Frog skin has been used by many workers in the study of bioelectric phenomena of living membranes. The isolated living skin mounted with Ringer’s solution on both sides maintains a potential of from 10 to 100 mV., the inside of the skin being positive. This potential can be modified by changing the chemical or physical conditions. Many workers have attempted to interpret the results of such changes in terms of simple membrane potentials but without much success. Dean & Gatty (1937) point out some of the difficulties of such interpretation and review the literature. There is evidence of at least two sources of potential in the skin which may at times be opposed (Steinbach, 1933), and also that the active surfaces are not uniform but present a mosaic structure (see also Dean, 1938).

While it is not possible to determine actual mobilities solely from changes in potential following ionic changes in the medium, we can interpret such changes in terms of relative mobilities. If, for example, Cl is replaced in the outside solution by an anion that has a higher mobility through some part of the skin, this ion will tend to diffuse inwards more rapidly than Cl diffuses outwards. This makes the outside solution less negative and decreases the skin potential. The magnitude of the change will depend not only on the charge and relative mobilities of the ions through the layers causing the potential, but also on the thickness, extent and the chemical nature of these layers as well as on the magnitude of the concentration gradients across them. It is not valid to assume that the entire concentration gradient across the skin is operative across that part of it that has selective ion permeability. Therefore, we can only deduce an averaged apparent ionic mobility and the absolute values for the mobilities of the ions as calculated by a Henderson equation have little meaning. We can say that one ion appears to have a greater or smaller mobility than another.

Some of the changes in potential, particularly those caused by poisons, are obviously due to metabolic changes in the skin. There is good reason, however, to regard most if not all of the effects reported in this paper as due to differences in ionic mobilities or the rate of entry of ions into a superficial layer of the skin.

Some of the work reported here has been done before by other workers and with similar results; although it so happens that much of the previous work was done without proper control of the various factors involved. In so far as is possible, these experiments have been done with solutions that differ in only one component from Ringer, pH, minor ions, and osmotic pressure (though not always ionic strength)1 being kept constant.

Potential was measured in cells similar to those described by Francis & Pumphrey (1933). No attempt was made to stretch the skin in the cells. The outlet syphons dipped into beakers of Ringer (or in later work on dilute solutions into strong NaCl) which received the overflow when solutions were changed. Saturated KCl-calomel electrodes also dipped into these beakers. Contact could be made between the two beakers by sucking the solutions into an inverted Y tube to check zero potentials in the measuring circuit. These were never more than 5 mV. and were usually corrected for. In some experiments it was convenient to insert a bias of 100 mV. by means of an extra storage battery and variable high resistance to facilitate reading of negative potentials.

Potentials were measured on a standard potentiometer reading to fifths of a millivolt using an Einthoven string galvanometer as a null point instrument in series with a resistance of 1 megohm. With this apparatus readings to 1 mV. could be made within 5 sec. after closing the key.

Five or six comparisons were made in all cases, comparing the test solution on one half of the abdominal skin of Rana temporaria with a control solution on the other half. The differences between test and control were analysed statistically, using Student’s t test (Fisher, 1925). A probability of chance difference of 0·05 was taken as significant and no results are reported unless they fulfil this requirement for significance.

The skins were mounted with Ringer on both sides and left for an hour or two to settle down. Even after several hours some skins were still very variable, and without proper controls would have yielded misleading results. Solutions were changed at frequent but irregular intervals, rarely more than 1 hr. in length when potentials were read. Controls received as nearly as possible the same treatment as the test. When solutions were changed by emptying and refilling test cells the controls were also emptied. Ordinarily, only frogs of the same sex were used in any one set of comparisons which were all tested together. Fen frogs recently caught and kept in tanks with fresh tap water were used throughout.

Resistance was measured by the method of Pumphrey (1934) in the same cells. The inner end of the inlet tube was shortened to 6 mm. from the skin. Silver electrodes were inserted in the inlet tubes and connected in series with a resistance of 10 megohms and a high tension battery. In this way a constant current of 10 μA. could be applied across the skin (average area one square centimetre) and the change in potential measured. The potential was measured before and again as soon as possible after closing the circuit. With such low current densities the potential drift is slow and the initial quick change was taken as proportional to the ohmic resistance of the skin. Calculations show that with a high resistance skin the deviations from parallel lines of current flow are less than 5 % and less than variations within the skin itself.

