ArticlePDF Available

Abstract

The serpent's forked tongue has intrigued humankind for millennia, but its function has remained obscure. Theory, anatomy, neural circuitry, function, and behavior now support a hypothesis of the forked tongue as a chemosensory edge detector used to follow pheromone trails of prey and conspecifics. The ability to sample simultaneously two points along a chemical gradient provides the basis for instantaneous assessment of trail location. Forked tongues have evolved at least twice, possibly four times, among squamate reptiles, and at higher taxonomic levels, forked tongues are always associated with a wide searching mode of foraging. The evolutionary success of advanced snakes might be due, in part, to perfection of this mechanism and its role in reproduction.
DOI: 10.1126/science.263.5153.1573
, 1573 (1994); 263Science
et al.Kurt Schwenk,
Why Snakes Have Forked Tongues
www.sciencemag.org (this information is current as of January 17, 2007 ):
The following resources related to this article are available online at
http://www.sciencemag.org
version of this article at:
including high-resolution figures, can be found in the onlineUpdated information and services,
http://www.sciencemag.org#otherarticles
, 2 of which can be accessed for free: cites 9 articlesThis article
http://www.sciencemag.org#otherarticles
4 articles hosted by HighWire Press; see: cited byThis article has been
http://www.sciencemag.org/help/about/permissions.dtl
in whole or in part can be found at: this article
permission to reproduce of this article or about obtaining reprintsInformation about obtaining
registered trademark of AAAS.
c 1994 by the American Association for the Advancement of Science; all rights reserved. The title SCIENCE is a
CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005.
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
on January 17, 2007 www.sciencemag.orgDownloaded from
ARTICLE
Why
Snakes
Have
Forked
Tongues
Kurt
Schwenk
The
serpent's
forked
tongue
has
intrigued
humankind
for
millennia,
but
its
function
has
remained
obscure.
Theory,
anatomy,
neural
circuitry,
function,
and
behavior
now
support
a
hypothesis
of
the
forked
tongue
as
a
chemosensory
edge
detector
used
to
follow
pheromone
trails
of
prey
and
conspecifics.
The
ability
to
sample
simultaneously
two
points
along
a
chemical
gradient
provides
the
basis
for
instantaneous
assessment
of
trail
location.
Forked
tongues
have
evolved
at
least
twice,
possibly
four
times,
among
squamate
reptiles,
and
at
higher
taxonomic
levels,
forked
tongues
are
always
associated
with
a
wide
search-
ing
mode
of
foraging.
The
evolutionary
success
of
advanced
snakes
might
be
due,
in
part,
to
perfection
of
this
mechanism
and
its
role
in
reproduction.
Deeply
embedded
in
the
popular
psyche,
serpents'
tongues
have
long
been
a
part
of
the
world's
religious
iconography
(I)
and
in
many
cultures
symbolize
malevolence
and
deceit.
Yet,
despite
this
prominence,
the
function
of
forked
tongues
has
eluded
ratio-
nal
scrutiny
for
more
than
two
millennia.
A
full
understanding
of
this
function
and
its
significance
to
the
evolution
of snakes
and
lizards
is
only
now
emerging,
yielding
to
a
multilevel
approach
highlighting
the
need
for
analysis
of the
whole
organism.
Snake
tongues,
it
turns
out,
tell
us
a
great
deal
about
snakes
and
about
evolution
as
well.
Recorded
inquiry
into
the
functional
significance
of
the
forked
tongue
begins
with
Aristotle
who,
reasoning
from
the
basis
of
his
own
tongue,
thought
that
it
would
provide
snakes
"a
twofold
pleasure
from
savours,
their
gustatory
sensation
be-
ing
as
it
were
doubled"
(2).
Some
19
centuries
later,
Hodierna
was
not
so
chari-
table
(3).
He
thought
that
snakes
used
their
forked
tongue
"for
picking
the
Dirt
out
of
their
Noses,
which
would
be
apt
else
to
stuff
them,
since
they
are
always
grovelling
on
the
Ground,
or
in
Caverns
of
the
Earth."
