The Journal of Physiology - 1952 - Hodgkin - A Quantitative Description of Membrane Current and Its Application To

14697793, 1952, 4, Downloaded from https://physoc.onlinelibrary.wiley.com/doi/10.1113/jphysiol.1952.sp004764 by Cochrane Peru, Wiley Online Library on [27/11/2022].
See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
500
J. Physiol. (I952) I I7, 500-544
A QUANTITATIVE DESCRIPTION OF MEMBRANE

CURRENT AND ITS APPLICATION TO CONDUCTION
AND EXCITATION IN NERVE
BY A. L. HODGKIN AND A. F. HUXLEY
From the Physiological Laboratory, University of Cambridge
(Received 10 March 1952)
This article concludes a series of papers concerned with the flow of electric
current through the surface membrane of a giant nerve fibre (Hodgkin,
Huxley & Katz, 1952; Hodgkin & Huxley, 1952 a-c). Its general object is to
discu the results of the preceding papers (Part I), to put them into
mathematical form (Part II) and to show that they will account for con-
duction and excitation in quantitative terms (Part III).
PART I. DISCUSSION OF EXPERIMENTAL RESULTS
The results described in the preceding papers suggest that the electrical
behaviour of the membrane may be represented by the network shown in
Fig. 1. Current can be carried through the membrane either by charging the
membrane capacity or by movement of ion-s through the resistances in parallel
with the capacity. The ionic current is divided into components carried by
sodium and potassium ions (INa and IK), and a small 'leakage current' (I,)
made up by chloride and other ions. Each component of the ionic current is
determined by a driving force which may conveniently be measured as an
electrical potential difference and a permeability coefficient which has the
dimensions of a conductance. Thus the sodium current (INa) is equal to the
sodium conductance (9Na) multiplied by the difference between the membrane
potential (E) and the equilibrium potential for the sodium ion (ENa). Similar
equations apply to 'K and I, and are collected on p. 505.
Our experiments suggest that gNa and 9E are functions of time and
membrane potential, but that ENa, EK, El, CM and g, may be taken as
constant. The influence of membrane potential on permeability can be sum-
marized by stating: first, that depolarization causes a transient increase in
sodium conductance and a slower but maintained increase in potassium con-
ductance; secondly, that these changes are graded and that they can be
reversed by repolarizing the membrane. In order to decide whether these
effects are sufficient to account for complicated phenomena such as the action
potential and refractory period, it is necessary to obtain expressions relating
14697793, 1952, 4, Downloaded from https://physoc.onlinelibrary.wiley.com/doi/10.1113/jphysiol.1952.sp004764 by Cochrane Peru, Wiley Online Library on [27/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MEMBRANE CURRENT IN NERVE 501
the sodium and potassium conductances to time and membrane potential.
Before attempting this we shall consider briefly what types of physical system
are likely to be consistent with the observed changes in permeability.
Outside
~~~~~
CM
T TIN TK ItE
~+ +
Inside )
Fig. 1. Electrical circuit representing membrane. RB =l/gNa; RK= l/9K; RI= 1/#1. RNw and
RK vary with time and membrane potential; the other components are constant.
The nature of the permewablity change8
At present the thickness and composition of the excitable membrane are
unknown. Our experiments are therefore unlikely to give any certain informa-
tion about the nature of the molecular events underlying changes in perme-
ability. The object of this section is to show that certain types of theory are
excluded by our experiments and that others are consistent with them.
The first point which emerges is that the changes in permeability appear to
depend on membrane potential and not on membrane current. At a fixed
depolarization the sodium current follows a time course whose form is inde-
pendent of the current through the membrane. If the sodium concentration
is such that ENaB<E, the sodium current is inward; if it is reduced until
ENa > E the current changes in sign but still appears to follow the same time
course. Further support for the view that membrane potential is the variable
controlling permeability is provided by the observation that restoration of the
normal membrane potential causes the sodium or potassium conductance to
decline to a low value at any stage of the response.
The dependence of 9Na and g9 on membrane potential suggests that the
permeability changes arise from the effect of the electric field on the distribu-
tion or orientation of molecules with a charge or dipole moment. By this we
do not mean to exclude chemical reactions, for the rate at which these occur
might depend on the position of a charged substrate or catalyst. All that is
intended is that small changes in membrane potential Would be most unlikely
502 A. L. HODGKIN AND A. F. HUXLEY
to cause large alterations in the state of a membrane which was composed
entirely of electrically neutral molecules.
The next question to consider is how changes in the distribution of a charged
particle might affect the ease with which sodium ions cross the membrane.
Here we can do little more than reject a suggestion which formed the original
basis of our experiments (Hodgkin, Huxley & Katz, 1949). According to this
view, sodium ions do not cross the membrane in ionic form but in combination
with a lipoid soluble carrier which bears a large negative charge and which can
combine with one sodium ion but no more. Since both combined and un-
combined carrier molecules bear a negative charge they are attracted to the
outside of the membrane in the resting state. Depolarization allows the carrier
molecules to move, so that sodium current increases as the membrane potential
is reduced. The steady state relation between sodium current and voltage
could be calculated for this system and was found to agree reasonably with the
observed curve at 0-2 msec after the onset of a sudden depolarization. This
was encouraging, but the analogy breaks down if it is pursued further. In the
model the first effect of depolarization is a movement of negatively charged
molecules from the outside to the inside of the membrane. This gives an initial
outward current, and an inward current does not occur until combined carriers
lose sodium to the internal solution and return to the outside of the membrane.
In our original treatment the initial outward current was reduced to vanishingly
small proportions by assuming a low density of carriers and a high rate of
movement and combination. Since we now know that the sodium current
takes an appreciable time to reach its maximum, it is necessary to suppose
that there are more carriers and that they react or move more slowly. This
means that any inward current should be preceded by a large outward current.
Our experiments show no sign of a component large enough to be consistent
with the model. This invalidates the detailed mechanism assumed for the
permeability change but it does not exclude the more general possibility that
sodium ions cross the membrane in combination with a lipoid soluble carrier.
A different form of hypothesis is to suppose that sodium movement depends
on the distribution of charged particles which do not act as carriers in the
usual sense, but which allow sodium to pass through the membrane when they
occupy particular sites in the membrane. On this view the rate of movement
of the activating particles determines the rate at which the sodium con-
ductance approaches its maximum but has little effect on the magnitude of
the conductance. It is therefore reasonable to find that temperature has
a large effect on the rate of rise of sodium conductance but a relatively small
effect on its maximum value. In terms of this hypothesis one might explain
the transient nature of the rise in sodium conductance by supposing that the
activating particles undergo a chemical change after moving from the position
which they occupy when the membrane potential is high. An alternative is to
attribute the decline of sodium conductance to the relatively slow movement
of another particle which blocks the how of sodium ions when it reaches a
certain position in the membrane.
Much of what has been said about the changes in sodium permeability
applies equally to the mechanism underlying the change in potassium perme-
ability. In this case one might suppose that there is a completely separate
system which differs from the sodium system in the following respects: (1) the
activating molecules have an affinity for potassium but not for sodium,
(2) they move more- slowly; (3) they are not blocked or inactivated. An
alternative hypothesis is that only one system is present but that its selectivity
changes soon after the membrane is depolarized. A situation of this kind would
arise if inactivation of the particles selective for sodium converted them into
particles selective for potassium. However, this hypothesis cannot be applied
in a simple form since the potassium conductance rises too slowly for a direct
conversion from a state of sodium permeability to one of potassium
permeability.
One of the most striking properties of the membrane is the extreme steepness
of the relation between ionic conductance and membrane potential. Thus gNa
may be increased e-fold by a reduction of only 4 mV, while the corresponding
figure for gE is 5-6 mV (Hodgkin & Huxley, 1952 a, figs. 9, 10). In order to
illustrate the possible meaning of this result we shall suppose that a charged
molecule which has some special affinity for sodium may rest either on the
inside or the outside of the membrane but is present in negligible concentra-
tions elsewhere. We shall also suppose that the sodium conductance is pro-
portional to the number of such molecules on the inside of the membrane but
is independent of the number on the outside. From Boltzmann's principle the
proportion Pi of the molecules on the inside of the membrane is related to the
proportion on the outside, PO, by
= exp[(w + zeE)/kT],
where E is the potential difference between the outside and the inside of the
membrane, w is the work required to move the molecule from the inside to the
outside of the membrane when E = 0, e is the absolute value of the electronic
charge, z is the valency of the molecule (i.e. the number of positive electronic
charges on it), k is Boltzmann's constant and T is the absolute temperature.
Since we have assumed that Pi + PO = 1 the expression for Pi is
Pi = 1/[1 + exp - kT )
For negative values of z and with E sufficiently large and positive this gives
Pi = constant x exp[zeE/kT].
5, .4 A. L. HODGKIN AND A. F. HUXLEY
In order to explain our results z must be about -6 since k( =R) is 25 mV
at room temperature and g9.oc exp - E/4 for B large. This suggests that the
particle whose distribution changes must bear six negative electronic charges,
or, if a similar theory is developed in terms of the orientation of a long
molecule with a dipole moment, it must have at least three negative charges
on one end and three positive charges on the other. A different but related
approach is to suppose that sodium movement depends on the presence of six
singly charged molecules at a particular site near the inside of the membrane.
The proportion of the time that each of the charged molecules spends at the
inside is determined by exp - E/25 so that the proportion of sites at which all
six are at the inside is exp - E/4.17. This suggestion may be given plausibility
but not mathematical simplicity by imagining that a nlumber of charges form
a bridge or chain which allows sodium ions to flow through the membrane
when it is depolarized. Details of the mechanism will probably not be settled
for some time, but it seems difficult to escape the conclusion that the changes
in ionic permeability depend on the movement of some component of the
membrane which behaves as though it had a large charge or dipole moment.
If such components exist it is necessary to suppose that their density is
relatively low and that a number of sodium ions cross the membrane at a single
active patch. Unless this were true one would expect the increase in sodium
permeability to be accompanied by an outward current comparable in
magnitude to the current carried by sodium ions. For movement of any
charged particle in the membrane should contribute to the total current and
the effect would be particularly marked with a molecule, or aggregate, bearing
a large charge. As was mentioned earlier, there is no evidence from our
experiments of any current associated with the change in sodium perme-
ability, apart from the contribution of the sodium ion itself. We cannot set
a definite upper limit to this hypothetical current, but it could hardly have
been more than a few per cent of the maximum sodium current without pro-
ducing a conspicuous effect at the sodium potential.