The experimental solutions were made up from stock solutions prepared by weight from B.D.H. chemicals without further purification. Frog Ringer containing no glucose was made according to the formula given by Bayliss (1927). The solutions were made to 1 1. instead of with 1000 g. of water and 0001 M Na2HPO4 was used instead of 0·0008M NaH2PO4. pH was always adjusted with HC1 or NaOH (or appropriate agents for substituted Ringer) using cresol red or phenol red and a colour chart. Osmotic deficiencies of polyvalent ions were made up with mannitol (C6H14O6) keeping the Na+ or Cl concentrations constant. In some cases all the Na+ or Cl was replaced by other ions, but for most of the experiments a solution containing only the minor salts of Ringer (KC1, CaCl2, NaHCO3 and Na2HPO4 which contains 0·0024M Na+ and 0·0030 M Cl was used as a base and 0·111M quantities of NaNO3, KC1, etc., were added to it before final dilution. Unless otherwise stated Q Ringer refers to a solution containing 0·111 mol. of Q ions as Q Cl or NaQ substituted for Na+ or Cl and other ions as above.

The normal potential of living skins mounted with Ringer on both sides can be maintained for 12 hr. with a current flow of 20−30 × 10−6 amp. (Francis, 1933). This current is sufficient to depolarize in less than one-fifth of a second any potential due to an oriented layer that is not being continuously reformed (Hubbard, Gatty & Rothschild, unpublished). The potential can, however, be maintained by the outward diffusion of two ions with different mobilities that are produced within the skin. Respiration seems to be the most probable method of forming these ions and many workers have shown the correlation between O2 uptake and potential. See Francis & Gatty (1938) for a summary of results.

If H+ and HCO3 are the respiratory ions and if they diffuse out of the skin through the outer surface, HCO3 must have a high diffusion mobility relative to H+ to give the correct sign to the observed potential and it must also have a higher diffusion mobility than Cl in order to carry an appreciable fraction of the current. Therefore high concentrations of HCO3 outside should decrease the potential. Fig. 1 shows the effect of substituting HCO3 Ringer for Cl Ringer. Each point represents a potential reading and a renewal of the solution. HCO3 Ringer outside lowers the potential to a negative value but it becomes positive again when HCO3 Ringer is placed on both sides. Experiments with lower concentrations of HCO3 confirm the above result. Low pH kills the skin irreversibly, making undoubted changes in the properties of the skin so that it is impossible to measure the relative mobility of H+ ions in the living skin by this method.

Fig. 1.

Line of circles, HCO3 Ringer inside at A; Ringer both sides again at C. Line of crosses, HCO3 Ringer outside at A, HCO3 Ringer both sides at B, return to Ringer both sides at C.

Fig. 1.

Line of circles, HCO3 Ringer inside at A; Ringer both sides again at C. Line of crosses, HCO3 Ringer outside at A, HCO3 Ringer both sides at B, return to Ringer both sides at C.

Calculations based on the data of Francis (1933) show that sufficient CO2 is formed to account for the current if the process is only 2% efficient. Francis & Gatty (1938) show however that the potential is more quickly killed in acrylate plus iodoacetate than in iodoacetate, yet the oxygen consumption and presumably the CO2 production is carried on at a high rate for 5 hr. in the former solution. There is little doubt that the respiratory ions are not as simple as suggested above. HCO3 Ringer increases the resistance whereas a decrease is expected if HCO3 acts only by virtue of its superior mobility. It is probable that one of the actions of HCO3 Ringer is to alter the skin surface, perhaps as a protection against unfavourable conditions.

When another anion, for example NO3, is substituted for Cl in the outside solution the potential becomes more positive but the skins die more quickly. Fig. 2 shows the effect of Cl free NOa Ringer on both sides for 21 hr. The effect is due to the Cl free solutions outside. With some exceptions noted below such solutions have no effect on the inside. The rise in potential is often much less than in Fig. 2, especially if the skins originally had a high potential. It is only reversible in the early stages. This change in potential can be considered to be caused by the greater mobility of Cl than NO3 through some part of the skin. All other anions tried except HCO3 gave the same effect. 0·05 M acetate substituted for part of the Cl in Ringer gave no effect nor did 0·0025 M succinate, but 0·05 M NO3 and·0·0025 M SO4 produce nearly as large an effect as Cl free solutions. That this effect is due to the lower mobility of inorganic ions is further corroborated by the higher potentials in NO3 than in Ac and in SO4− than in Sue−. The experiments show the following series for apparent relative effective mobilities of anions in Na+ Ringer :
Fig. 2.

