collagen fiber orientation and establishes an anolog to the "split-
line" technique of the modern crocodilians. The last is a review of
30 genera of labyrinthodont amphibians with respect to skull pro-
portions and intensity of ornament.

ACKNOWLEDGMENTS

I would like to thank Dr. P. P. Vaughn and Dr. R. Molnar for
materials and advice on approaches to the experiments. Drs. T. H.
Eaton, L. Martin, and Linda Trueb deserve many thanks for
further advice and suggestions. Mrs. Fox of the UCLA Rehabilita-
tion Center was most helpful for preparation of the alligator thin
sections. Drs. E. C. Olson, T. H. Eaton, and Linda Trueb offered
helpful criticisms of the manuscript. Lastly, I am indebted to the
American Museum of Natural History for making their collections
available to me.

MATERIALS AND METHODS

The biological materials are as follows. An immature Alligator
mississippiensis was used for the split-line technique. The length
from snout to occipital condyle was 16 cm. For the histological
study, a formalin-preserved head, 7.5 cm in length, of Caimon
sclerops was used. Because different taxa are involved, positions of
ornament differ. However, absolute position is important only in
the mechanical studies, where ornament position affects skull trajec-
tories brought out by the split-line technique. In the histological
studies, ornament is considered only in terms of surrounding tissues.
Dermal bone fragments from various parts of the skull of Eryops
were used for thin sections.

The materials and methods involved in the split-line technique
are discussed thoroughly by Benninghoff (1925). A more general
discussion is found in Tappen (1953) and Seipel (1946). A brief
summary follows: 1. — Fix specimen in 10% formalin for one week;
2. — decalcify in 5% nitric acid, the acid being stirred occasionally
and changed once a day (usual time is three clays followed by one
or two days of washing); 3. — bleach in 3% H202 for one day to
facilitate later photography; 4. — dessicate in alcohol for three days
(60%, 70%, and finally 80%) and preserve specimen in 80% alcohol

ORNAMENT IN LABYRINTHODONT AMPHIBIANS 5

during time of preparation; 5. — puncture the surface of the bone
with a sharpened teasing needle; 6. — inject with a hypodermic
needle india ink into the fractures. The fractures occur parallel to
the collagen fibers. Hence the split-line technique reveals the
orientation of the bone microstructure. The result is the formation
of trajectories along the bone surface when the small fracture lines
are joined.

The method used in sectioning, staining, and embedding are
systematically covered in Geyer ( 1936 ) . The sections were cut to
1 cm- X & inch thick, and the celloidin method of embedding was
employed. A triple stain of hematoxylin, eosin, and azure was used
to differentiate the tissues.

The Eryops bone thin sections were made as follows: 1. — Mount
dermal bone fragments on a glass slide with Lakeside, a thermal
plastic cement; 2. — grind specimen to the desired level with water
and coarse to fine grit and powder ( coarse, #400 grit silicon carbide;
medium, aluminum oxide No. 9M; fine, aluminum oxide No. 3); 3. —
reheat cement, turn specimen over, and repeat procedure (after
each phase of grinding the slide should be washed thoroughly with
fast running water to remove all grinding material); 4. — place cover
slip over the finished specimen using Canada balsam as a fixative.

Data for the 30 genera of labyrinthodonts came from the litera-
ture (see Appendix) and from specimens at the American Museum
of Natural History and Museum of Natural History, The University
of Kansas (see Specimens Examined).

RESULTS

The alligator dermal skull roofing bones exhibit three general
collagen fiber orientations by the split-line technique. First, the

Fig. 1. — Composite view of skull of Alligator mississippiensis. Below, loca-
tion of ornament; above, stress trajectories, where the dashes represent fractures
in the hone brought out by the split-line technique.

6 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

fractures along unornamented sections of bone are linear and
parallel, running antero-posteriorly along the skull. Second, the
lines converge toward ornamented areas. Third, within orna-
mented areas, lines of stress (trajectories) do not appear. In-
stead, either the bone fractures non-linearly in random directions,
or there is no fracture but only round puncture holes.

Specifically, there are five main ornamented areas around which
trajectories converge ( Fig. 1 ) . The two main areas are ( 1 ) at the
anterior end of the maxilla directly over the largest tooth and (2) on
the squamosal. Two of the smaller, less ornamented areas occur at
the prefrontal along the medial orbital boundary and on the jugal
at the lateral orbital boundary. The least ornamented areas occur
on the prenasal at the lateral narial border.

Trajectories originate from the most anterior teeth and converge
on either the prenasal or the maxillary ornament. A second group
of trajectories run parallel to the more posterior teeth (past the
sixth tooth). These parallel trajectories converge on the jugal
ornament (lateral orbital border) and originate from the main
ornamented area at the anterior portion of the maxillary. More
trajectories from this maxillary ornament run anteriorly to the
prenasal ornament and posteriorly to the jugal ornament (medial
orbital border). The medial skull trajectories deflect toward the
maxillary ornament and run parallel between the orbits and radiate
out into the heavily ornamented squamosal (skull table).

epidermis

r

dermis <

s=— ^__cornified
M cells

cs

:%-yy.<-i-y.i*>. <*>•<■>:•:

■>!■••+• •'. "■»" --

~ ...... . . j«.

vessel

connective
tissue

haversian
canal

bone

Fig. 2. — Cross section of Caimon sclerops maxillary showing the relation
of ornamented bone to dermal and epidermal vascularization.