A
third
theory
has
snakes
catching
flies
"with
wonderful
nimbleness"
between
the
tines
of
their
forked
tongues
(4).
As
it
happens,
Aristotle
was
close
to
the
truth.
We
now
know
that
snake
tongues
are,
in
fact,
involved
in
chemoreception,
but
probably
not
gustation
as
Aristotle
sur-
mised.
Rather,
the
tongue
is
a
delivery
mechanism
for
paired
chemosensors
in
the
snout,
called
the
vomeronasal
(Jacobson's)
organs
(VNO)
(5).
These
organs
commu-
nicate
with
the
oral
cavity
through
two
tiny
openings
in
the
palate,
the
vomeronasal
fenestrae.
The
tongue
samples
environmen-
tal
chemicals
by
means
of
tongue
flicking,
a
behavior
in
which
the
tongue
is
rapidly
protruded,
sometimes
oscillated,
and
then
The
author
is
in
the
Department
of
Ecology
and
Evolutionary
Biology,
U-43,
University
of
Connecticut,
Storrs,
CT
06269-3043,
USA.
retracted
into
the
mouth,
usually
after
the
tongue
tip
contacts
the
ground
or
some
object
(6).
Odor
molecules
adherent
to
the
tongue
are
delivered
to
the
vomeronasal
fenestrae
where
they
make
their
way
to
the
sensory
epithelia
of the
VNO
(vomerolfac-
tion)
(7,
8).
Tongue
flicking,
then,
in-
cludes
a
sampling phase
(protrusion,
oscil-
lation,
and
chemical
pickup)
and
a
delivery
phase
(deposition
of the
sampled
chemical
within
the
oral
cavity
for
delivery
to
the
VNO).
For
many
species
of
squamate
rep-
tiles,
particularly
snakes,
vomerolfaction
may
be
the
dominant
sensory
mode
under-
lying
many
complex
behaviors
(2,
5).
Chemical
Delivery
and
Tongue
Form:
A
Red
Herring
Despite
Aristotle's
early
association
of
the
snake's
tongue
with
chemoreception,
this
concept
was
lost
in
a
myriad
of
fanciful
theories
that
were
proposed
over
the
cen-
turies
(4),
two
of
which
are
noted
above.
By
the
20th
century,
most
scientists
be-
lieved
that
the
tongue
was
a
tactile
organ.
Work
in
the
1920s
and
'30s
then
estab-
lished
chemical
sampling
and
delivery
to
the
VNO
as
the
function
of
tongue
flicking
(9-1i).
These
and
later
workers
were
se-
duced
by
the
natural
association
between
the
paired
VNO,
their
fenestrae
in
the
palate,
and
the
forked
form
of the
tongue
tip.
It
was
first
suggested
by
Broman
in
1920
that
the
attenuate
tips
of
the
forked
tongue
in
snakes
were
inserted
directly
into
the
vomeronasal
fenestrae
to
deliver
odor
mol-
ecules
to
the
VNO
(9).
In
so
doing,
Bro-
man
indelibly
linked
the
forked
tongue
morphology
with
the
delivery
phase
of
tongue
flicking,
a
link
that
has
remained
until
nearly
the
present
time.
Unfortunately,
Broman's
hypothesis
of
forked
tongue
function
must
be
rejected
on
the
basis
of
experimental
and
comparative
evidence.
First,
cineradiographic
(x-ray
movie)
studies
of
tongue
flicks
in
snakes
and
varanid
lizards
(with
a
similar
forked
tongue
morphology)
have
failed
to
reveal
lingual
movement
within
the
oral
cavity
consistent
with
such
a
hypothesis
(12).
In
addition,
the
tongue
was
seen
to
be
fully
retracted
into
its
sheath
before
mouth
clo-
sure
(13),
suggesting
that
sampled
mole-
cules
are
deposited
by
the
tongue
onto
paired
pads
on
the
floor
of
the
mouth
(anterior
processes
of
the
sublingual
plicae)
and
that
these,
not
the
tongue,
are
ap-
pressed
to
the
vomeronasal
fenestrae
to
effect
delivery.