PART II. MATHEMATICAL DESCRIPTION OF MEMBRANE

CURRENT DURING A VOLTAGE CLAMP
Total membrane current
The first step in our analysis is to divide the total membrane current into
a capacity current and an ionic current. Thus
I = CM d+ i, (1)
MEMBRANE CURRENT IN NERVE 505,A
where
I is the total membrane current density (inward current positive);
Ii is the ionic current density (inward current positive);
V is the displacement of the membrane potential from its resting value
(depolarization negative);
CM is the membrane capacity per unit area (assumed constant);
t is time.
The justification for this equation is that it is the simplest which can be used
and that it gives values for the membrane capacity which are independent of
the magnitude or sign of V and are little affected by the time course of V (see,
for example, table 1 of Hodgkin et al. 1952). Evidence that the capacity
current and ionic current are in parallel (as suggested by eqn. (1)) is provided
by the similarity between ionic currents measured with dV= 0 and those
dV~~~~~~~~~d
calculated from-CM dt with I= 0 (Hodgkin et al. 1952).
The only major reservation which must be made about eqn. (1) is that it
takes no account of dielectric loss in the membrane. There is no simple way of
estimating the error introduced by this approximation, but it is not thought
to be large since the time course of the capacitative surge was reasonably close
to that calculated for a perfect condenser (Hodgkin et al. 1952).
The ionic current
A further subdivision of the membrane current can be made by splitting the
ionic current into components carried by sodium ions (INa), potassium ions
(IK) and other ions (I): i=INa + IK + I (2)
The individual ionic currents
In the third paper of this series (Hodgkin & Huxley, 1952 b), we showed that
the ionic permeability of the membrane could be satisfactorily expressed in
terms of ionic conductances (9Na,I and gl). The individual ionic currents are
obtained from these by the relations
'Na =gNa (E ENa),
IK=gK (E-EK),
II=lg= (E-.El),
where ENa and EK are the equilibrium potentials for the sodium and potassium
ions. El is the potential at which the 'leakage current' due to chloride and
other ions is zero. For practical application it is convenient to write these
equations in the form INa=gNa (V-VNa) (3)
IK 9EK (V -VK), 4
Itl91 (VV-Vl), (5)
where V=E-Er,
VNa = ENa -Er,
VK= EK-Er,
VI=Ei-Er,
and Er is the absolute value of the resting potential. V VgN, VK and VI can
then be measured directly as displacements from the resting potential.
The ionic conductances
The discussion in Part I shows that there is little hope of calculating the
time course of the sodium and potassium conductances from first principles.
Our object here is to find equations which describe the conductances with
reasonable accuracy and are sufficiently simple for theoretical calculation of
the action potential and refractory period. For the sake of illustration we
shall try to provide a physical basis for the equations, but must emphasize
that the interpretation given is unlikely to provide a correct picture of the
membrane.
8
,,E5 A
004
EE 3 -
0
I I I I I I .1 I I I I
0 1 2 3 4 5 0 1 2 3 4 5
msec msec
Fig. 2. A, rise of potassium conductance associated with depolarization of 25 mV; B, fall of
potassium conductance associated with repolarization to the resting potential. Circles:
experimental points replotted from Hodgkin & Huxley (1952b, Fig. 13). The last point of
A is the same as the first point in B. Axon 18, 210 C in choline sea water. The smooth curve
is drawn according to eqn. (11) with the following parameters:
Curve A Curve B
(V= -25 mV) (V=0)
9qKo 0 09 m.mho/cm2 7.0 m.mho/Cm2
9gOo 7-06 m.mho/cm2 009 m.mho/cm'
n075 msec 1 1 msec
At the outset there is the difficulty that both sodium and potassium con-
ductances increase with a delay when the axon is depolarized but fall with no
appreciable infiexion when it is repolarized. This is illustrated by the circles in
Fig. 2, which shows the change in potassium conductance associated with
a depolarization of 25 mV lasting 4-9 msec. If g9 is used as a variable the end
of the record can be fitted by a first-order equation but a third- or fourth-order
equation is needed to describe the beginning. A useful simplification is
achieved by supposing that g9 is proportional to the fourth power of a variable
which obeys a first-order equation. In this case the rise of potassium con-
ductance from zero to a finite value is described by (1 - exp (-t))4, while the
fall is given by exp (-4t). The rise in conductance therefore shows a marked
inflexion, while the fall is a simple exponential. A similar assumption using
a cube instead of a fourth power describes the initial rise of sodium con-
ductance, but a term representing inactivation must be included to cover the
behaviour at long times.
The potassium conductance
The formal assumptions used to describe the potassium conductance are:
- n4 (6)
dn
ddn =n (1- n)-Bnn, (7 )
where g is a constant with the dimensions of conductance/cm2, cx and /n
are rate constants which vary with voltage but not with time and have
dimensions of [time]-', n is a dimensionless variable which can vary between
0 and 1.
These equations may be given a physical basis if we assume that potassium
ions can only cross the membrane when four similar particles occupy a certain
region of the membrane. n represents the proportion of the particles in a certain
position (for example at the inside of the membrane) and 1- n represents
the proportion that are somewhere else (for example at the outside of the
membrane). an determines the rate of transfer from outside to inside, while
fln determines the transfer in the opposite direction. If the particle has a
negative charge an should increase and fin should decrease when the membrane
is depolarized.
Application of these equations will be discussed in terms of the family of
curves in Fig. 3. Here the circles are experimental observations of the rise of
potassium conductance associated with depolarization, while the smooth
curves are theoretical solutions of eqns. (6) and (7).
In the resting state, defined by V = 0, n has a resting value given by
anO
-n + finO'
no
If V is changed suddenly can and n instantly take up values appropriate to the

new voltage. The solution of (7) which satisfies the boundary condition that
n=no when t=O is
n=na,,-(nn-no) exp (-t/rn), (8)
where no, = an/(Xn + fAn) (9)
and rTn = l/(An + n). (10)
From eqn. (6) this may be transformed into a form suitable for comparison
with the experimental results, i.e.
9 ={(gK) - [(9K1) - (gKo)*] exp (-t/T)}', (11)
where gt- is the value which the conductance finaly attains and gEo is the
conductance at t=O. The smooth curves in Fig. 3 were calculated from
20 _ _9L. ~~~~~~~~~~~~~109
-
A
15 100
-B
10
'5
-V~
~ ~~ ~ ~ ~ ~ ~ ~ ~ 8
~~~o E
-C
0~~~~~~~~~~~~~~
_ ~ ~~~~~~~ c
10
E L
E~~~~~~~~~~~~~~
4,J
co
0 3
1 _1 F
0~
n 1 23 4 s 6 7 8 9 10 11msec6
0
Fig. 3. Rise of potassium conductance associated with different depolarizations. The circles are
experimental points obtained on axon 17, temperature 6-70 C, using observations in sea water
and choline sea water (see Hodgkin & Huxley, 1952a). The smooth curves were drawn from
eqn. (11) with 9KO==0 24 m.mho/cm2 and other parameters as shown in Table 1. The time
scale applies to all records. The ordinate scale is the same in the upper ten curves (A to J)
and is increased fourfold in the lower two curves (K and L). The number on each curve gives
the depolarization in mV.
eqn. (11) with a value of Zr chosen to give the best fit. It will be seen that
there is reasonable agreement between theoretical and experimental curves,
except that the latter show more initial delay. Better agreement might have
been obtained with a fifth or sixth power, but the improvement was not con-
sidered to be worth the additional complication.
The rate constants an and f,n. At large depolarizations gK. seems to approach
an asymptote about 20-50% greater than the conductance at -100 mV.
TABLE 1. Analysis of curves in Fig. 3
v 9KO n,x, Tnpnfn
(mV) (m.mho/cm') (msec) (msec-1) (msec'L)
Curve (1) (2) (3) (4) (5) (6)
- (-0c) (24.31) (1-000) -
A - 109 20*70 0*961 1.05 0.915 0 037
B - 100 20-00 0.953 1-10 0*866 0*043
C - 88 18-60 0*935 1*25 0*748 0-052
D - 76 17-00 0*915 1.50 0*610 0 057
E - 63 15-30 0-891 1-70 0*524 0 064
F - 51 13*27 0.859 2-05 0*419 0-069
G - 38 10*29 0*806 2;60 0.310 0*075
H - 32 8-62 0*772 3 20 0-241 0-071
I - 26 6*84 0-728 3-80 0*192 0-072
J - 19 5*00 0*674 4 50 0.150 0*072
K - 10 1*47 0*496 5-25 0 095 0*096
L - 6 0*98 0.448 5-25 0*085 0*105
(0) (0.24) (0-315) - -
Col. 1 shows depolarization in mV; col. 2, final potassium conductance; col. 3, n00 =(9K0/§F);
col. 4, time constant used to compute curve; col. 5, n,. =n0v/r; col. 6, fin =(1 - n0)/v,,. The figure
of 24-31 was chosen for ,K because it made the asymptotic value of n00 5% greater than the value
at -100 mV.
For the purpose of calculation we assume that n= 1 at the asymptote which

is taken as about 20 % greater than the value Of 9gK0 at V =-100 mV. These
assumptions are somewhat arbitrary, but should introduce little error since
we are not concerned with the behaviour of 9g at depolarizations greater than
about 110 mV. In the experiment illustrated by Fig. 3, gK. = 20 m.mho/cm2
at V = -100 mV. 9E was therefore chosen to be near 24 m.mho/cm2. This
value was used to calculate n0 at various voltages by means of eqn. (6). an and
fn could then be obtained from the following relations which are derived from
eqns. (9) and (10):
aBn = (l-o)1r'W
The results of analysing the curves in Fig. 3 by this method are shown in
Table 1.
An estimate of the resting values of an and fin could be obtained from the
decline in potassium conductance associated with repolarization. The pro-
cedure was essentially the same but the results were approximate because the
resting value of the potassium conductance was not known with any accuracy
when the membrane potential was high. Fig. 2 illustrates an experiment in
which the membrane potential was restored to its resting value after a de-
polarization of 25 mV. It will be seen that both the rise and fall of the
potassium conductance agree reasonably with theoretical curves calculated
from eqn. (11) after an appropriate choice of parameters. The rate constants
derived from these parameters were (in msec-1): o= 021, f,n- = 070 when
V=0 and ocn= 0 90, fin = 043 when V= -25 mV.
In order to find functions connecting Ocn and fiP with membrane potential we
collected all our measurements and plotted them against V, as in Fig. 4.