Effect of NO3Ringer on both sides.

Fig. 2.

Effect of NO3Ringer on both sides.

Other organic and inorganic ions are grouped by analogy, no other direct comparisons having been made between members of the groups.

Citrate, Oxalate, F and HP04 Ringer all cause a decrease in potential when applied inside. With Cit this effect is strong enough to mask the Cl free effect when the solution is on both sides. The potential may become negative and the effect is reversible. F and Ox are irreversible as might be expected. The esults suggest a fixation or removal of some Ca++ maintained layer but further evidence is necessary. Ca++ free Ringer has practically no effect on the potential.

K+, Cs+, Rb+, NH4+, Mg++, or Ca++ Ringer on both sides reduced the potential, which fell to zero with K+ Ringer (see p. 136), or to about + 3 mV. with the other ions. Fig. 3 shows the effect of K+ Ringer on either side and also on both sides. When K+ Ringer is on one side only the potential is decreased but not to such low values. The peculiar shape of the curve on replacing Na+ Ringer on both sides is seen to be due to the composite effects on the two sides. Steinbach (1933) finds no effect of Ca++ solutions on the inside, but the writer found the effect of Ca++ Ringer to be very similar to K+ Ringer on either side.

Fig. 3.

Line of crosses, K+ Ringer inside at A; Ringer both sides again at B. Line of triangles (comparison with line of crosses), K+ Ringer outside at A, Ringer both sides again at B. Line of circles (separate experiment), K+ Ringer both sides at A, Ringer both sides again at B.

Fig. 3.

Line of crosses, K+ Ringer inside at A; Ringer both sides again at B. Line of triangles (comparison with line of crosses), K+ Ringer outside at A, Ringer both sides again at B. Line of circles (separate experiment), K+ Ringer both sides at A, Ringer both sides again at B.

The low potential in potassium solutions reported by nearly every worker on frog skin is accompanied by a large drop in the resistance noted by Pumphrey (1934). If this change in resistance were due to the high mobility of K+ ion, as has been assumed for frog skin and other tissues, then there would be a large K+ membrane potential when K+ Ringer is on one side only. This is not the case as can be seen from the figure. It seems more likely that K+ increases the mobilities of other ions and the following experiments support this theory.

When Cl is substituted by other anions in K+ Ringer the relative mobilities are in quite a different order. Fig. 4 shows the effect of KNO3 Ringer on either side, with KC1 Ringer on the other side, after KC1 Ringer has been on both sides, also of KHCO3 and KAc Ringers. NO3 now lowers the potential from the outside and increases it from the inside. Br and CNS have a similar effect. I has practically no effect at all. HCO3 increases the potential when outside and decreases it inside. Ac and all other anions tested behave as in Na+ Ringer. The following series is obtained (compare with p. 137):
Fig. 4.

Three experiments all starting with K+ Ringer both sides.

Fig. 4.

Three experiments all starting with K+ Ringer both sides.

It is evident that K+ has greatly altered the relative mobilities of the anions. (Note. The effect with KHCO3 Ringer suggests that K+ Ringer also destroys the respiratory potential if this is caused by HCO3 ions.)

When dilute KC1 is put outside and K+ Ringer inside most skins develop a positive potential without an abnormally high resistance. This effect suggests that under these conditions Cl has a high mobility relative to K+. However these positive potentials which some experimenters failed to get (Amberson & Klein, 1928) are easily reversed by 0·002M KMnO4, ether or 0·1 % saponin outside or 0·002 M CN inside and may not be indicative of more normal conditions. There is however sufficient evidence to say that K+ solutions alter ionic mobilities. They probably increase Cl mobility through some part of the skin since a major ion must be involved to account for the low resistance.

When the outside solution is greatly diluted the potential decreases and is negative in Na+ free dilute solutions. It is impractical to measure potentials with pure water on one side because of the high resistance, and even a small diffusion of salts out of the skin would render the results meaningless. With 0·00194M KC1 (equal K+ concentration to Ringer) outside the potential is about–30 mV.