Second,
and
ultimately
damning
to
the
Broman
hypothesis,
is
the
fact
that
vomeronasal
function
is
highly
developed
in
nearly
all
squamate
reptiles
(lizards,
snakes,
and
amphisbaenians),
and
these
share
a
similar,
derived
vomeronasal
form,
but
only
a
few
of
them
have
forked,
attenuate
tongue
tips
(Fig.
1).
In
most
lineages,
the
tongue
is
only
slightly
notched
and
is
rather
blunt;
the
tips
could
not
possibly
be
inserted
into
the
vomeronasal
fenestrae.
Despite
this
supposed
mechanical
limitation,
such
species
have
been
shown
to
deliver
chemicals
effectively
to
the
VNO
(8).
These
findings
refute
Broman's
hypoth-
esis
and
call
into
question
the
role
of
the
forked
tongue
tip
in
the
delivery
phase
of
tongue
flicking.
Despite
these
observations,
the
idea
that
Fig.
1.
Variation
in
the
form
of
the
tongue
tip
in
squamate
reptiles
in-
cluding
the
plesiomorphic
notched
morphology
and
several
forked
forms.
Shown
are
representatives
of
each
of
the
four
principal
lin-
.
:..
eages
of
Squamata.
The
drawings
/'
'
"-
are
not
to
scale,
but
the
approxi-
mate
resting
tongue
length
(RTL)
for
the
specimens
on
which
they
are
based
is
indicated
here.
From
left
to
right:
Sceloporus
(Iguania)
(RTL
=
7
mm);
Coleonyx
(Gekkonidae)
(RTL
=
6.5
mm);
Cnemidophorus
(Teiidae)
(RTL
=
16
mm);
Lacerta
(Lacertidae)
(RTL
=
7
mm);
Bipes
(Amphisbae-
nia)
(RTL
=
3.5
mm);
Scincella
(Scincidae)
(RTL
=
5.5
mm);
Abronia
(Anguidae)
(RTL
=
14
mm);
and
Varanus
(Varanidae)
(RTL
=
50
mm).
Most
snake
tongues
would
appear
similar
to
that
of
Varanus.
SCIENCE
*
VOL.
263
*
18
MARCH
1994
1573
on January 17, 2007 www.sciencemag.orgDownloaded from
the
forked
tip
morphology
and
the
delivery
phase
of
tongue
flicking
are
linked
has
persisted
in
the
notion
that
there
is
a
correlation
between
the
degree
of
lingual
bifurcation
(forkedness),
sensitivity
of
the
VNO,
and
importance
of
vomerolfaction
in
the
life
of
the
animal
(7,
8,
14,
15).
This
erroneous
notion
has
acted
as
a
red
herring
in
chemosensory
studies
because
it
has
di-
verted
attention
away
from
the
actual
func-
tion
of
the
forked
tongue,
namely,
its
me-
chanical
role
outside
the
oral
cavity
in
sampling
the
chemical
environment.
Chemical
Sampling
and
Tongue
Form:
Evidence
for
Tropotaxis
The
turn-of-the-century
concept
of
animal
tropisms
was
replaced
by
Fraenkel
and
Gunn
(16)
with
several
categories
of
move-
ment
they
termed
kineses
and
taxes.
For
directed
movement
toward
a
stimulus
source,
they
distinguished
klinotaxis
and
tropotaxis.
Both
taxes
involve
repeated
(or
continuous)
sensory
assessment
of
the
envi-
ronment
and
the
behavioral
response
of
movement
toward
(or
away
from)
the
stim-
ulus
but
are
differentiated
by
their
temporal
nature:
Klinotaxis
involves
comparison
of
successive
stimulus
intensities
between
dis-
parate
points,
whereas
tropotaxis
involves
simultaneous
comparison
of
stimulus
inten-
sities
on
two
sides
of
the
body.