ORNAMENT IN LABYRINTHODONT AMPHIBIANS

ornament
pit

haversian
canal

mm

superficial
section

deeper
section

Fig. 3. — Thin sections of Enjops dermal bone fragments showing the
orientation ( small dashes ) of lacunae within the bone.

The histological sections of the ornamented maxillary (Fig. 2)
and squamosal exhibit four important features. The bone has very
large Haversian canals in what Bystrow (1947) calls the middle
layer. Above and below this layer no canals are present. The
vascularization of the subcutaneous connective tissue and smooth
muscle above the bone is slight. Finally, above the epidermis lies
relatively thick cornified epithelium. The cells are flattened, com-
pact, and apparently dead throughout this layer.

Thin sections of ornamented dermal skull roofing bone in Eryops
(Fig. 3) show different collagen fiber orientations according to bone
depth. Collagen fibers themselves cannot be seen. Their orientation
is inferred from the long-axis orientation of the bone lacunae. In
the lower and middle layers, collagen fibers parallel one another.
Where ornament interrupts the upper layer, the collagen fibers are
non-parallel and random.

Ornament rugosity within different labyrinthodont lines shows
a general trend relative to the ratio of skull height to skull width
(measured at the occiput), and shape of the subtemporal fossa
(Table 1). Two trends are present: (1) with wider and longer
subtemporal fossa the ornament is usually more rugose; and (2)
with a smaller height/width ratio of skull the ornament is most often
better developed. These trends are not apparent in individuals of
only slightly differing characters. However, large differences in the

OCCASIONAL PAPERS MUSEUM OE NATURAL HISTORY

CO

1-1

M
GO.

O

Q
O

3

o

C/3

w

a

<
>

o

(C
U
H
U

I

U

w
H

Cc/j

01

►J o

0) v:

1) li

60-H
1JW

£ 3

C £

^>

•So
> &

- °o

£L O -H

«) «

-2 §H

N a 1

w a

>, CO

.■a s

£ c
aO

£1

■P71

.£f"o

a

r"

p

p

r*

a

o

o

o

a

O

o

on

o

co

o

in

in

i— i

i—i

i— i

i-H

CD

CD CD

be bo

13 13

cp CD

? ?

a a

<LI CD

CD 0)

13 T3

0)

<Li

CD

CD

<M

q
V»

03

C

s

(H
CD

c
be

-C CD

cm

to

3

oj a.

is o

CD 173 5j

C J" O ** C ->

r- CD is CD iS to

= 13 S H3 C "TS

CO

CO

in
co

Co

3

c/i

*0

^

fi

bli

en

CD

O

C/5

«

0

CI

'^

^

—

-IH

W

-C

O

tlJO

to

co

a

1

5b

•4->

1

'5b

03

03

*■£

05

p

p

on

r*
2

"3>

CD

is

CD

J— 1

05
p

CD
>

13
03
O
Ih

0
a

b£

c

13

CD

a
0

V3

a

CD
CD

13

13
03

=

ID
>

15

- —

CD

* **

CD

•*->

CD

13

3

rf

'1

CD

O

£

3

•4-1

X

3

CD

<u

cd

S

_

*r"

>

G

pj

>>

13

!>

03
i-
CD

o

CO

CD __

> CD

CD ~ 13

T3 ^ —

CD

a

CD
13

co

CD

co

in

o
in

o

o

-O.

-2

-Q

-Q

^Q

^2

^O.

0..

o3

o3

05

o3

03

03

03

1 — "

CD

CD

CD

CD

CD

CD

CD

T3

T3

T)

>

>

>

^*

>

>

>

CD

m

CD

O

O

O

O

O

O

O

V3

w

«3

X

<-;

t r-;

<->

^*i

r|

t i~]

^

4J

-w

■4-f

*j2

?j

-w

T3

TS

13

13

13

13

u

13

1

^

CD

CD

K*

^

?

!>

?

?

CD

CD

O

rt

CD

CD

CD

CD

CD

CD

4-1

4->

C

•4-1

4-i

4-1

03

05

H

r-1

"03

ts

n3

«

03

03

05

03

o3

03

03

bJ

03

05

a

a

a

a

a

a

a

a

CD

IS

1
03

1

g

'>

•l-H

"bb

be

•—

Ih

03

03

a

p

P

CD

CD

CD

CD

U

?
^

C

be

13

13

_ ri

-c

M

en

-

^

^

>>

O

CD

CD

'pi

U

13

CD
>

o3

-

n

'be

. «x

. ^

CD

j-"

CD

CD

13

03

75 S

75

£

'$

CD

»** 4-*

O

CD T CD "r

j3 CD 4L CD 4^ CD

o5 a 03 a 03 a

!h o '-1 O '— O

CD -^ Cu-i CD —

i-q CD i-Q CD "C CD

C ^ C -** O ^

r- CD r- CD f CD

C 13 =13 =13

CD

CO

CI

in

o

o

•l"H

s
s

o
x

o

o

to

a.