Differences in temperature were allowed for by adopting a temperature
coefficient of 3 (Hodgkin et al. 1952) and scaling to 60 C. The effect of replacing
sodium by choline on the resting potential was taken into account by dis-
placing the origin for values in choline sea water by + 4 mV. The continuous
curves, which are clearly a good fit to the experimental data, were calculated
from the following expressions:
0X=001 (V + 10)exp 0 - (12)

n
~ O/L 10 ~,(2
Pn=0 125 exp (V/80), (13)
where an and fin are given in reciprocal msec and V is the displacement of the
membrane potential from its resting value in mV.
These expressions should also give a satisfactory formula for the steady
potassium conductance (gK-) at any membrane potential (V), for this relation
is implicit in the measurement of xn and fin. This is illustrated by Fig. 5, in
which the abscissa is the membrane potential and the ordinate is (g9/9,K)1.
The smooth curve was calculated from eqn. (9) with an and fn substituted
from eqns. (12) and (13).
Fig. 4 shows that fin is small compared to an over most of the range; we
therefore do not attach much weight to the curve relating fin to V and have
used the simplest expression which gave a reasonable fit. The function for on
was chosen for two reasons. First, it is one of the simplest which fits the
experimental results and, secondly, it bears a close resemblance to the
equation derived by Goldman (1943) for the movements of a charged particle
in a constant field. Our equations can therefore be given a qualitative physical
basis if it is supposed that the variation of oc and fi with membrane potential
arises from the effect of the;; electric field on the movement of a negatively
charged particle which rests on the outside of the membrane when V is large
and positive, and on the inside when it is large and negative. The analogy
cannot be pressed since ac and f are not symmetrical about B =0, as they
should be if Goldman's theory held in a simple form. Better agreement might
MEMBRANE CURRENT IN NERVE 511.
et n /8, Axon
1.0 v v 15
o * 17
0*9 o~ * 18
o * 20
_ 08 A A 21
x 23
N4 0-7 ccn + 24
E " 26
c 06
0
0
2'V 04
a-v0 0,4
0-3
0-2
01
0
-110-100-90-80-70--60-50-40-30-20-10 0 10 20 30 40 50
V(mV)
Fig. 4. Abscissa: membrane potential minus resting potential in sea water. Ordinate: rate
constants determining rise (a") or fall (Bn) of potassium conductance at 6° The resting
0.
potential was assumed to be 4 mV higher in choline sea water than in ordinary sea wate:
Temperature differences were allowed for by assuming a Qlo of 3. All values for V < 0 were
obtained by the method illustrated by Fig. 3 and Table 1; those for V> 0 were obtained
from the decline of potassium conductance associated with an increase of membrane potential
or from repolarization to the resting potential in choline sea water (e.g. Fig. 2). Axons 17-21
at 6-11' C, the remainder at about 200 C. The smooth curves were drawn from eqns. (12)
and (13).
Axon
1
0t
-036
-
Fe A 21
01~
~~~ ~ ~~~~~~~~~~~0 2
0-4 -
0-3 -
0-5~ -
~ ~ ~ ~
0-1 _
-1 10 -100-90 -80 -70 -60 -50 -40 -30 -2o0-10 0 +1 0 +20 +30 +40 +50 +60
V (mV)
Fig. 5. Abscissa: membrane potential minus resting potential in sea water. Ordinate: experi-
mental measurements of n, calculated from the steady potassium conductance by the
relation n. = $gK./190), where 9K is the 'maximum' potassium conductance. The smooth
curve is drawn according to eqn. (9).
be obtained by postulating some asymmetry in the structure of the membrane,
but this assumption was regarded as too speculative for profitable considera-
tion.
The sodium conductance
There are at least two general methods of describing the transient changes
in sodium conductance. First, we might assume that the sodium conductance
is determined by a variable which obeys a second-order differential equation.
Secondly, we might suppose that it is determined by two variables, each of
which obeys a first-order equation. These two alternatives correspond roughly
to the two general types of mechanism mentioned in connexion with the
nature of inactivation (pp. 502-503). The second alternative was chosen since
it was simpler to apply to the experimental results.
The formal assumptions made are:
gNa=-- n0hgNa, (14)
diJm (1-m)-Pmm, (15)
dt = (1-h)-#,%hi (16)
where 9Na is a constant and the oc's and P's are functions of V but not of t.
These equations may be given a physical basis if sodium conductance i's
assumed to be proportional to the number ofsites on the inside of the membrane
which are occupied simultaneously by three activating molecules but are not
blocked by an inactivating molecule. m then represents the proportion of
activating molecules on the inside and 1-rn the proportion on the outside;
h is the proportion of inactivating molecules on the outside and 1-h the
proportion on the inside. a,m or # and Pm or Oh represent the transfer rate
constants in the two directions.
Application of these equations will be discussed first in terms of the family
of curves in Fig. 6. Here the circles are experimental estimates of the rise and
fall of sodium conductance during a voltage clamp, while the smooth curves
were calculated from eqns. (14)-(16).
The solutions of eqns. (15) and (16) which satisfy the boundary conditions
m=mo and h=ho at t=0 are
M = m,0-(mo- mo) exrp (tm)(17)
h = h.-(h. -ho) exp (-t/lrh), (18)
where m or/(oc, + fl) and Tmm=l/(am+fPm)9
h> = (I/(Ch +Ph) and Th = I /(h +Ph)-
In the resting state the sodium conductance is very small compared with the
value attained during a large depolarization. We therefore neglect rn if the
depolarization is greater than 30 mV. Further, inactivation is very nearly
complete if V < -30 mV so that h., may also be neglected. The expression for
the sodium conductance then becomes
9Na=gNa [1- exp (-rtI)]3 exp (- tI'T), (19)
109 o o o A
100L
,88 _
E
0 1-
E 76
u
63 0
E~~~~~~~~~~~~~~~
0 51 i
E 38G
°l 32 < ~
19
10 E I L
0.
6 ]0
0 1 2 3 4 5 6 7 8 9 10 11 msec
Fig. 6. Changes of sodium conductance associated with different depolarizations. The circles are
experimental estimates of sodium conductance obtained on axon 17, temperature 6-70 C
(of. Fig. 3). The smooth curves are theoretical curves with parameters shown in Table 2;
A to H drawn from eqn. 19, I to L from 14, 17, 18 with gN, = 70*7 m.mho/cm2. The ordinate
scales on the right are given in m.mho/cm2. The numbers on the left show the depolarization
in mV. The time scale applies to all curves.
where g.=9N.m3. ho and is the value which the sodium conductance would
attain if h remained at its resting level (ho). Eqn. (19) was fitted to an experi-
mental curve by plotting the latter on double log paper and comparing it with
a similar plot of a family of theoretical curves drawn with different ratios of
.m to rh. Curves A to H in Fig. 6 were obtained by this method and gave the
PH. CXVII. 33
A. L. HODGKIN AND A. F. HUXLEY
0rI)~O
C
d~~~~~0
eDo~~ ~ ~ (
.#
0
0
'4a
w
co
C-a
9
-
El~~~~~~~~~~~~~~~~C
mmCaC
~~
0000~~~M to=08
_o
00_
M00-0000000000
<>ee0000
~~~o m
>u>c
o os
oo
e
00c
_
oo
0
e
o
~ ~~~a
5;:
00
10
_o
ron ce
00ca
t>o
~
Lo_ oto
~
~
xo
:c
cq ~
Co
$
~
___
000
~
~
cS
~
o
~
OQ
o
_
_
0~
rs
2
Q
<
~8
r
Ca
Hs
514
values of g9, Tm and -rh shown in Table 2. Curves I to L were obtained from
eqns. (17) and (18) assuming that h. and Th had values calculated from experi-
ments described in a previous paper (Hodgkin & Huxley, 1952 c).
The rate constants aCm and Pm. Having fitted theoretical curves to the experi-
mental points, 0wm and Pm were found by a procedure similar to that used with
an and fln, i.e.
amm=m,oom,
/ mT=(P -m)ITm.
the value of m0O being obtained from -/gN on the basis that m. approaches
unity at large depolarizations.
°tm fm Axon
10 _ a 17
VV
9 v 0 v 20
8 Ak* 21
v v 21
E
0 v .~~~~~~~~41
-6 c m
C
0~~~~
40
8-110-0-08-0-05-03-01 0 0
2
1
e
-110 -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 +10
V (rr)
Fig. 7. Abscissa: membrane potential minus resting potential in sea water. Ordinate: rate
constanta (a. and f,m) determining initial changes in sodium conductance at 60 C. All values
for V < 0 were obtained by the method illustrated by Fig. 6 and Table 2; the value at V = 0
was obtained from the decline in sodium conductance associated with repolarization to the
resting potential. The temperature varied between 3 and 1100 and was allowed for by
assuming a Qlo of 3. The smooth curves were drawn from eqns. (20) and (21).
Values of om and Pm were collected from different experiments, reduced to
a temperature of 60 C by adopting a Qlo of 3 and plotted in the manner shown
in Fig. 7. The point for V= 0 was obtained from what we regard as the most
reliable estimate of the rate constant determining the decline of sodium con-
ductance when the membrane is repolarized (Hodgkin & Huxley, 1952b,
table 1, axon 41). The smooth curves in Fig. 7 were drawn according to the
equations: V +25
um=to1 (V+ 25)/(exp 10 -1), (20)
Pm= 4 exp (V/18), (21)
where xm and Pm are expressed in msec-I and V is in mV.
33-2
Fig. 8 illustrates the relation between m. and V. The symbols are experi-
mental estimates and the smooth curve was calculated from the equation
moo =ocmI(om +Pm) (22)
where am and Pm have the values given by eqns. (20) and (21).
The rate constants owh and Ph. The rate constants for the inactivation process
were calculated from the expressions
x,h hx./,Ir,
=
Axon
o 17
m.m o 2
1*09 , v 21
08 V
07
06
0,5
04
03
02
01
-110-100 -90 -80 -70 -60 -50 -40-30-20-10 0 +10
V(mV)
Fig. 8. Abscissa: membrane potential minus resting potential in sa water. Ordinate: mt
obtained by fitting curves to observed changes in sodium conductance at different depolariza-
tions (e.g. Fig. 6 and Table 2). The smooth curve is drawn according to eqn. (22). The
experimental points are proportional to the cube root of the sodium conductance which
would have been obtained if there were no inactivation.