Fig. 5 shows the effect of placing 0·0019 M KC1 outside compared with 0·0019 M KC1 +0·0004 M NaCl on the other halves of the skin. The negative potential is sensitive to small changes in Na+ concentration but not to anions nor other cations except Li+ which can partially substitute for Na+. The effect indicates a higher relative mobility of Na+ and Li+ than Cl and other ions in dilute solutions.

Fig. 5.

Effect of NaCl on potentials in dilute solutions.

Fig. 5.

Effect of NaCl on potentials in dilute solutions.

Fig. 6 shows the effect of replacing dilute KC1 outside with 0·111M NaCl + 0·0019 M KC1 and compared with the effect of 0·1130M KC1. Similar “surges” which build up in one or two seconds to +100 mV. or more and die out in a few minutes to normal Ringer-Ringer potentials have been obtained with Ringer, Li+ Ringer and to a smaller degree with 0·056M FeCl3 (pH 1−2?). No “surge” potentials were obtained with K+, Cs+, Rb+, NH4+, Mg++, Sr++, Hg++, Ca++, Ce+++ or HC1 in Ringer.

Fig. 6.

Both lines, 0·0019M KC1 outside at A.

Fig. 6.

Both lines, 0·0019M KC1 outside at A.

The dilute solutions seem to wash out Na+ from a thin outer cation permeable layer of the skin. Then when Na+ is replaced a high concentration gradient across this layer is produced for a short time resulting in the surge. During the steady state with a dilute solution outside the gradient is across the whole skin though only a thin layer of it may be selectively permeable to cations and capable of producing a membrane potential. The “surge” results confirm the supposition that Na+ and Li+ have higher mobilities than all other ions through this layer. Similar “surges” of smaller magnitude are obtained when Na+ Ringer replaces K+ Ringer or other cation substituted Ringer on the outside.

With dilute solutions outside the resistance of the skin, after correction for cell resistance, is increased tenfold. This may be due to the formation of a membrane relatively impermeable to salts and would be of advantage to the frog since a high resistance membrane would restrain salt diffusion and loss. It is not safe therefore to generalize about relative mobilities of Na+ and Cl from dilute solution effects as the skin surface has obviously been changed.

  1. Bicarbonate substituted for chloride in Ringer affects the potential of frog skin as if the bicarbonate ion were more permeable than chloride ion.

  2. Bicarbonate ion behaves as would be expected if it were one of the respiratory ions maintaining the normal potential of frog skin with Ringer on both sides.

  3. Chloride ion shows a higher apparent mobility than all other ions through the outside of the skin in sodium Ringer.

  4. Chloride ion also shows a higher apparent mobility than calcium removing anions through the inside of the skin.

  5. Potassium changes the order of apparent relative mobility of anions and increases the apparent mobility of chloride ion.

  6. With dilute solutions outside sodium has a higher apparent mobility than all other ions but can be partially substituted for by lithium.

The author wishes to thank Prof. J. Gray and O. Gatty for their helpful advice and interest throughout this work.

Amberson
,
W. R.
&
Klein
,
H.
(
1938
).
J. gen. Physiol
.
11
,
823
.
Bayliss
,
W. M.
(
1927
).
Principles of General Physiology
, 4th ed. p.
211
.
London
:
Longmans, Green and Co
.
Dean
,
R. B.
(
1938
).
Thesis; Cambridge University
.
Dean
,
R. B.
&
Gatty
,
O.
(
1937
).
Trans. Faraday Soc
.
33
,
1040
.
Fisher
,
R. A.
(
1925
).
Statistical Methods for Research Workers
, p.
106
.
London
:
Oliver and Boyd
.
Francis
,
W. L.
(
1933
).
Nature, Land
.,
131
,
805
.
Francis
,
W. L.
&
Gatty
,
O.
(
1938
).
J. exp. Biol
.
15
,
132
.
Francis
,
W. L.
&
Pumphrey
,
R. J.
(
1933
).
J. exp. Biol
.
10
,
379
.
Pumphrey
,
R. J.
(
1934
).
J-exp. Biol
.
11
,
423
and 429.
Steinbach
,
H. B.
(
1933
).
J. cell. comp. Physiol
.
3
,
1
.
1

Ionic strength cannot be kept constant if osmotic pressure and the concentrations of all other ions are kept constant when a polyvalent ion is substituted for a monovalent ion.