CJ

e

i

5 .2

■a "5

■3 1

to

C
O

a.

S
en

w

03

-J=

Ph

o rs

w

a
H

5-i

o

Rh

to
to

ORNAMENT IN LABYRINTHODONT AMPHIBIANS

g

g

s

g £ g

g

g

g

g

g

g

o

o

o

o o o

o

o

o

o

o

o

o

o

00

o

O I- t>

^H

10

oo

m

t^

t-

>>

cq

^

CO .-1

|-H

00

tM

CM

eg

rH

.— 1

cd cd

pfi

-Q

-D.

a

03

rt

<o

13

13

0)

CD

13

13

13

>

CD

0)

^

>

CD

CD

CD

o

Cn

KS

o

O

en

en

en

R

3

3

r-

3

3

3

13
03

13

u

13

0

CD

en

en

3

3

13

CD

t/3

3

p

13

12

13 13 13

'"O

13

13

13

13

13

13

•cH

i>

^

">

> > >

>

!<•

">

>

>

i>

g

•^

^

?*

> r> E>

^

^

**

r^

**

»**

CD

CD

CD

CD CD CD

CD

Cv

CD

CD

03

CD

CO

4->

+J

4-> 4-> 4-)

4->

+4

4^

4_*

4-1

nJ

OS

Is

rt rt Cfl

ctf

cd

ca

o3

o3

cS

03

as

03

a

fl rt cC

a

15

oS

03

03

03

o3

a

ft

a

a a a

a

&

a

a

a

a

a

'5b

OS

g

CD

4J
en

a

CD
CD

'5c
g

ctj

g

'to

•—

13

CD

a

J3

g

"5b

13
>>

CD

13

£

rj

03

CD

~o3
>

CD

13

CD

CD
>

CD

a

—

4-"

3

CD

4->

03

(— (

3^

•—

CD
>

13"

CD

a

£*

>

CD

13

'5b
2

CD

"So

2

CD

CD

13

C

"13

5

■ fH
4-»
OS
g

"2
—

s

CD

U

>

'3
>

CD

13

15

T3

a

CD
03

13

a

CD

CD
13

cd"

—

4-'

03

'Si

of

3

"S

3

. «

. «

CD

13

CD

CD

>..2

—

c5

Cd C

CD

13

CD

13

a

ID

—

CX! -T 5 13 X 13 13 3^3 g cd 13 fl
<" * > 3 "S OS "S '3 a c a Ji "3 C

60

03

3

53
a

CD

CD

13

•4-1

03

>H

CD

13

03

3

en

cd"

C

> 3

CD
03

? C 3 U

13

a
s JJ

? 3 13

O

3 c 13 c

O >3 'r* '^

^-> be % bo

-r o3 oS

CD ,-i .ft r*

13 3 cd 3

^13

O ^

4j a t

03 J>> cS

CD _S CD -3

4_i en -t-> en

C8 OJ

•- CD •- CD

J; Mjj tit

^3 13 13 13

.-a

CD

a

O cu

CD

a

CD

CD

13

>.

"a3

4-<
OS

(13

13

c

CD .2
Tj XJ
•r1 o3

^.g

CD

cS

CD

CD

4-*

Ctf

CD

13 13

CD 1> >

a a -

c c z,

13 'q "i3

13

CD

a

CD

^jCar^CD_^;_,r--CD

13

CD

a

o

CD ^Z -5

« a S a id cd osacpo

>^-3

CD
>

13

CD

a
_c

"a;
>

CD
13

CD

ii a3

cs a

13 g
E 13

CD

c

^IJ
1 a

13 >

C CD

g 13

11 k-
z ^

cS ~

2 _5j

— —

o; a
53-2

13 g

2 £

3 13

CO r^

CO

co co t

CD

00

I-

CD
CO

CD

00

CO

O
CO

-a

' —

T—

S

a.
o

tq

O c^ Q

N

o

o

=n a

o e

CO

3

CO

3

CO

CO

S

>v

3

p^a

u

•c:

en

5

E

o

~

—

a

CO

O

&

v.

OS
en

44>

CD

12

O

O

J^

o

u

-~

c»

OS

=

cd

S-,

JZ

ca

CD

tc

>-l

o

CD

CO

-

Z

10

OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

eg

l-l o

CP t/1
0) CD

^ c

I—) Oj

■S o
3 5°

a

o

—

c

"a J3

h

■M

rt

O

a c

a u

■ —

SQ

be

1H

«

-

(9

r

d

cd

w

C/3

<D

(H

s

O

M-l

OCA)

0)

a

4-*

d

c/5

Cfl

o

<

fe.t:

o

—

d

>>

CD

c

CO

d
u

d

E

j=0

HH

~

-

?

d

—

.^

w

V.