Values obtained by these equations are plotted against membrane potential
in Fig. 9. The points for V < -30 mV were derived from the analysis described
in this paper (e.g. Table 2), while those for V > -30 mV were obtained from
the results given in a previous paper (Hodgkin & Huxley, 1952 c). A tem-
perature coefficient of 3 was assumed and differences in resting potential were
allowed for by taking the origin at a potential corresponding to h. = 0-6.
The smooth curves in this figure were calculated from the expressions
ahX=0-07 exp (V/20), (23)
and II V+301)
Ph --Jrx 10 +I(24)
The steady state relation between h.0 and V is shown in Fig. 10. The smooth
curve is calculated from the relation
h0 = (ch/(ch + Ph)' (25)
15 - a o 15
14 -
Axon 1-4
13 -
17 _ 13
1.2 -
v I
|20 Ah. 13
21 methodA 1-2
~~~~~~~~~~~~~~V21J
1-1
v 37
1 10 ;v38 1 fih, 1.0
el *I 139
a method B
0-9 -As v v *39*
A _ 09
A
ul
08 ° 2 ° 8 ° \D o- I| V
37
o 38 °xh, _ 08
O 0-7 -
A
^ \I I ca 39 method B 0-7
U Q J~~~~~~~~
x v \ x 39*0
0,,06 -
A 0-6
05 -
L_________- 0-5 -
0,4 It 0-4
03 -: ah 0-3
0,2 -
\ S/ _02
01 _ e 01
0 ~~~~~~~~~~~~~~~0
-110-100-90-80-70-60-50-40-30-20-10 0 10 20 30
V (mV)
Fig. 9. Rate constants of inactivation (ah and fPh) as functions of membrane potential (V). The
smooth curves were calculated from eqns. (23) and (24). The experimental values of ah and fih
were obtained from data such as those in Table 2 of this paper (methodA) or from the values of
Th and ho:> given in Table 1 of Hodgkin & Huxley (1952c) (method B). Temperature differences
were allowed for by scaling with a Qlo of 3. Axon 39 was at 190C; all others at 3-9° C. The
values for axons 37 and 39* were displaced by - 1-5 and - 12 mV in order to give h. = 0-6 at
V =0.
with ah and Ph given by eqns. (23) and (24). If V > -30 mV this expression
approximates to the simple expression used in a previous paper (Hodgkin &
Huxley, 1952 c), i.e. h=/(+expVh - V)
where Vh is about -2 and is the potential at which h., = 0-5. This equation is
the same as that giving the effect of a potential difference on the proportion
of negatively charged particles on the outside of a membrane to the total
number of such particles on both sides of the membrane (see p. 503). It is there-
fore consistent with the suggestion that inactivation might be due to the
movement of a negatively charged particle which blocks the flow of sodium
ions when it reaches the inside of the membrane. This is encouraging, but it
must be mentioned that a physical theory of this kind does not lead to
satisfactory functions for och and h without further ad hoc assumptions.
h,x, h.
10 -
, 10
0;9 Axon 09
O 38
08 o 39 08
07 & 39* 07
06- 06
05 0_5
04 04
0-3 0-3
02- 02
0*1 0 01
0 0
-100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50
V (mV)
Fig. 10. Steady state relation between h and V. The smooth curve is drawn according to eqn. (25).
The experimental points are those given in Table 1 of Hodgkin & Huxley (1952c). Axon 38
(50 C) as measured. Axon 39 (190 C) displaced - 1-5 mV. Axon 39* (30 C, fibre in derelict
state) displaced - 12 mV. The curve gives the fraction of the sodium-carrying system which
is readily available, as a function of membrane potential, in the steady state.
PART III. RECONSTRUCTION OF NERVE BEHAVIOUR

The remainder of this paper will be devoted to calculations of the electrical
behaviour of a model nerve whose properties are defined by the equations
which were fitted in Part II to the voltage clamp records described in the
earlier papers of this series.
Summary of equations and parameters
We may first collect the equations which give the total membrane current I
as a function of time and voltage. These are:
dV
I=CMdt + #Kn (V-VK) + 9Nam3h (V-VNa) + 91 (VV-) (26)
where dnl/dt ocn(l-n)-Pn,
= (7)
dm/dt= m(1-m)-P.m, (15)
clh/clt oc,(l1-h) # h)
= - ( 16)
and yn=001 (V+10)/(exp 1O -1), (12)
/n3=0.125 exp (V/80), (13)
om=O-l (V+25)/(exp VZ2 -1), (20)

f3m=4 exp (V/18), (21)
Xh= 007 exp (V120), (23)
h=1/(exp V+30 +1). (24)
Equation (26) is derived simply from eqns. (1)-(6) and (14) in Part II. The
four terms on the right-hand side give respectively the capacity current, the
current carried by K ions, the current carried by Na ions and the leak current,
for 1 cm2 of membrane. These four components are in parallel and add up to
give the total current density through the membrane I. The conductances to
K and Na are given by the constants gK and 9NaX together with the dimension-
less quantities n, m and h, whose variation with time after a change of
membrane potential is determined by the three subsidiary equations (7), (15)
and (16). The a's and ,B's in these equations depend only on the instantaneous
value of the membrane potential, and are given by the remaining six equations.
Potentials are given in mV, current density in IzA/cm2, conductances in
m.mho/cm2, capacity in juF/cm2, and time in msec. The expressions for the
cc's and P's are appropriate to a temperature of 6.30 C; for other temperatures
they must be scaled with a Qlo of 3.
The constants in eqn. (26) are taken as independent of temperature. The
values chosen are given in Table 3, column 2, and may be compared with the
experimental values in columns 3 and 4.
Membrane currents during a voltage clamp
Before applying eqn. (26) to the action potential it is well to check that it
predicts correctly the total current during a voltage clamp. At constant
voltage dV/dt=0 and the coefficients a and ,B are constant. The solution is
then obtained directly in terms of the expressions already given for n, m and h
(eqns. (8), (17) and (18)). The total ionic current was computed from these for
a number of different voltages and is compared with a series of experimental
curves in Fig. 11. The only important difference is that the theoretical current
has too little delay at the sodium potential; this reflects the inability of our
equations to account fully for the delay in the rise of g9 (p. 509).
'Membrane' and propagated action potentials
By a 'membrane' action potential is meant one in which the membrane
potential is uniform, at each instant, over the whole of the length of fibre
A. L. HODGKIN AND A. F. HUXLEY
.-4
- 0>
o >z
°
000
._0
eqeoq±
*1
C)
P
0
0
-CB
C)
14.
3~
0
o0
e
1
*
Ca
_0 eq
i'iv
'0CBql
es
-
EH
co
CO
eq
eq
P-
_
I--
*
X
Cz w
+
'+
1+q
_
-+
-434
d
-
^4
?
:.
m
aI
C)
&
0
10
0
O
V
r
eo-
P..
520
a ~ 5 mV 120!LA/cm2 -5 mV I20yA/cm2
-10 mV
C
-2O~~~mVI1OOpA JCM -18 mV I1 pA/cm2
D I05uAIcM2
d f00ouAIcM2 -7m
-30 mV
0 10 20 msec
I-I
0 10
I -
20 m.ec
. . . _
I
_
I I I
N-130mV~
oI * 2I I 4I 5I I * Im * 0 1*2 3 4 5 6 msec | |
0 1 2 3 4 5 6 7 msec 0 1 2 3 4 5 6 7 msec
Fig. 11. Left-hand column: time course of membrane current during voltage clamp, calculated
for temperature of 40 C from eqn. (26) and subsidiaries and plotted on the same scale as the
experimental curves in the right-hand column. Right-hand column: observed time course of
membrane currents during voltage clamp. Axon 31 at 40 C; compensated feedback. The
time scale changes between d, D and e, E. The current scale changes after b, B; c, C; d, D
andf, F.
considered. There is no current along the axis cylinder and the net membrane
current must therefore always be zero, except during the stimulus. If the
stimulus is a short shock at t = 0, the form of the action potential should be
given by solving eqn. (26) with I = 0 and the initial conditions that V = Vo and
m, n and h have their resting steady state values, when t = 0.
The situation is more complicated in a propagated action potential. The fact
that the local circuit currents have to be provided by the net membrane
current leads to the well-known relation
1 a2V
ax2' (27)
where i is the membrane current per unit length, r1 and r2 are the external and
internal resistances per unit length, and x is distance along the fibre. For an
axon surrounded by a large volume of conducting fluid, r, i8 negligible com-
pared with r2. Hence 1a V
r2 ax2
or =2R
a
Va (28)
where I is the membrane current density, a is the radius of the fibre and R2 is
the specific resistance of the axoplasm. Inserting this relation in eqn. (26),
we have
a a2V av 4V-
2R2 ax2 =CMat+SEn4 ( V-VK) + gNam3h(V-VNa) + gl (V-VI), (29)
the subsidiary equations being unchanged.
Equation (29) is a partial differential equation, and it is not practicable to
solve it as it stands. During steady propagation, however, the curve of V
against time at any one position is similar in shape to that of V against distance
at any one time, and it follows that
a2v 1 a2v
ax2= 2 at2)
where 0 is the velocity of conduction. Hence
a d2V dV
2R202 dt2 CM dt +Kn4 (V-VK)+gNm3kh(V-VNa)+g9(V-V). (30)
This is an ordinary differential equation and can be solved numerically, but
the procedure is still complicated by the fact that 0 is not known in advance.
It is necessary to guess a value of 0, insert it in eqn. (30) and carry out the
numerical solution starting from the resting state at the foot of the action
potential. It is then found that V goes off towards either + oo or -oo,
according as the guessed 0 was too small or too large. A new value of 0 is
then chosen and the procedure repeated, and so on. The correct value brings
V back to zero (the resting condition) when the action potential is over.
The solutions which go towards + oo correspond to action potentials
travelling slower than normal under a travelling anode or faster than normal
under a travelling cathode. We suspect that a system which tends to - c for
all values of 0 after an initial negative displacement of V is one which is
incapable of propagating an action potential.
NUMERICAL METHODS
Membrane action potentisal
Integration procedure. The equations to be solved are the four simultaneous first-order equations
(26), (7), (15), and (16) (p. 518). After slight rearrangement (which will be omitted in this descrip-
tion) these were integrated by the method of Hartree (1932-3). Denoting the beginning and end
of a step by to and t1 ( = to +St) the procedure for each step was as follows:
(1) Estimate V1 from VO and its backward differences.
(2) Estimate n, from no and its backward differences.
(3) Calculate (dn/dt), from eqn. 7 using the estimated na and the values of a,, and P., appropriate
to the estimated V1.