Klinotaxis
is
common
in
chemoreception,
but
tropotaxis
is
theoretically
possible
only
when
the
chemical
gradient
is
steep
enough
for
two
parts
of
the
animal's
body
to
be
stimulated
with
different
intensities
(16).
For
a
snake
or
lizard
to
use
chemosensory
tropotaxis,
it
must
be
able
to
sense
simul-
taneously
the
chemical
stimulus
at
two
points.
This
requirement
is
met
admirably
by
the
forked
tongue.
The
more
deeply
forked
the
tongue,
the
greater
the
potential
distance
between
simultaneous
sampling
points.
The
distance
between
sampling
points
(the
tip
of
each
tine)
is
a
function
of
absolute
size,
fork
depth,
tongue
width,
and
the
degree
to
which
the
tines
of
the
fork
are
spread.
A
simple
tongue
fork
score
can
be
calculated
(Table
1).
In
some
species
(snakes,
varanid
and
teiid
lizards,
and
amphisbaenians)
the
potential
distance
between
the
tips
is
considerable,
exceeding
the
width
of the
head
(Fig.
2)
(17).
This
mechanism
requires
that
chemical
stimuli
on
each
tine
be
delivered
to
the
ipsilateral
VNO,
which
is
the
actual
sensory
organ.
Although
the
mechanism
of
chemi-
cal
delivery
remains
controversial
(18),
evi-
dence
indicates
that
same-side
delivery
would
be
possible
either
by
the
tongue
itself
or
by
the
sublingual
plicae.
The
chemical
source
must
exhibit
a
gradient
sufficiently
steep
that
a
differential
can
be
sensed
within
the
distance
available
1574
between
the
separated
tines
of
the
forked
tongue
tip;
in
other
words,
the
chemical
source
must
be
scaled
to
the
receiver.
This
requirement
is
met,
for
example,
by
trails
of
pheromones
left
by
passing
conspecifics
and
prey.
Such
trails,
a
known
target
of
tongue
flicking
by
squamates
(17,
19,
20),
are
of
biological
significance,
are
narrow,
and
not
particularly
volatile,
and
thus
provide
a
steep
chemical
gradient
of
the
appropriate
physical
dimension.
Does
the
forked
tongue,
in
fact,
func-
tion
in
tropotaxis?
Auffenberg
suggested
this
possibility
with
regard
to
prey
trailing
by
the
Komodo
monitor
(20),
pointing
out
that
each
tine
of
the
deeply
forked
tongue
could
deliver
a
different
stimulus
intensity
to
the
VNO.
However,
the
distinction
between
klino-
and
tropotaxis
is
a
key
factor
in
this
discussion,
because
it
is
clear
that
virtually
all
squamates
utilize
tongue
flicking
and
vomerolfaction
for
the
former,
but
not
necessarily
the
latter.
In
klinotaxis,
gross
movements
of the
head
and
body
would
allow
sequential
tongue
flicks
to
sam-
ple
disparate
points
along
the
ground.
The
stimulus
strength
of
the
temporally
and
spatially
sequential
points
could
then
be
compared
by
the
central
nervous
system.
Klinotaxis
is
characterized
by
wide-ranging,
exploratory
movements
that
sequentially
define
the
chemical
source
(16).
In
con-
trast,
tropotaxis
mediated
by
a
forked
tongue
would
assess
the
gradient
in
a
single
tongue
flick.
It
would
be
most
effective
once
a
circumscribed
chemical
source,
such
as
a
pheromone
trail,
was
encountered.
The
forked
tongue
could
then
function
as
an
edge
detector
to
delimit
the
chemical
zone
and
follow
it
with
minimal
deviation.
Although
no
single
piece
of
evidence
is,
itself,
definitive,
a
survey
of
the
data
leads
to
one
ineluctable
conclusion:
forked
tongues
function
during
tropotaxis
by
pro-
viding
a
mechanism
for
instantaneous
chemical
edge
detection
that
enhances
a
squamate's
ability
to
follow
pheromone
trails,
thus
accomplishing
the
biologically
critical
activities
of
seeking
both
prey
and
mates.