A

-u

_M

O

'5

in

O CJ

o

o

o

o o

r-l CO

CD

o

t>

O -H

cxi i— l

i — I

cq

1— (

I— 1 1—1

xj t3

Ws. «<

_o

CT3

.5 v

cJ oj ID
<U > as

-O ? ;r

<d

n3

OT3 £

o .2

-g cu G ns
? "S - G

s 6 'SB

> C -G G
£ S w5 0)

3 G

o

Wi_3

g> S

cu £

■"0 on.

-H >,

"^ <D

in

o
u

CD a

c1 a

o

as

G

a

*i— c

— i
6

g

>

o

H

13

r^

&

O-.

*&0

o

■> — '

^

M

UJ

CD

^

fH

-T5

5

-2

O

be

•~-

T3

C
—

cb"

-^

—

s

£

03

a

G

03

-2

H

E

>

>.

ctj

— 1

a)
IS

—

>.

rt

>

—

CD

*>

—

CD

C

CD

CD

X3
O

W)

bO

be

CD CD

-Q

-Q

XI

X

X

C3

c«

a!

rt

03

TS T)

CD

CD

CD

CD

CD

CD CD

>

>

>

>

>

Crt C/5

O

O

O

O

O

P G

£

G

£

^ ^

CD

CD

CD

CD

CD

CD CD

4-» •+-»

r^

r^

r-*

G

—

O

C

5

O

C

03 03

G

£

G

G

as 03

a a

c

a

2

<D

i— i

CD

-^

ri

-cs

H

>>

5b

IH

!h

CD

rt

>

.—

—

«-h

CD

CD

bo

3

G >

7 O

G «X!

'G S rt

• - r'bb

1/ CD !_

<D

a

o

CD

CD

+j

^>

£8

CD

CD

X3

—

;

13

CO

(M

CO

co oi

esi

co

f-H

<si co

i— t

*—t

(>1

>— 1 i-H

CO

s

en

reospondyls

Eupelor

B

3

Q

0

C/3

CD

(4

CD

s

bJO

o

C3
ft*

•S

CO

o

■to

&

o
£

.2

3
O

£

.a
5

ca

cV3

.2

CO

O

CD

S
W

CD
CO

ORNAMENT IN LABYRINTHODONT AMPHIBIANS 11

above characters show these trends to be predictable in widely
different lines of labyrinthodonts. Ornament rugosity seems to be
independent of other variable skull features such as length of skull,
relative proportions of cheek and snout, degree of kinesis, and width
of inter-pterygoid vacuities.

DISCUSSION

The histological sections show that cutaneous respiration is im-
possible. The top layer of ornamented bone is free of the large
Haversian canals as is the subcutaneous connective tissue above.
Hence, the blood supply below the living epidermis is meager, at
best, and could not serve for efficient gas exchange. Furthermore,
the cornified epithelium prevents any possible exchange that could
take place between the living epidermis and the poorly vascularized
connective tissue below. Frogs, which do respire through their
skin, lack this outer cornified layer so that gas exchange can occur.
Also, the alligator has, at the bottom of many ornament pits, an
opening like that seen in some labyrinthodont ornament. Bystrow
(1947) interpreted these to be openings in the bone for blood
supply to a respirating skin. Clearly, even if these foramina are
associated with blood vessels, they can not be related to skin respira-
tion. All these factors, then, indicate an alternate explanation for
the occurrence and function of ornament.

The results of the histological study must be interpreted cau-
tiously. The dermis of the ancient amphibians could have been ar-
ranged quite differently. In the case of the modern alligatorids, the
ornament shows no correlation with cutaneous respiration. There-
fore, the histological evidence, being negative, is only indirect.

The split-line technique might offer a mechanical explanation to
ornament. Parrallel fractures along smooth bone indicate the gen-
eral orientation of the collagen matrix in a particular region. Be-
cause the collagen fiber orientation is an indicator of stress direc-
tion, the stress is clearly shown to run from either the tooth row to
an ornamented region or from one region of ornament to another.
The fact that the fracture lines run antero-posteriorly between
ornament demonstrates the overall stress direction over the skull.
With the origin of stress at the posterior attachments of the jaw
musculature and at the more anterior teeth, one would expect the
fracture lines to run in a general antero-posterior direction.

Within the ornamented region the disorganized non-linear frac-
tures demonstrate a random collagen orientation. The simple round
punctures indicate that the collagen fibers are oriented perpendicular
to the skull surface. Round punctures would occur in the walls of
ornamented pits and small foramina. The non-oriented collagen
on the other hand, shows stress is diffused throughout that region
of bone. Hence, the ornament acts as a reinforcement against stress

12 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

Melosaurus

Eryops

Stegops

MORE INTENSE
ORNAMENT

>

Acheloma

Chenoprosopus

Trematops

Fig. 4. — Trends seen within the labyrinthodonts concerning skull profile,
jaw musculature, and ornament intensity; (a) shows increasing area of anterior
adductors from Melosaurus to Stegops; (b) shows lower skull profiles (occipital
view ) from Acheloma to Trematops.

by diverting the stress down into thicker bone and by diffusing
stress throughout the ornamented area. Since stress converges on
ornament, fractures cannot form where stress is best defined.