(4) Calculate n1 from the equation
n
t tdn\ Idn\ 1 F2 Idn\ Idn
nl ~ =2 {(dti)o + (d-t 12 L (dt)o ( dt,1jj
A2(dn/dt) is the second difference of dn/dt; its value at tL has to be estimated.
(5) If this value of n1 differs from that estimated in (2), repeat (3) and (4) using the new n1. If
necessary, repeat again until successive values of na are the same.
(6) Find ml and hI by procedures analogous to steps (2)-(5).
(7) Calculate g{Kn and gN.m3hk.
(8) Calculate (dV/dt), from eqn. 26 using the values found in (7) and the originally estimated
VI-
(9) Calculate a corrected V1 by procedures analogous to steps (4) and (5). This result never
differed enough from the original estimated value to necessitate repeating the whole procedure
from step (3) onwards.
The step value had to be very small initially (since there are no differences at t =0) and it also
had to be changed repeatedly during a run, because the differences became uinmanageable if it
was too large. It varied between about 0-01 msec at the beginning of a run or 0-02 msec during
the rising phase of the action potential, and 1 msec during the small oscillations which follow the
spike.
Accuracy. The last digit retained in V corresponded to microvolts. Sufficient digits were kept
in the other variables for the resulting errors in the change of V at each step to be only occasionally
as large as 1 IV. It is difficult to estimate the degree to which the errors at successive steps
accumulate, but we are confident that the overall errors are not large enough to be detected in the
illustrations of this paper.
Temrperature differences. In calculating the action potential it was convenient to use tables
giving the cc's and ,'s at intervals of 1 mV. The tabulated values were appropriate to a fibre at
6.30 C. To obtain the action potential at some other temperature T' °C the direct method would
be to multiply all a's and P's by a factor q =o3(T'-63)/10, this being correct for a QLO of 3. Inspection
of eqn. 26 shows that the same result is achieved by calculating the action potential at 6.30 C with
a membrane capacity of #CMj !F/cm2, the unit of time being 1/0 msec. This method was adopted
since it saved recalculating the tables.
Propagated action potential
Equation. The main equation for a propagated action potential is eqn. (30). Introducing
a quantity K =2R2(VMm/a, this becomes
dPjid$=KRW(1dtdt + 0'M
C SgKn4 ( V - Vs + ^M3h ( V - VNJ) + g, ( V - VI)]) * (31)
The subsidiary equations (7), (15) and (16), and the a's and ,'s, are the same as for the membrane
equation.
Integration procedure. Steps (1)-(7) were the same as for the membrane action potential. After
that the procedure was as follows:
(8) Estimate (dV/dt), from (dV/dt)o and its backward differences.
(9) Calculate (d3V/dt2)1 from eqn. (31), using the values found in (7) and the estimated values
of V1 and (dV/dt)1.
(10) Calculate a corrected (dV/dt), by procedures analogous to steps (4) and (5).
(11) Calculate a corrected V1 by a procedure analogous to step (4), using the corrected (dV/dt)1.
(12) If necessary, repeat (9)-(11) using the new V1 and (dV/dt)1, until successive values of V1
agree.
Starting conditions. In practice it is necessary to start with V deviating from zero by a finite
amount (0.1 mV was used). The first few values of V, and hence the differences, were obtained as
follows. Neglecting the changes in 9K and gNa, eqn. (31) is
d2V K d V gV
dt' dt -C3
where go is the resting conductance of the membrane. The solution of this equation is V = VoeMt,
where , is a solution of
IL" - KjL - Kgo/lm=0. (32)
When K has been chosen, ju can thus be found and hence V1, V., etc. (Voeimti, V0eMtu, etc.).
After several runs had been calculated, so that K was known within fairly narrow limits, time
was saved by starting new runs not from near V = 0 but from a set of values interpolated between
corresponding points on a run which had gone towards + oo and another which had gone
towards - o.
Choice of K. The value of K chosen for the first run makes no difference to the final result, but
the nearer it is to the correct value the fewer runs will need to be evaluated. The starting value
was found by inserting in eqn. (32) a value of , found by measuring the foot of an observed action
potential.
Caculation of falling phase. The procedure outlined above is satisfactory for the rising phase
and peak of the action potential but becomes excessively tedious in the falling phase and the
oscillations which follow the spike. A different method, which for other reasons is not applicable
in the earlier phases, was therefore employed. The solution was continued as a membrane action
potential, and the value of d2V/dt2 calculated at each step from the differences of dV/dt. From
these it was possible to derive an estimate of the values (denoted by z) that d2 V/dt2 would have
taken in a propagated action potential. The membrane solution was then re-calculated using the
following equation instead of eqn. (31):
dV 1
t
K
(V - VK) + 9Nam8h (V - VN.) + 91 (VV-Vd} + K (33)
This was repeated until the z's assumed for a particular run agreed with the d2V/dt2's derived
from the same run. When this is the case, eqn. (33) is identical with eqn. (31), the main equation
for the propagated action potential.
RESULTS
Membrane action potentials
Form of action potential at 60 C. Three calculated miembrane action potentials,
with different strengths of stimulus, are shown in the upper part of Fig. 12.
Only one, in which the initial displacement of membrane potential was 15 mV,
is complete; in the other two the calculation was not carried beyond the middle
of the falling phase because of the labour involved and because the solution
110
100
80
70
---
E 60
::.50
1 40
20
10
0
80
~7030 1is3 4
E 60
>.50 msec
1 40
20
10
0
0 1 2 3 4 5 6
msec
Fig. 12. Upper family: solutions of eqn. (26) for initial depolarizations of 90, 15, 7 and 6 mV
(calculated for 60C). Lower family: tracings of membrane action potentials recorded at
6' C from axon 17. The numbers attached to the curves give the shock strength in
m1Lcoulomb/cm2. The vertical and horizontal scales are the same in both families (apart from
the slight curvature indicated by the 110 mV calibration line). In this and all subsequent
figures depolarizations (or negative displacements of V) are plotted upwards.
had become almost identical with the 15 mV action potential, apart from the
displacement in time. One solution for a stimulus just below thireshold is also
shown.
The lower half of Fig. 12 shows a corresponding series of experimental
membrane action potentials. It will be seen that the general agreement is
good, as regards amplitude, form and time-scale. The calculated action
potentials do, however, differ from the experimental in the following respects:
(1) The drop during the first 01 msec is smaller. (2) The peaks are sharper.
(3) There is a small hump in the lower part of the falling phase. (4) The ending
of the falling phase is too sharp. The extent to which these differences are the
result of known shortcomings in our formulation will be discussed on pp. 542-3.
The positive phase of the calculated action potential has approximately the
correct form and duration, as may be seen from Fig. 13 in which a pair of
curves are plotted on a slower time scale.
100 _
90 \
80
70 _
> 60 -
E 50 -
>40
30
20 -
10
0
0 5 1 15 20msec
100 lOOMV
90
80
70
>.
E 60 s
40-
30-
20-
10'
0 -155mse~c
Fig. 13. Upper curve: solution of eqn. (26) for initial depolarization of 15 mV, calculated for
60 C. Lower curve: tracing of membrane action potential recorded at 9 10 C (axon 14). The
vertical scales are the same in both curves (apart from curvature in the lower record).
The horizontal scales differ by a factor appropriate to the temperature difference.
Certain measurements of these and other calculated action potentials are
collected in Table 4.
Form of action potential at 18-50 C. Fig. 14 shows a comparison between
a calculated membrane action potential at 18.50 C and an experimental one
at 20.50 C. The same differences can be seen as at the low temperature, but,
except for the initial diop, they are less marked. In both the calculated and the
experimental case, the rise of temperature has greatly reduced the duration of
the spike, the difference being more marked in the falling than in the rising
phase (Table 4), as was shown in propagated action potentials by Hodgkin &
Katz (1949).
The durations of both falling phase and positive phase are reduced at the
higher temperature by factors which are not far short of that (3 84) by which
the rate constants of the permeability changes are raised (Qlo = 3.0). This is the
justification for the differences in time scale between the upper and lower parts
in Figs. 13 and 14.
IQz,=°B$=~ ~ EC).l 14697793, 1952, 4, Downloaded from https://physoc.onlinelibrary.wiley.com/doi/10.1113/jphysiol.1952.sp004764 by Cochrane Peru, Wiley Online Library on [27/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
_4Cs4I4-4
~Oq C)
co ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~~~~3
0 ~ ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~U 0
4. ~
~ ~ ~ ~ ~ ~ ~ ~ -
* 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3j s $II.Ie
10
000 4)~~~~~~~~~~~~~~
e *S
4 C0C 4L4
~~~C0C0~~(4 4co
4D
ia?14 '0t- 10"4*e

e4
00C 01 U
S X -~~~~~~~- ro
_o "4:,X
Ca
0 0 0~~~~~~~~~~~
14 ~ ~ ~ ~ ~ ~
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~-
a?~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~oC
H ~~~~~~~~~~~~~~~~~~~~
i
a?
a?3 a?;?a
a
14 1414 1-4 °4 4
Propagated action potential
Form of propagated action potential. Fig. 15 compares the calculated pro-
pagated action potential, at 18.50 C, with experimental records on both fast
and slow time bases. As in the case of the membrane action potential, the only
differences are in certain details of the form of the spike.
Velocity of conduction. The value of the constant K that was found to be
needed in the equation for the propagated action potential (eqn. 31) was
10-47 msec-1. This constant, which depends only on properties ofthe membrane,
100 100:
A B
E 50 ~~~~~~~~~~~E
1 2 msec
lOOm.0
t~~~
4
d00m on ion Dc
E
> ~~~~~~C>onl
E
0 12 msec~
0 54,fo5 10
msec
Fig.neve
15. A, solution of eqn. (31) calculated for K of 10*47 msec-1 and temperature of 18.50 C.
B, fibr cnieeasacbeThreainiginbytedeitonfK
same solution plotted on slower time scale. C, tracing of propagated action potential on
same vertical and horizontal scales a A. Temperature 18.50 C. D, tracing of propagated
action potential from another axon on approximately the same vertical and horizontal scales
as B. Temperature 19.20 C. This axon had been used for several hours; its spike was initially
100 mV.
determines the conduction velocity in conjunction with the constants of the

nerve fibre considered as a cable. The relation is
given by the definition of K
(p. 524), from which
= V(KaI22ICm), (34)
where 0= conduction velocity, a = radius of axis cylinder, 14= specific
resistance of axoplasm, and C = capacity per unit area of membrane.