The
evidence
is
of
four
general
types.
1)
Behavior
constitutes
the
richest
source
of
support
for
the
tropotaxis
hypothesis.
The
be-
havior
most
clearly
implicated
in
the
forked
tongue
mechanism
is
following
of
a
phero-
mone
trail.
Therefore,
there
should
be
a
rough
correlation
between
depth
of
tongue
bifurcation
and
ability
to
follow
scent
trails.
Insufficient
data
exist
to
allow
a
quantita-
tive
comparison
of
trail-following
ability
among
taxa,
but
those
taxa
with
deeply
forked
tongues
(snakes,
amphisbaenians,
teiids,
varanids,
and
helodermatids)
are
highly
proficient
trail
followers
(2,
20-23).
Iguanians
and
gekkonids,
on
the
other
hand,
which
have
only
slightly
notched
SCIENCE
·
VOL.
263
*
18
MARCH
1994
tongues,
appear
not
to
follow
scent
trails.
Experimental
removal
of
the
forked
por-
tion
of
the
tongue
should
eliminate
the
ability
to
follow
scent
trails
but
not
delivery
of
stimulus
particles
to
the
VNO.
Remark-
ably,
these
experiments
were
performed
on
snakes
by
Kahmann
in
the
1930s
(11)
and
produced
the
predicted
results:
The
only
deficit
found
in
the
experimental
animals
was
reduction
or
loss
of
the
ability
to
follow
scent
trails.
Kahmann,
however,
interpret-
ed
his
results
in
light
of the
theory
that
the
tongue
tips
deliver
chemicals
directly
to
the
VNO.
Waters
(24)
confirmed
these
results
by
blocking
the
vomeronasal
fenestrae
of
garter
snakes
on
one
side.
Treated
snakes
were
unable
to
trail,
but
instead
turned
a
circle
toward
the
unblocked
side.
Ford
(17)
observed
male
garter
snakes
following
pheromone
trails
left
by
females
Fig.
2.
Tongue
flicks
in
a
teiid
(Tupinambis
nigropunctatus)
(A)
and
a
varanid
(Varanus
salvator)
(B)
showing
active
separation
of
tongue
fork
tines
at
the
moment
of
substrate
contact
(sampling).
Photos
are
enlarged
from
individual
16-mm
cine
frames
(64
frames
per
second)
extracted
from
sequences
of
unre-
strained
animals
tongue
flicking
the
floor
of
the
filming
chamber.
Sequences
show
that
when
the
tongue
is
initially
protruded,
the
tines
are
held
together.
They
are
progressively
spread
during
the
downward
phase
of
the
tongue
flick
and
reach
maximal
spread
during
contact
with
the
substrate.
During
retraction
of
the
tongue
into
the
mouth,
the
tines
come
closer
together
but
remain
separate
until
they
disappear
from
sight.
Photos
were
made
from
films
taken
by
G.
S.
Throckmorton,
with
permission.
on January 17, 2007 www.sciencemag.orgDownloaded from
,,,,,,,,,,,,,,,,,,,,,,,,,,~...
,~,:,,:.........
.~,
and
suggested
that
the
forked
tongue
func-
tions
as
part
of
a
tropotactic
mechanism.
He
found
that
"When
the
edge
of
the
trail
was
exceeded
by
one
tongue
tip
during
a
tongue-flick,
the
snake
reversed
direction
and
swung
his
head
back
into
the
phero-
mone
field
during
the
period
before
the
next
tongue-flick."
The
tongue
tips
were
widely
spread
during
such
trail-following
behavior,
as
much
as
twice the
width
of
the
snake's
head,
and
trail
following
was
accu-
rate
and
directed.
Occasional
loss
of
the
trial
only
occurred
when
the
male's
head
left
the
female's
pheromone
field
between
tongue
flicks.