The thin sections of Eryops dermal bone demonstrate a structural
mechanism for diffusing stress throughout ornamented areas. In-
dividual collagen fibers propagate stress because of their greater
organization and density different from the bone mineral. In lower
layers of ornamented bone the collagen fibers are oriented parallel
to one another and stress direction is well defined. This is seen in
the smooth areas of alligator bone, where the fractures are straight
or gently curving and parallel. In the upper layers of Eryops, how-
ever, the collagen fibers are random, as were, presumably, the stress
directions. Ornamented areas in the modern alligator show the
same random stress directions by the split-line technique. In
labyrinthodonts the ornament is distributed over the entire dorsal
skull surface. Thus, ornament could diffuse stress wherever it occurs
on the skull. Therefore stress could not form in preferred areas.

The random pattern of collagen fibers should be expected in
ornamented bone. Exostosis produces sculpturing on bone surfaces
in modern frogs (Trueb, 1973). In this process resorption, second-
ary deposition, and subsequent modification of sculpture patterns
produce all three basic ornament patterns. Because of the extent
of reorganization involved one would expect a more complicated
collagen fiber arrangement than in simple bone. In smooth bone,
layer after layer is laid down in simple sheets without differential

ORNAMENT IN LABYRINTHODONT AMPHIBIANS 13

growth to create ridges or resorption to produce pits. A simpler
but familiar model of sculpturing is seen in the production of
Haversian canals. Resorption and secondary deposition create a
more complex collagen fiber arrangement adjacent to a canal than
in the surrounding bone.

Trends of greater ornament rugosity with larger subtemporal
fossa and/or lower skull profile indicate a correlation of ornament
with labyrinthodont jaw mechanics. Labyrinthodonts have a
distinct jaw musculature which is termed by Olson ( 1961 ) as the
kinetic-inertial (K-I) system. In this system there are two main
divisions of the adductor muscles, anterior and posterior. The
anterior adductors exert the greatest force at maximum gape. At
occlusion the anterior adductors exert no force since the adductor
fossa and subtemporal fossa are on nearly the same plane. The
posterior adductors are most important in holding jaw position and
have some importance in adduction.

The subtemporal fossa often has an anterior emargination ( Fig.
4, a) variably developed. Olson interpreted the anterior emargina-
tion of the subtemporal fossa to be a slot over which the tendon of
the anterior adductors passed. However, in many forms this an-
terior emargination is wide and deep, indicating that it also accom-
modated better developed anterior adductors.

In reptiles, the adductor musculature is divided into three parts
— anterior, middle, and posterior. The middle and posterior adduct-
ors become differentiated, whereas the posterior adductors are
emphasized. The anterior adductors are small or absent. Olson
termed this as the static-pressure (S-P) system. In this system
maximum adductor force occurs at or near occlusion. Because the
adductors are differentiated over the K-I system, the S-P system
is more susceptible to evolutionary modification. The crocodilians
are such an example. They have a secondarily derived K-I system.

Concerning adductor muscle insertion, the K-I system has a
longer and more powerful lever arm than does the S-P system
(Fig. 5a). The resultant angular acceleration is much greater in the
K-I system, as seen in the following analysis (also see Fig. 5c):

K-I: a =

_ XiFiSin^ + 3C2F2sini/r + xsF3shixJ/

I

„ p , _ .x^F'isini// -|- 3c2F,2sini/' -j- ^F^sin^

~T~

where a and d = angular acceleration.

-Vi, x2, xz = distance from the jaw articulation to the point of
muscle insertion on the lower jaw (= lever arm
length ) .

14 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

Fi, F2, F3, and F\, F'2, F'3 — forces of the adductor muscles.

I = moment of inertia

ip = angle at which the adductor muscles attach to the lower jaw.

From figure 5, b and c, one sees that F2 = F'2, Fx = F'3, and F3 =
FY All other variables are assumed to be equal for simplicity (see
below for further comment). F3 and F\ represent the maximum
adductor force. In this simple model the muscles differ only in
position of insertion. Therefore, the difference in angular accelera-
tion will be directly related to x3/.Ti. Examination of table 1 indi-
cates that the anterior emargination of the adductor fossa in some
labyrinthodonts is quite deep. This suggests that the main adductor
force (F3) has a large value for x3. In both the K-I and S-P systems
the posterior adductors are reasonably close to the jaw articulation;
Xs/xi, then, would be reasonably large. The result is a significantly
greater angular acceleration, a at occlusion in the K-I system.

Stress is defined as force per unit area. Because force is directly
related to the angular acceleration, the stress exerted on the skull
also will be significantly greater in the K-I system. This force will,
of course, be greater in those forms with better developed anterior
adductors. Greater stress created with a better developed sub-
temporal fossa might explain the trend toward more intense orna-
mentation.