The propagated action potential was calculated for the temperature at which
the record C of Fig. 15 was obtained, and with the value of Cm (F-01F/cm2)
that was measured on the fibre from which that record was made. Since 6, a
and 14 were also measured on that fibre, a direct comparison between calcu-
lated and observed velocities is possible. The values of a and R2 were 238 p and
35-4 Q. cm respectively. Hence the calculated conduction velocity 'is
(10470 x 0-0238/2 x 35-4 xJ10-6)1 cm/sec = 18-8 in/sec.
The velocity found experimentally in this fibre was 21-2 m/sec.
Impedance changes
Time course of conductance change. Cole & Curtis (1939) showed that the
impedance of the membrane fell during a spike, and that the fall was due to
a great increase in the conductance which is in parallel with the membrane
capacity. An effect of this kind is to be expected on our formulation, since the
entry of Na+ which causes the rising phase, and the loss of K+ which causes
the falling phase, are consequent on increases in the conductance of the
membrane to currents carried by these ions. These component conductances
are evaluated during the calculation, and the total conductance is obtained by
adding them and the constant 'leak conductance', 9.
mV A m.mho/cm2 B
100 40
50
j20
1._. I 1. .1 1 1 11111
0 1 2 3 4 5678910msec
0 1 2 3 4 5678910msec
Fig. 16. A, solution of eqn. (26) for initial depolarization of 15 mV at a temperature of 6° C. The
broken curve shows the membrane action potential in mV; the continuous curve shows the
total membrane conductance (gN +g9 + 9) as a function of time. B, records of propagated
action potential (dotted curve) and conductance change reproduced from Cole & Curtis (1939).
The time scales are the same in A and B.
Fig. 16A shows the membrane potential and conductance in a calculated

membrane action potential. For comparison, Fig. 16B shows superposed
records of potential and impedance bridge output (jroportional to conductance
change), taken from Cole & Curtis's paper. The time scale is the same inBas inA,
and the curves have been drawn with the same peak height. It will be seen that
the main features of Cole & Curtis's record are reproduced in the calculated
curve. Thus (1) the main rise in conductance begins later than the rise of
potential; (2) the conductance does not fall to its resting value until late in
the positive phase; and (3) the peak of the conductance change occurs at
nearly the same time as the peak of potential. The exact time relation between
the peaks depends on the conditions, as can be seen from Table 4.
We chose a membrane action potential for the comparison in Fig. 16 because
the spike duration shows that the experimental records were obtained at about
60 C, and our propagated action potential was calculated for 18.50 C. The
conductance during the latter is plotted together with the potential in Fig. 17.
The same features are seen as in the membrane action potential, the delay
PH. CXVII. 34
between the rise of potential and the rise of conductance being even more
marked.
Absolute value of peak conductance. The agreement between the height of the
conductance peak in Fig. 16A and the half-amplitude of the bridge output in
Fig. 16B is due simply to the choice of scale. Nevertheless, our calculated
action potentials agree well with Cole & Curtis's results in this respect.
These authors found that the average membrane resistance at the peak of
the impedance change was 25 Q.cm2, corresponding to a conductance of
40 m.mho/cm2. The peak conductances in our calculated action potentials
ranged from 31 to 53 m.mho/cm2 according to the conditions, as shown in
Table 4.
mV m.mho/cm2
903 -V
80 -
30
70-
60 -25
50 20 g
30
10,
20
10 gK
0 C,
Fig. 17. Numerical solution of eqn. (31) showing components of membrane conductance (g) during
propagated action potential ( - V). Details of the analysis are as in Fig. 15.
Components of conductance change. The manner in which the conductances
to Na+ and K+ contribute to the change in total conductance is shown in
Fig. 17 for the calculated propagated action potential. The rapid rise is due
almost entirely to sodium conductance, but after the peak the potassium con-
ductance takes a progressively larger share until, by the beginning of the
positive phase, the sodium conductance has become negligible. The tail of
raised conductance that falls away gradually during the positive phase is due
solely to potassium conductance, the small constant leak conductance being of
course present throughout.
Ionic movements
Time course of ionic currents. The time course ofthe components of membrane
current carried by sodium and potassium ions during the calculated pro-
pagated spike is shown in Fig. 18 C. The total ionic current contains also
a small contribution from 'leak current' which is not plotted separately.
Two courses are open to current which is carried into the axis cylinder by
ions crossing the membrane: it may leave the axis cylinder again by altering
the charge on the membrane capacity, or it may turn either way along the
axis cylinder making a net contribution, I, to the local circuit current. The
magnitudes of these two terms during steady propagation are -CCMdV/dt and
(Cm/K) d2V/dt2 respectively, and the manner in which the ionic current is
divided between them at the different stages of the spike is shown in Fig. 18B.
It will be seen that the ionic current is very small until the potential is well
beyond the threshold level, which is shown by Fig. 12A to be about 6 mV.
100
> 50E /
soo w B
ce00 _CMCL)aV
E 300 dt
2300-\ /
1400-\ /
-600 ' I
-200 -\/
-o00 . 1
During this perio

suthn creqnt f) charing he o membrane capaity com
palmotenatirelypfromnloal. circuits.aTe fatenthattheV)

B ionic current does, nacty
partlynto th smalld nes tofthlmmbaecurrentsreached-
in an c,ircmtniacuresnea the,soiu
th rreshold,&and ptartly toctheedelay withewhich soaedppium conucante rises. whenal
thoftheantalysis lowredani.15
Tlost movementsofro ionaicis.
toanrfasoiu
The tand loss
ofni poretassium cno
be obtained by integrating the corresponding ionic currents over the whole

34-2
impulse. This has been done for the four complete action potentials that we
calculated, and the results are given in Table 5. It will be seen that the results
at 18.50 C are in good agreement with the values found exp6imentally by
Keynes (1951) and Keynes & Lewis (1951), which were obtained at com-
parable temperatures.
Ionicfluxes. The flux in either direction of an ion can be obtained from the
net current and the equilibrium potential for that ion, if the independence
principle (Hodgkin & Huxley, 1952 a) is assumed to hold. Thus the outward
flux of sodium ions is INa/(exp (VV-Na) F/RT- 1), and the inward flux of
potassium ions is -IK/(exp (VK- V) FIRT -1). These two quantities were
evaluated at each step of the calculated action potentials, and integrated over
the whole impulse. The integrated flux in the opposite direction is given in
each case by adding the total net movement. The results are given in Table 5,
where they can be compared with the results obtained with radioactive tracers
by Keynes (1951) on Sepia axons. It will be seen that our theory predicts too
little exchange of Na and too much exchange of K during an impulse. This
discrepancy will be discussed later.
Refractory period
Time course of inactivation and delayed rectification. According to our theory,
there are two changes resulting from the depolarization during a spike which
make the membrane unable to respond to another stimulus until a certain
time has elapsed. These are 'inactivation', which reduces the level to which
mV
100 -v
50_I\
m.mho/cm2 06
gK 0-5h
10 h 04
0-3
5* 002
0 5 10 15 20 msec
Fig. 19. Numerical solution of eqn. (26) for initial depolarization of 15 mV and temperature of
6° C. Upper curve: membrane potential, as in Fig. 13. Lower curves show time course of
9E and h during action potential and refractory period.
the sodium conductance can be raised by a depolarization, and the delayed

rise in potassium conductance, which tends to hold the membrane potential
near to the equilibrium value for potassium ions. These two effects are shown
in Fig. 19 for the calculated membrane action potential at 60 C. Both curves
reach their normal levels again near the end of the positive phase, and finally
533
MEMBRANE CURRENT IN NERVE
BE4
tH
Ca
C 4
4
Wt
as~
k 4
C
1i 0= toj 10~~~~~~~~~~1
Q
A~~~~~ t 15~~C)
.
Ct
COCDC
t
--s
i 2P
m
X
o
~~ *- 'C
.4
c
10bo
C
,~4~
Pq°
~~~~C)
z
o4
C)
B]
0
@
0
M
g
.C~~
O.
~
r @
C)
~
H ~^ C
4,D rO,,
D
.
Ca
oX
H~ ~
4P4
r V
s
44
k
g
Ca
E~.s
o
bO
C)
0
b-
Ez
settle down after a heavily damped oscillation of small amplitude which is not
seen in the figure.
Responses to stimuli during positive phase. We calculated the responses of
the membrane when it was suddenly depolarized by 90 mV at various times
during the positive phase of the membrane action potential at 6° C. These are
shown by the upper curves in Fig. 20. After the earliest stimulus the
-V (mV) -V (mV)
100 _ 100
c D E
-
50
50
0O
msec msec
-V (mV)
100p
50
0
V L C 0 1 2mOeC
Fig. 20. Theoretical basis of refractory period. Upper curves: numerical solutions of eqn. (26) for
temperature of 60 C. Curve A gives the response to 15 mpcoulomb/cm2 applied instan-
taneously at t =0. Curve E gives the response to 90 mpcoulomb/cm2 again applied in the
resting state. Curves B to D show effect of applying 90 mucoulomb/cm2 at various times after
curve A. Lower curves: a similar experiment with an actual nerve, temperature 90 C. The
voltage scales are the same throughout. The time scales differ by a factor appropriate to the
temperature difference.
membrane potential falls again with hardly a sign of activity, and the
membrane can be said to be in the 'absolute refractory period'. The later
stimuli produce action potentials of increasing amplitude, but still smaller
than the control; these are in the 'relative refractory period '. Corresponding
experimental curves are shown in the lower part of Fig. 20. The agreement is
good, as regards both the duration of the absolute refractory period and the
changes in shape of the spike as recovery progresses.
*Excitation
Our calculations of excitation processes were all made for the case where the
membrane potential is uniform over the whole area considered, and not for the
case of local stimulation of a whole nerve. There were two reasons for this:
first, that such data from the squid giant fibre as we had for comparison were
obtained by uniform stimulation of the membrane with the long electrode;
and, secondly, that calculations for the whole nerve case would have been
extremely laborious since the main equation is then a partial differential
equation.
Threshold. The curves in Figs. 12 and 21 show that the theoretical
'membrane' has a definite threshold when stimulated by a sudden displace-
ment of membrane potential. Since the initial fall after the stimulus is much
less marked in these than in the experimental curves, it is relevant to compare
the lowest point reached in a just threshold curve, rather than the magnitude
of the original displacement. In the calculated series this is about 6 mV and in
the experimental about 8 mV. This agreement is satisfactory, especially as the
value for the calculated series must depend critically on such things as the
leak conductance, whose value was not very well determined experimentally.