Loss
of
the
trail
caused
the
male
to
stop
for
one
or
several
tongue
flicks
and
then
swing
the
head
from
side
to
side
between
flicks
until
the
trail
was
relocated.
The
latter
behavior
is
consistent
with
kli-
notaxis,
typical
of
all
squamates,
including
those
without
forked
tongues.
2)
A
forked
tongue
tip
morphology
would
be
most
significant
during
the
sampling
phase
of
tongue
flicking
and
not
the
delivery
phase.
Therefore,
a
forked
tongue
tip
should
not
be
necessary
for
chemical
delivery
to
the
VNO.
This
is
supported
by
comparative
analysis
(above),
as
well
as
experimental
evidence
showing
that
removal
of
the
forked
portion
of the
tongue
in
snakes
and
lizards
does
not
prevent
stimulus
delivery
to
the
VNO
after
tongue
flicking
(7,
11,
25).
3)
The
tips
of
the
tongue
should
be
spread
laterally
during
a
tongue
flick
when
odor
mol-
ecules
are
retrieved,
because
the
greater
the
distance
between
sampling
points,
the
greater
the
likelihood
of
sampling
a
chemical
gradient
in
a
single
flick.
Observations
of
lizards
with
forked
tongues
(teiids
and
varanids)
(Fig.
2)
and
snakes
(17)
confirm
that
the
tines
are
spread
rapidly
and
widely
just
before
con-
tact
with
the
substrate.
Histology
of
the
tongue
tip
in
snakes
suggests
that
muscle
fiber
architecture
optimizes
bending
of
the
tines
rather
than
elongation,
in
contrast
to
the
remainder
of
the
tongue
(26).
4)
Central
nervous
system
projections
of
the
vomeronasal
receptor
cells
must
provide
a
neu-
ral
substrate
for
comparing
signal
strength
from
each
side
of
the
tongue,
that
is,
a
mechanism
that
enables
the
animal
to
locate
the
signal
in
space.
Owl
hearing
offers
an
analogy,
as
owls
use
their
paired
ears
to
localize
a
sound
source.
They
compare
sound
intensity
(and
the
time
of
arrival
of
sound)
at
each
ear
(27).
Similarly,
squamates
compare
differ-
ential
chemical
signal
strength
delivered
from each
tine
to
the
paired
VNO.
In
owls,
each
auditory
pathway
projects
to
a
nucleus
that
communicates
through
commissural
connections
to
the
contralateral
nucleus.
Direct
input
to
the
nucleus
is
excitatory,
whereas
commissural
input
is
inhibitory,
thereby
providing
the
basis
for
neural
com-
putation
of
interaural
differences
(28).
In
squamates
the
VNO
projects
to
a
nucleus
(nucleus
sphericus)
(29),
and
in
snakes the
nucleus
sphericus
communicates
with
the
contralateral
nucleus
through
the
anterior
commissure
(30),
circuitry
that
is
similar
to
the
owl's
(that
is,
central
projection
to
a
nucleus
with
commissural
communication
between
contralateral
nuclei).
Although
not
conclusive,
such
circuitry
is
consistent
Table
1.
Tongue
tip
form
and
foraging
ecology
of
lizards.
Shown
are
families
for
which
both
tongue
tip
data
and
foraging
mode
data
were
available.
Snakes
were
excluded
from
this
analysis
because
data
were
not
available
to
calculate
tongue
fork
scores
(TFS)
by
family.
The
presence
of
a
forked
tongue
is
positively
correlated
with
wide
foraging
(Wilcoxon
rank
sum
test,
P
=
0.016).
Lizard
families
are
ranked
in
order
of
increasing
degree
of
tongue
tip
bifurcation.
Numbers
in
parentheses
after
each
taxon
indicate
[number
of
species
used
to
calculate
TFS,
number
of
species
used
to
calculate
foraging
mode
score
(FMS)].
TFS
is
the
length
of
the
lingual
bifurcation
(tine
length)
divided
by
the
width
of
the
tongue
at
the
base
of
the
bifurcation.
Each
score
is
given
as
a
family
mean
followed
by
the
standard
error.