The skull profile is also important in ornament intensity. The
flatter skull is more susceptible to both compressive and tensile
stress. The type of profile determines the resistance offered by the
skull roof to the occlusal force of the lower jaw (Fig. 4, b). With
a smaller height/width ratio, there is a smaller vertical component
of the resistance force. This bears on the trend of stronger orna-
mentation with flatter skulls. If ornament diffuses stress and
strengthens bone, then more intense ornament may be necessary to
compensate for such a small resistance by the skull roof. Tensile
stress is generated by raising the skull roof relative to the lower jaw.
Watson ( 1951 ) stated that there is a dependent corollary in a large
retroarticular process with flat-profiled skulls. This process serves

S P system

sin ^

Fig. 5. — Different adductor muscle forces (a, Fi, F2, Fs; b, F'a, F'», F's),
and lever arm lengths (xi, x2, and x3) of the K-I and S-P system, and force
exerted l>y an adductor muscle (c, Fsin^ for any angle of muscle insertion, \p,
where ^ < 90°; for ^ > 90° read F sin ( 180°—^ ) ).

ORNAMENT IN LABYRINTIIODONT AMPHIBIANS 15

as an origin for the depressor mandibuli which inserts on the occiput
as high as possible (nearest to the skull table). This muscle is
responsible for raising the skull. Because of the short lever arm
distance between the end of the process and the jaw articulation,
the muscle must act at a great mechanical disadvantage. The force
needed to raise the skull then must be large. The result is a large
tensile stress imparted on the skull table behind the orbits. In the
alligator, which has a well-developed retroarticular process, the
skull table is one of the most heavily pitted regions of the skull.

In forms with large height/width ratios, ornament would not
need to be as well-developed, because a large vertical resistance to
the occluding jaw is present. The change of skull profile and orna-
ment intensity also can help to explain the reduction of ornament
in the cotylosaurian reptiles. The anthracosaurs ( labyrinthodonts )
have high skull profiles and only slight to moderate ornamentation.
The captorhinomorphs and procolophonids, with narrower and
higher skulls, finally dispose of ornament except for a few forms
(Captorhinus and some paraiesaurs, for example). It should be
added that these cotylosaurs had acquired a S-P jaw musculature.

Drawing conclusions from ornament intensity and trends in skull
parameters is tentative. Because many of the skulls could only be
examined by photographs, any direct measurements were impossible.
Only 30 genera of labyrinthodonts were observed; thus, the trends
described here are not necessarily conclusive. Because a dead sys-
tem is being studied, only qualitative differences and comparisons
among and within the different jaw mechanisms can be made.
Other problems exist in comparisons between the two jaw mechan-
isms. In many forms which have a S-P jaw musculature, there is
also a reasonably well-developed coronoid process. A coronoid
process lengthens the effective lever arm of the lower jaw. This
reduces the difference between the S-P and the K-I systems. How-
ever, the S-P jaw musculature is also characterized by differentiation
of the adductors for lateral and antero-posterior movements ( Olson,
1961). The result is a decrease in vertical adductor force (lower
value for sini// in the equation above). By dividing the adductor
force into different components in the S-P system, the resultant stress
on the skull is less than if the adductor force were directed vertically
(K-I system). Thus, there are conflicting factors which make com-
parisons between labyrinthodonts and reptiles ambiguous. The
trends of ornament intensity have been restricted to the labyrintho-
donts where trends in jaw musculature conform to the K-I system.

Many of the conclusions concerning ornament function in laby-
rinthodonts have involved the crocodilian skull. A comparison of
sculptured bone in the two groups is warranted. In the labyrintho-
donts ornament is evenly distributed over all the skull roof. The
ornament in modern crocodilians is widespread over the skull but

16 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

is distributed unevenly; the more rugose regions are usually associ-
ated with thicker bone. Similar irregularities exist in only a few
labyrinthodonts (i.e., Intasuchus) . Also the crocodilians have rela-
tively long, non-linear ridges and furrows unlike most of the an-
cient amphibians. However, the random pits and ridges are similar
in both groups. The uneven distribution of ornament over the skull
and its slightly different form in crocodilians might imply a function
different from that in the labyrinthodonts. However, the differences
in ornament between the two groups are in degree, not in kind.

There is also the possibility that ornament ridges may act to
reinforce bone in a girder fashion, as mentioned by Fox (1964).
Trueb ( 1973 ) also suggested that the ornamentation in modern
frogs may reinforce bone. In all long-snouted labyrinthodonts longi-
tudinal ridges occur in the snout region. Orientation of these ridges
is not a developmental artifact. Ornament is added onto bone
surfaces at early but separate stages. In Branchiosaurus (Credner,
1883) radial ridges appear from the earliest stages of growth to
maturity. The same process of sculpturing is seen in the modern
alligator. Random sculpture appears in localized areas first; at later
stages longitudinal ridges appear in the snout region and persist in
later growth stages. Trueb (1970), studying casque-headed tree
frogs (Hylidae), observed that ornament initially is localized and
then spreads over the bone surface. Olson ( 1951 ) showed bone
growth in Diplocaulus to occur at different rates. However, the
ornamented surface is uniform with random pits and interconnect-
ing ridges. Because ornament patterns are not due to differential
bone growth of perimeter relative to center, the patterns may neces-
sarily act to reinforce long dermal bones by resisting stress in a
general antero-posterior direction. Stress exists in this general
direction, as is seen in stress lines ( trajectories ) of the alligator ( Fig.
1), which is also long-snouted. The same reinforcement of bone
may be acting in radial ridging.