The agreement might have been somewhat less good if the comparison had
been made at a higher temperature. The calculated value would have been
much the same, but the experimental value in the series at 230 C shown in
Fig. 8 of Hodgkin et al. (1952) is about 15 mV. However, this fibre had been
stored for 5 hr before use and was therefore not in exactly the same state as
those on which our measurements were based.
Subthreshold responses. When the displacement of membrane potential was
less than the threshold for setting up a spike, characteristic subthreshold
responses were seen. One such response is shown in Fig. 12, while several
are plotted on a larger scale in Fig. 21 B. Fig. 21 A shows for comparison
the corresponding calculated responses of our model. The only appreciable
differences, in the size of the initial fall and in the threshold level, have been
mentioned already in other connexions.
During the positive phase which follows each calculated subthreshold
response, the potassium conductance is raised and there is a higher degree of
'inactivation' than in the resting state. The threshold must therefore be
raised in the same way as it is during the relative refractory period following
a spike. This agrees with the experimental findings of Pumphrey, Schmitt &
Young (1940).
Anode break excitation. Our axons with the long electrode in place often
gave anode break responses at the end of a period during which current was
made to flow inward through the membrane. The corresponding response of
out theoretical model was calculated for the case in which a current sufficient
to bring the membrane potential to 30 mV above the resting potential was
suddenly stopped after passing for a time long compared with all the time-
constants of the membrane. To do this, eqn. (26) was solved with I = 0 and the
initial conditions that V = + 30 mV, and m, n and h have their steady state
values for V = + 30 mV, when t = 0. The calculation was made for a temperature
5
E
t 0
_S
-101
I I I I I I I I I I I
0 1 2 3 4 5 6 7 8 9 10
msec
Fig. 21. A, numerical solutions of eqn. (26) for 6° C. The numbers attached to the curves give the
initial depolarization in mV (also the quantity of charge applied in m,ucoulomb/cm').
B, response of nerve membrane at 60C to short shocks; the numbers show the charge applied
in m,ucoulomb/cm2. The curves have been replotted from records taken at low amplification
and a relatively high time-base speed.
of 6 3° C. A spike resulted, and the time course of membrane potential is

plotted in Fig. 22A. A tracing of an experimental anode break response is
shown in Fig. 22B; the temperature is 18-50 C, no record near 6° being avail-
able. It will be seen that there is good general agreement. (The oscillations
after the positive phase in Fig. 22B are exceptionally large; the response of
this axon to a small constant current was also unusually oscillatory as shown
in Fig. 23.)
The basis of the anode break excitation is that anodal polarization decreases
the potassium conductance and removes inactivation. These effects persist for
an appreciable time so that the membrane potential reaches its resting value
with a reduced outward potassium current and an increased inward sodium
current. The total ionic current is therefore inward at V = 0 and the membrane
undergoes a depolarization which rapidly becomes regenerative.
1000
90
80 -
70-
60 -
. 40
30
20-
0
-10
-20 -
E0
50
0 1 20 436 msec
40-
-30
Fig. 22. Theoretical basis of anode break excitation. A, numerical solution of eqn. (26) for
boundary condition - V = - 30 mV for t <0; temperature 60 C. B, anode break excitation
following sudden cessation of external current which had raised the membrane potential by
26-5 mV; giant axron with long electrode at 18.50 0. Time scales differ by a factor appropriate
to the temperature difference.
Accommodation. No measurements of accommodation were made nor did we

make any corresponding calculations for our model. It is clear, however, that
the model will show 'accommodation' in appropriate cases. This may be
shown in two ways. First, during the passage of a constant cathodal current
through the membrane, the potassium conductance and the degree of inactiva-
tion will rise, both factors raising the threshold. Secondly, the steady state
ionic current at all strengths of depolarization is outward (Fig. 11), so that an
applied cathodal current which rises sufficiently slowly will never evoke a
regenerative response from the membrane, and excitation will not occur.
Oscillations
In all the calculated action potentials and subthreshold responses the
membrane potential finally returns to its resting value with a heavily damped
oscillation. This is well seen after subthreshold stimuli in Figs. 21A and 24,
but the action potentials are not plotted on a slow enough time base or with
a large enough vertical scale to show the oscillations which follow the positive
phase.
The corresponding oscillatory behaviour of the real nerve could be seen after
a spike or a subthreshold short shock, but was best studied by passing a small
constant current through the membrane and recording the changes of
membrane potential that resulted. The current was supplied by the long
internal electrode so that the whole area of membrane was subjected to
a uniform current density. It was found that when the current was very weak
the potential changes resulting from inward current (anodal) were almost
exactly similar to those resulting from an equal outward current, but -with
opposite sign. This is shown in Fig. 23B and C, where the potential changes
are about + 1 mV. This symmetry with weak currents is to be expected from
our equations, since they can be reduced to a linear form when the displace-
ments of all the variables from their resting values are small. Thus, neglecting
products, squares and higher powers of 81V, Sm, Sn and Ah, the deviations of V,
m, n and h from their resting values (0, moi, no and ho respectively), eqn. (26)
(p. 518) becomes
8 CM d8V
dt +9Kn408V-4gKn'3VK8n +gNa0 mho SV
- 39NaMohoVNa8m- Nano VNaSh + 9l 8 V. (35)
Similarly, eqn. (7) (p. 518) becomes
dSn
V (c~ +~)
-~-=~SV~ Sn-nO
P) Snn0 a (In+cn)
~ SY,
or (ac
n_
_____
(p+on+Pn)8n=n no ,V (36)
where p represents d/dt, the operation of differentiating with respect to
time.
The quantity Sn can be eliminated between eqns. (35) and (36). This process
is repeated for Sm and Sh, yielding a fourth-order linear differential equation
with constant coefficients for S V. This can be solved by standard methods for
any particular time course of the applied current density SI.
Fig. 23A shows the response of the membrane to a constant current pulse
calculated in this way. The constants in the equations are chosen to be
appropriate to a temperature of 18.50 C so as to make the result comparable
with the tracings of experimental records shown in B and C. It will be seen
that the calculated curve agrees well with the records in B, while those in C,
obtained from another axon, are much less heavily damped and show a higher
r,,11I,*, *,*I,,,I, . .,IuIII
i
0 5 10 15 20 25 msec
A T
/ \ T1~~~~~~~~~jkQ.f.cm2
140 E|Bi kQ1.cm2

E
-05tB2 f
k1l.cm2
0 5 10 15 20 25 msec
1-0
o.5F Ci Tikfl.CM2
-
0 C2 II kl.cm2
Fig. 23. A, solution of eqn. (35) for small constant current pulse; temperature 18.50 C; linear
approximation. The curve shows 8V/8I (plotted upwards) as a function of time. B, changes
in membrane potential associated with application of weak constant currents of duration
15 msec and strength ± 1.49 pA/cm2. B1, cathodic current; B2, anodic current. Depolariza.
tion is shown upward. Temperature 1900. C, similar records from another fibre enlarged to
have same time scale. Current strengths are ±0 55IA/cm2. Temperature 18° C. The response
is unusually oscillatory.
frequency of oscillation. A fair degree of variability is to be expected in these

respects since both frequency and damping depend on the values of the com-
ponents of the resting conductance. Of these, gNa and gK depend critically on
the resting potential, while gi is very variable from one fibre to another.
Both theory and experiment indicate a greater degree of oscillatory behaviour
than is usually seen in a cephalopod nerve in a medium of normal ionic com-
position. We believe that this is largely a direct result of using the long internal
electrode. If current is applied to a whole nerve through a point electrode,
neighbouring points on the membrane will have different membrane potentials
and the resulting currents in the axis cylinder will increase the damping.
The linear solution for the behaviour of the theoretical membrane at small
displacements provided a convenient check on our step-by-step numerical
procedure. The response of the membrane at 6.30 C to a small short shock was
calculated by this means and compared with the step-by-step solution for an
initial depolarization of the membrane by 2 mV. The results are plotted in
Fig. 24. The agreement is very close, the step-by-step solution deviating in the
direction that would be expected to result from its finite amplitude (cf. Fig. 21).
21
O0 10 20
o'Q ~~~~~~~~~msec
Fig. 24. Comparison of step-by-step solution and linear approximation. Eqn. (26), temperature
60 C; initial displacement of - V =2 mV. Continuous line: step-by-step solution. Circles:
ear approximation with same initial displacement.
As pointed out by Cole (1941), the process underlying oscillations in

membrane potential must be closely connected with the inductive reactance
observed with alternating currents. In our theoretical model the inductance is
due partly to the inactivation process and partly to the change in potassium
conductance, the latter being somewhat more important. For small displace-
ments of the resting potential the variations in potassium current in 1 cm2 of
membrane are identical with those in a circuit containing a resistance of 820 Q
in series with an inductance which is shunted by a resistance of 1900 Q. The
value of the inductance is 0-39 H at 250 C, which is of the. same order as the
0-2H found by Cole & Baker (1941). The calculated inductance increases
3-fold for a 100 C fall in temperature and decreases rapidly as the membrane
potential is increased; it disappears at the potassium potential and is replaced
by a capacity for E > E-.
DISCUSSION
The results presented here show that the equations derived in Part II of this
paper predict with fair accuracy many of the electrical properties of the squid
giant axon: the form, duration and amplitude of spike, both 'membrane'
and propagated; the conduction velocity; the impedance changes during the
spike; the refractory period; ionic exchanges; subthreshold responses; and
oscillations. In addition, they account at least qualitatively for many of the
phenomena of excitation, including anode break excitation and accommoda-
tion. This is a satisfactory degree of agreement, since the equations and
constants were derived entirely from 'voltage clamp' records, without any
adjustment to make them fit the phenomena to which they were subsequently
applied. Indeed any such adjustment would be extremely difficult, because
in most cases it is impossible to tell in advance what effect a given change in
one of the equations will have on the final solution.
The agreement must not be taken as evidence that our equations are any-
thing more than an empirical description of the time-course of the changes in
permeability to sodium and potassium. An equally satisfactory description of
the voltage clamp data could no doubt have been achieved with equations of
very different form, which would probably have been equally successful in
predicting the electrical behaviour of the membrane. It was pointed out in
Part II of this paper that certain features of our equations were capable of
a physical interpretation, but the success of the equations is no evidence in
favour of the mechanism of permeability change that we tentatively had in
mind when formulating them.
The point that we do consider to be established is that fairly simple perme-
ability changes in response to alterations in membrane potential, of the kind
deduced from the voltage clamp results, are a sufficient explanation of the
wide range of phenomena that have been fitted by solutions of the equations.