Measurements
are
from
photographs
taken
by
the
author.
I
calculated
FMS
from
characterizations
of
foraging
mode
by
species
(43,
44).
Sit-and-wait
(SAW)
species
were
scored
0
and
wide
foragers
(WF)
1;
FMS
is
the
mean
value
for
the
family.
For
the
dichotomous
classifications,
TFS
>
1
is
considered
to
be
"forked"
and
FMS
>
0.5
is
considered
to
be
"wide
foraging."
All
snakes
have
forked
tongues
[TFSs
range
from
1.5
(Scolecophidia)
to
values
comparable
with
varanids
(6.0+)
(Caenophidia)]
and
virtually
all
are
wide
foragers
(43).
The
traditional
families
Iguanidae
and
Agamidae
(Iguania)
may
be
paraphyletic
assemblages
(45)
but,
in
any
case,
are
relatively
uniform
in
terms
of
tongue
tip
form
and
foraging
mode.
Dichotomous
Taxon
Mean
TFS
FMS
TFS
FMS
Xantusiidae
(1,
3)
0.10
0
-
SAW
Agamidae*
(3,
6)
0.13
(0.06)
0
-
SAW
Iguanidae*
(11,
68)
0.17
(0.02)
0
-
SAW
Scincidae
(6,
3)
0.28
(0.02)
0.67
-
WF
Anguidae
(3,
5)
0.53
(0.02)
0
-
SAW
Helodermatidae
(1,
2)
1.08
1.00
+
WF
Lacertidae
(3,
5)
1.70
(0.32)
0.80
+
WF
Teiidae
(5,
19)
3.26
(0.25)
1.00
+
WF
Varanidae
(2,
9)
6.43
(0.58)
1.00
+
WF
*Potentially
paraphyletic
taxa
(45).
with
the
tropotaxis
hypothesis.
Physiologi-
cal
evidence
of
direct
nucleus
sphericus
excitation
and
commissural
inhibition
would
lend
additional
support.
The
proposed
mechanism
of
forked
tongue
function
does
not
suggest
that
liz-
ards
and
snakes
use
it
to
sense
the
direction
taken
by
the
prey
or
conspecific.
This
would
require
detection
of
some
sort
of
polarized
trail.
Trail
direction
probably
is
determined
by
tongue
flicking
of
objects
in
the
environment
against
which
the
target
has
pushed
during
locomotion
(31).
During
snake
locomotion,
chemical
cues
would
be
deposited
only
on
the
side
facing
the
direc-
tion
of
travel,
and
therefore
direction
can
be
determined
by
tongue
flicking
both
sides
of
the
object.
Some
venomous
snakes
can
differentiate
between
trails
made
by
en-
venomated
and
nonenvenomated
prey;
hence
they
can
follow
a
trail
in
the
direc-
tion
of
the
prey
item
once
it
has
been
struck
(32).
It
remains
unclear
how
directionality
can
be
determined,
if
at
all,
for
trails
laid
by
limbed
prey
or
conspecifics
that
do
not
push
off
against
environmental
objects.
Tongue
Form
and
Ecology
The
tropotaxis
function
of
the
forked
tongue
is
indicated
by
a
wide
range
of
organismal
data
and
suggests
certain
predic-
tions.
For
example,
if
a
forked
tongue
en-
hances
the
ability
to
follow
prey
trails,
there
may
be
a
correlation
between
the
presence
of a
forked
tongue
and
foraging
mode
(33).
Wide-ranging
foragers
should
profit
from
a
trail-following
mechanism
be-
cause
they
seek
out
prey
by
moving
widely
through
the
environment.
The
ability
to
follow
a
trail
efficiently
to
its
source
would
be
an
advantage
to
an
animal
that
expends
a
great
deal
of
its
energy
in
search.
Con-
versely,
sit-and-wait
foragers
have
no
need
to
follow
prey
trails,
because
they
wait
for
prey
to
come
to
them.
Klinotaxis
would
be
sufficient