Another possible function of ornament is to prevent microscopic
surface fractures from spreading. Currey (1962) examined thin
sections of partially cracked bone. He found that fractures stopped
at bone lacunae more often than would be expected from a random
fracture pattern. Alexander (1968) stated that stress is concentrated
at the end of a crack. This stress can be blunted if the crack ends
in a rounded cavity. Stress would be redistributed and diffused
around a much larger surface area (the surface of the lacunae).
This same principle may act at the surface where dermal pitting
occurs. Any fracture that starts at or near the surface immediately
would run into a pit. The leading edge of the crack would be
transformed to the larger surface area of the pit. Hence, stress would
be less concentrated because of the larger surface area. Parts of
automobile engines which are under stress are often pitted (shot-

ORNAMENT IN LABYRINTHODONT AMPHIBIANS 17

peened) in order to stop fractures from spreading along the surface
and into the interior.

By assigning a strengthening function to ornament, one can ex-
plain its presence on any structure. As mentioned previously, orna-
ment appears on widely different structures, including the pectoral
girdle (most prominent on the interclavicle) and dermal plates
capping the neural arches. The interclavicle has a wide contact
with the clavicles. From this contact, torsion (stress) from the
limbs would be transferred to the interclavicle via the clavicles.
The unusual dermal caps are assumed to lend rigidity to the
vertebral column. This allows the vertebral column to support the
weight of the animal on land. In this case stress would be applied
to these elements through the vertebral column and by adjacent
plates. Of course, whether the forces involved would warrant the
ornament is speculative.

SUMMARY

This paper presents possible explanations for the function of
ornament on dermal skull roofing bones of labyrinthodont amphib-
ians. Ornament is ubiquitous within the Labyrinthodontia and
appears in essentially three forms: random pits and ridges, and longi-
tudinal and radial ridges and furrows. Previous interpretations of
ornament function have proven either inadequate or incomplete.
The histology and morphology of sculptured dermal bone in both
ancient amphibians and modern crocodilians show that ornament
may strengthen bone by reinforcement and diffusion of stress. Study
of the vascular system in modern crocodilians contradicts previous
interpretations of function of ornament. Trends in skull proportions
and jaw musculature as they relate to skull mechanics correlate with
changes in ornament types and relief. These trends imply a positive
association of ornament with a bone-strengthening function.

SPECIMENS EXAMINED

The following is an alphabetical list of most of the specimens examined in
this study ( see Appendix I for other specimens studied through the literature ) .
The specimens show the dorsal view of the skull unless otherwise noted.
"AMNH": American Museum of Natural History; "KU": University of Kansas
Museum of Natural History.

Achdoma cumminsi. — AMNH 4205. Archegosaurus. — AMNH 5704. Broiliel-
lus texensis. — AMNH 1824. Capitosaurus nasatus, — AMNH 5744. Cheno-
prosopus melleri. — AMNH 1831. Colosteus scutellatus. — AMNH 6916. Edops
craigi. — AMNH 7614. Erpetosaurus radiatus. — AMNH 6924, 6927. Eryops.—
AMNH 4175, 4183, 4901 (palatal view), KU 695. Eupelor browni.— AMNH
1832. Genothorax.— AMNH 3868. Ichthyostega.— AMNH 1058. Macrerpeton
huxk'iji.— AMNH 6944, 6834. Parioxys ferricolus.— AMNH 4310. Parotosaurus
peabodyi. — AMNH 2001. Saurerpeton tabulatus. — AMNH 6837. Stegops
divaricata.— AMNH 6952. Trimerorhachis.— AMNH 4591, 4557. Zatrachys.—
AMNH 7501.

18 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

LITERATURE CITED

Alexander, R. McN. 1968. Animal Mechanics. Univ. Washington Press,

Seattle. 346p.
Baird, D. 1964. The aistopod amphibians surveyed. Mus. Comp. Zool.,

206:1-17.
Bexxlnghoff, A. 1925. Spatlinien am Knochen, eine Methode zur Ermittlnng

der Architektur platter Knochen. Verhandl. anat. Ges., 34:189-206.
Bystrow, A. P. 1944. Kotlassia prima Amalitzky. Geol. Soc. Amer. Bull.,

55:379-416.
Bystrow, A. P. 1947. Hydrophilous and zenophilous labyrinthodonts. Acta

Zool., 28:137-164.
Carroll, R. L. 1969. Problems of the origin of reptiles. Phil. Trans. Roy.

Soc. (London), B., 257:267-308.
Colbert, E. H. 1955. Scales in the Permian amphibian Trimcrorhachis.