Range of applicability of the equations
The range of phenomena to which our equations are relevant is limited in
two respects: in the first place, they cover only the short-term responses of the
membrane, and in the second, they apply in their present form only to the
isolated squid giant axon.
Slow changes. A nerve fibre whose membrane was described by our equations
would run down gradually, since even in the resting state potassium leaves and
sodium enters the axis cylinder, and both processes are accelerated by activity.
This is no defect in describing the isolated squid giant axon, which does in fact
run down in this way, but some additional process must take place in a nerve
in the living animal to maintain the ionic gradients which are the immediate
source of the energy used in impulse conduction.
After-potentials. Our equations give no account of after-potentials, apart
from the positive phase and subsequent oscillations.
Conditions qf isolated giant axon. There are many reasons for supposing that
the resting potential of the squid giant axon is considerably lower after isola-
tion than when it is intact in the living animal. Further evidence for this view
is provided by the observation (Hodgkin & Huxley, 1952c) that the maximum
inward current that the membrane can pass on depolarization is increased by
previously raising the resting potential by 10-20 mV by means of anodally
directed current. Our equations could easily be modified to increase the resting
potential (e.g. by reducing the leak conductance and adding a small outward
current representing metabolic extrusion of sodium ions). We have not made
any calculations for such a case, but certain qualitative results are evident
from inspection of other solutions. If, for instance, the resting potential were
raised (by 12 mV) to the potassium potential, the positive phase and sub-
sequent oscillations after the spike would disappear, the rate of rise of the
spike would be increased, the exchange of internal and external sodium in
a spike would be increased, the membrane would not be oscillatory unless
depolarized, and accommodation and the tendency to give anode break
responses would be greatly reduced. Several of these phenomena have been
observed when the resting potential of frog nerve is raised (Lbrente de No,
1947), but no corresponding information exists about the squid giant
axon.
Applicability to other tissues. The similarity of the effects of changing the
concentrations of sodium and potassium on the resting and action potentials
of many excitable tissues (Hodgkin, 1951) suggests that the basic mechanism
of conduction may be the same as implied by our equations, but the great
differences in the shape of action potentials show that even if equations of
the same form as ours are applicable in other cases, some at least of the
parameters must have very different values.
Differences between calculated and observed behaviour
In the Results section, a number of points were noted on which the calcu-
lated behaviour of our model did not agree with the experimental results. We
shall now discuss the extent to which these discrepancies can be attributed to
known shortcomings in our equations. Two such shortcomings were pointed
out in Part II of this paper, and were accepted for the sake of keeping the
equations simple. One was that the membrane capacity was assumed to
behave as a 'perfect' condenser (phase angle 900; p. 505), and the other was
that the equations governing the potassium conductance do not give as much
delay in the conductance rise on depolarization (e.g. to the sodium potential)
as was observed in voltage clamps (p. 509).
The assumption of a perfect capacity probably accounts for the fact that
the initial fall in potential after application of a short shock is much less
marked in the calculated than in the experimental curves (Figs. 12 and 21).
Some of the initial drop in the experimental curves may also be due to end-
effects, the guard system being designed for the voltage clamp procedure but
not for stimulation by short shocks.
The inadequacy of the delay in the rise of potassium conductance has several
effects. In the first place the falling phase of the spike develops too early,
reducing the spike amplitude slightly and making the peak too pointed in
shape (p. 525). In the membrane action potentials these effects become more
marked the smaller the stimulus, since the potassium conductance begins to
rise during the latent period. This causes the spike amplitude to decrease
more in the calculated than in the experimental curves (Fig. 12).
The low calculated value for the exchange of internal and external sodium
ions is probably due to this cause. Most of the sodium exchange occurs near
the peak of the spike, when the potential is close to the sodium potential. The
early rise of potassium conductance prevents the potential from getting as close
to the socdium potential, and from staying there for as long a time, as it should.
A check on these points is provided by the 'anode break' action potential.
Until the break of the applied current, the quantity n has the steady state
value appropriate to V = + 30 mV, i.e. it is much smaller than in the usual
resting condition. This greatly increases the delay in the rise of potassium
conductance when the membrane is depolarized. It was found that the spike
height was greater (Table 4), the peak was more rounded, and the exchange
of internal and external sodium was greater (Table 5), than in an action
potential which followed a cathodal short shock.
The other important respect in which the model results disagreed with the
experimental was that the calculated exchange of internal and external
potassium ions per impulse was too large. This exchange took place largely
during the positive phase, when the potential is close to the potassium potential
and the potassium conductance is still fairly high. We have no satisfactory
explanation for this discrepancy, but it is probably connected with the fact
that the value of the potassium potential was less strongly affected by changes
in external potassium concentration than is required by the Nernst equation.
SUMMARY
1. The voltage clamp data obtained previously are used to find equations
which describe the changes in sodium and potassium conductance associated
with an alteration of membrane potential. The parameters in these equations
were determined by fitting solutions to the experimental curves relating
sodium or potassium conductance to time at various membrane potentials.
2. The equations, given on pp. 518-19, were used to predict the quantitative
behaviour of a model nerve under a variety of conditions which corresponded
to those in actual experiments. Good agreement was obtained in the following
cases:
(a) The form, amplitude and threshold of an action potential under zero
membrane current at two temperatures.
(b) The form, amplitude and velocity of a propagated action potential.
(c) The form and amplitude of the impedance changes associated with an
action potential.
(d) The total inward movement of sodium ions and the total outward
movement of potassium ions associated with an impulse.
(e) The threshold and response during the refractory period.
(f) The existence and form of subthreshold responses.
(g) The existence and form of an anode break response.
(h) The properties of the subthreshold oscillations seen in cephalopod axons.
3. The theory also predicts that a direct current will not excite if it rises
sufficiently slowly.
4. Of the minor defects the only one for which there is no fairly simple
explanation is that the calculated exchange of potassium ions is higher than
that found in Sepia axons.
5. It is concluded that the responses of an isolated giant axon of Lr5ligo to
electrical stimuli are due to reversible alterations in sodium and potassium
permeability arising from changes in membrane potential.
REFERENCES
COLE, K. S. (1941). Rectification and inductance in the squid giant axon. J. gen. Physiol. 26,
29-51.
COLE, K. S. & BAKEB, R. F. (1941). Longitudinal impedance of the squid giant axon. J. gsn.
Phy8iol. 24, 771-788.
COLE, K. S. & CuRTIs, H. J. (1939). Electric impedance of the squid giant axon during activity.
J. gen. Phy8iol. 22, 649-670.
GOLDMAN, D. E. (1943). Potential, impedance, and rectification in membranes. J. gen. Physiol.
27, 37-60.
HARTREE, D. R. (1932-3). A practical method for the numerical solution of differential equations.
Mem. Manchr lit. phil. Soc. 77, 91-107.
HODGKIN, A. L. (1951). The ionic basis of electrical activity in nerve and muscle. Biol. Rev. 26,
339-409.
HODGKN, A. L. & HUXLEY, A. F. (1952a). Currents carried by sodium and potassium ions through
the membrane of the giant axon of Loligo. J. Physiol. 116, 449-472.
HODGKIN, A. L. & HuXLEY, A. F. (1952b). The components of membrane conductance in the
giant axon of Loligo. J. Phy8iol. 1I6, 473-496.
HODGKN, A. L. & HUxLEY, A. F. (1952c). The dual effect of membrane potential on sodium
conductance in the giant axon of Loligo. J. Phys8ol. 116, 497-506.
HODGKIN, A. L., HUXLEY, A. F. & KATZ, B. (1949). Ionic currents underlying activity in the
giant axon of the squid. Arch. Sci. phy8iol. 3, 129-150.
HODGKIN, A. L., HUXLEY, A. F. & KATZ, B. (1952). Measurement of current-voltage relations in
the membrane of the giant axon of Loligo. J. Physiol. 116, 424-448.
HODGKIN, A. L. & KATZ, B. (1949). The effect of temperature on the electrical activity of the
giant axon of the squid. J. Phy8iol. 109, 240-249.
KEYNES, R. D. (1951). The ionic movements during nervous activity. J. Physiol. 114, 119-150.
KEYNES, R. D. & LEWIS, P. R. (1951). The sodium and potassium content of cephalopod nerve
fibres. J. Physiol. 114, 151-182.
LORENTE DE N6, R. (1947). A study of nerve physiology. Stud. RockefeUer In8t. med. Re8. 131,
132.
PUMPHREY, R. J., SCHMITT, 0. H. & YouNx, J. Z. (1940). Correlation of local excitability with
local physiological response in the giant axon of the squid (Loligo). J. Phy8iol. 98, 47-72.

The Journal of Physiology - 1952 - Hodgkin - A Quantitative Description of Membrane Current and Its Application To

Uploaded by

Copyright:

Available Formats

The Journal of Physiology - 1952 - Hodgkin - A Quantitative Description of Membrane Current and Its Application To

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Journal of Physiology - 1952 - Hodgkin - A Quantitative Description of Membrane Current and Its Application To

Uploaded by

Copyright:

Available Formats

14697793, 1952, 4, Downloaded from https://physoc.onlinelibrary.wiley.com/doi/10.1113/jphysiol.1952.sp004764 by Cochrane Peru, Wiley Online Library on [27/11/2022].

J. Physiol. (I952) I I7, 500-544

A QUANTITATIVE DESCRIPTION OF MEMBRANE

PART II. MATHEMATICAL DESCRIPTION OF MEMBRANE

If V is changed suddenly can and n instantly take up values appropriate to the

For the purpose of calculation we assume that n= 1 at the asymptote which

0X=001 (V + 10)exp 0 - (12)

PART III. RECONSTRUCTION OF NERVE BEHAVIOUR

/n3=0.125 exp (V/80), (13)

om=O-l (V+25)/(exp VZ2 -1), (20)

ia?14 '0t- 10"4*e

determines the conduction velocity in conjunction with the constants of the

Fig. 16A shows the membrane potential and conductance in a calculated

During this perio

palmotenatirelypfromnloal. circuits.aTe fatenthattheV)

be obtained by integrating the corresponding ionic currents over the whole

the sodium conductance can be raised by a depolarization, and the delayed

of 6 3° C. A spike resulted, and the time course of membrane potential is

Accommodation. No measurements of accommodation were made nor did we

140 E|Bi kQ1.cm2

frequency of oscillation. A fair degree of variability is to be expected in these

As pointed out by Cole (1941), the process underlying oscillations in

You might also like