Amer. Mus. Nov., 17:1-17.
Credxer, H. 1883. Die Stegocephalen aus clem Rothliegenden des Plauen'schen

Grundes bei Dresden. IV. Theil. Zeitschr. deutsch. geol. Ges., 35:275-

300.
Currey, J- D. 1962. Stress concentrations in bone. Quart. Jour. Microsc. Sci.,

103:111-133.
Fox, R. C. 1964. The adductor muscles of the jaw in some primitive reptiles.

Univ. Kansas Publ. Mus. Nat. Hist., 12 ( 15): 657-680.
Guyer, M. F. 1936. Animal Micrology. Univ. Chicago Press, Chicago. 331p.
Olson, E. C. 1951. Diplocaulus, a study in growth and variation. Fieldiana:

Geology, (11) 2:59-154.
Olsox, E. C. 1961. Taw mechanisms: rhipidistians, amphibians, reptiles.

Amer. Zool., 1:205-215.
Romer, A. S. 1947. Review of the Labyrinthodontia. Bull. Mus. Comp. Zool.,

99:1-368.
Romer, A. S. 1972. Skin breathing-primary or secondary? Respiration Physiol.,

14:183-192.
Seipel, C. M. 1946. Trajectories of the jaw. Acta Odont. Scand., 8:81-191.
Tappax, N. C. 1953. A functional analysis of the facial skeleton with the

split-line technique. Amer. Jour. Phys. Anthro., 12:503-532.
Trueb, L. 1970. The evolutionary relationships of casque-headed treefrogs

with co-ossified skulls (family Hvlidae). Univ. Kansas Publ. Mus. Nat.

Hist., 18:547-716.
Trueb, L. 1973. Bones, frogs, and evolution. In James L. Vial, ed., Evolu-
tionary Biology of the Anurans (ch. 2). Univ. Missouri Press, Columbia:

65-132.
Watsox, D. M. S. 1951. Paleontology and Modern Biology. Yale Univ. Press,

New Haven. 216p.

APPENDIX I

The following references were used for data on skull measure-
ments and photographs. This list does not cover all genera men-
tioned in this study, but is supplemental to specimens studied
directly. There are a few specimens, however, which were studied
completely through the literature.

Acheloma

Olsox, E. C. 1941. The family Trematopsidae. Jour. Geol, 49:149-176.
Archeria (Cricotusj

Case, E. C. 1911. Revision of the Amphibia and Pisces of the Permian of
North America. Publ. Carnegie Inst. Washington, 146:1-176.

ORNAMENT IN LABYRINTHODONT AMPHIBIANS 19

Broiliellus

Whxiston, S. W. 1914. Broiliellus, a new genus of amphibian from the
Permian of Texas. Jour. Geo]., 22:49-56.
Cacops

Williston, S. W. 1910. Cacops; new genera of Permian vertebrates. Bull.
Geol. Soc. Amer., 21:249-284.
Chenoprosopus

Langston, W., Jr. 1953. Permian amphibians from New Mexico. Univ.
Calif. Publ. Geol. Sci., (29) 7:349-414.
Colosteus, Macrerpeton

Romer, A. S. 1930. The Pennsylvania!! tetrapods of Linton, Ohio. Bull.

Amer. Mus. Nat. Hist., 59:119-126.
Cope, E. D. 1875. Synopsis of the extinct Batrachia from the Coal Measures.
Geol. Survey Ohio, Paleont., 11:349-411.
Dissoroph us

Willistox, S. W. 1910. Dissorophus Cope. Jour. Geo!., 18:526-536.
Erpetosaurus, Saurerpeton, Stegops

Steex, M. 1931. The British Museum collection of Amphibia from the
Middle Coal Measures of Linton, Ohio. Proc. Zool. Soc. London,
(B), 1930 (1931): 849-891.
Eugyrinus

Watson, D. M. S. 1940. The origin of frogs. Trans. Roy. Soc. Edinburgh,
60:195-231.
Lyrocephalus

Save-Sodebergh, G. 1936. On the morphology of Triassic stegocephalians
from Spitzbergen, and the interpretation of the endocranium in the
Labyrinthodontia. K. Svenska Vetenskapsakad. Handl. (3), (16)
1:1-181.
Melosaurus

Efremov, J. A. 1937. Notes on the Permian Tetrapoda and the localities
of their remains. Trav. Isnt. Pal. Acad. Sci. URSS, (8) 1:1-44.
Parotosaurus

Piveteau, J. 1955. Traite de Paleontologie. Masson et Cie, Paris, 5: 1-1113.
Seymouria

Hottox, N. 1968. The Evidence of Evolution. Amer. Heritage Publ. Co.,
160p.
Trematosaurus

Watson, D. M. S. 1919. The structure, evolution, origin of Amphibia —
the orders Rhaehitomi and Stereospondyli. Philos. Trans. Roy. Soc.
London (B ), 209:38-41.
Piveteau, J. 1955. Traite de Paleontologie. Masson et Cie, Paris, 5: 1-1113.

Date Due

APft-3i)1983-

Oti

cmc

Bookbinding Co., Inc.

300 Summer Street

Boston. Mass. 02210

2044 093 361 665