The model envisages two components in the process, whereby Ht equivalents of co-substrate and St equivalents of substrate accumulate in the cellular compartment in time t. The first is the flow through the symport, n equivalents of co-substrate entering or leaving with each substrate equivalent. The second is the basal flow of co-substrate outside the symport. In certain specific circumstances n can be derived by plotting Ht/t against St/t. The principal requirement is that, whereas the ratio of the component flows must change in the interval t, the magnitude of the basal flow must either be zero or constant. The procedure is applied to published observations [West & Mitchell (1973) Biochem. J. 132, 587-592] on the lactose-proton symport of Escherichia coli [n = 1.075 +/- 0.064(7)] and to new observations on the isoleucine-Na+ symport of mouse ascites-tumour cells [n = 1.136 +/- 0.120(18)].

1. Various ways of computing the proton stoichiometry of glycine absorption were examined in relation to the problem of distinguishing the proton flow (i) through the symport from the basal proton flow (ii) outside it. By depolarizing the plasma membrane, i will tend to inhibit ii. 2. A series of 23 yeast (Saccharomyces carlsbergensis) preparations grown with proline or glutamate were used, some of which were starved in the presence of glucose. Consequently, after ATP depletion, the rate of glycine uptake from a 0.2 mM solution varied through the series from 3 to 14 nmol.min-1.mg-1. Basal proton uptake in the absence of glycine was fairly constant at 3-4 nmol.min-1.mg-1. 3. After addition of glycine, the number of extra equivalents of protons entering the yeast with each amino acid equivalent in 30 s was 0.5 at the lowest rate of glycine absorption and 1.8 equivalents at the fastest rate. However, total proton absorption in 30 s increased in direct proportion to the amount of glycine absorbed. The proportionality factor, indicative of the carrier stoichiometry, was 2.25 +/- 0.13 (23) S.E.M. The effective basal proton uptake was negligibly small. 4. Progress of proton and glycine absorption by each yeast preparation in the period up to 180 s fitted the mathematical model described in the preceding paper by Eddy, Hopkins & Johnson [(1988) Biochem. J. 251, 111-114]. The analysis led to two estimates of the constant ratio of the inflow of protons to the inflow of glycine that would apply when the basal proton flow vanished. These further estimates of the carrier stoichiometry were also near 2, being 2.07 +/- 0.24 (6) and 2.22 +/- 0.07 (17).

Strain N.C.Y.C. 193 of Candida utilis was grown aerobically at 30 degrees C with nitrate as limiting nutrient in a chemostat. The washed yeast cells depleted of ATP absorbed up to 5 nmol of nitrate/mg dry wt. of yeast. At pH 4-6, extra protons and nitrate entered the yeast cells together, in a ratio of about 2:1. Charge balance was maintained by an outflow of about 1 equiv. of K+. Nitrate stimulated the uptake of about 1 proton equivalent during glycolysis or aerobic energy metabolism. Studies with 3,3′-dipropylthiadicarbocyanine indicated that the proton-linked absorption of nitrate, amino acids or glucose depolarized the yeast cells. Proton uptake along with lactate led neither to net expulsion of K+ nor to membrane depolarization.

Mouse ascites-tumour cells oxidizing lactate, in a modified Ringer solution, concentrated 2-aminoisobutyrate, L-methionine or 2-(methylamino)isobutyrate about 20-fold from a 0.4 mM solution in the presence of 2-3 micrograms of nigericin/mg cellular dry wt. The ionophore increased cellular [Na+] to almost 100 mM when extracellular [Na+] was about 45 mM. Either valinomycin or the two mitochondrial inhibitors oligomycin and antimycin acting together each markedly lowered the extent to which the tumour cells concentrated amino acid, from the above factor of about 20 to roughly 2-fold. Ouabain (1 mM) had a similar effect, and further raised cellular [Na+]. The sodium pump appeared to be closely involved in amino acid uptake under these conditions.

1. The fluorescent dye 3,3′-dipropyloxadicarbocyanine was used to show that the tumour cells absorbed 2-aminoisobutyrate, glycine, L-leucine and L-isoleucine and certain other amino acids electrogenically. The Km values with respect to amino acid concentration ([A]o), obtained from the fluorescence assays, varied through the above series from 0.8 to 26 mM, with Vmax. fairly constant. 2. Similar Km values described the uptake of the 14C-labelled amino acids in five instances where this was measured. 3. Each amino acid lowered the membrane potential (E) by 10-20 mV when its cellular concentration ([A]i) had reached a steady value and [A]o was 10mM. In these experiments energy metabolism was maintained by glycolysis, 2,4-dinitrophenol was present and cellular respiration was inhibited. The corresponding net flow of amino acid through the Na+ symport was deduced by making use of the fact that the depolarization an amino acid initially caused was roughly proportional to the net influx of amino acid itself. 4. The steady-state depolarization was attributed to the presence of a leak pathway for the amino acid with a rate coefficient PA. As assayed in the absence of Na+, PA was about 5-fold larger for isoleucine than for glycine. 5. Direct estimates of Vmax./PA were similar to those inferred from the extent of depolarization in the steady state and [A]i. 6. A mathematical model was used to predict [A]i/[A]o in term of the measured values of [Na]o, [Na]i, E, Km and Vmax./PA. The predicted and observed values agreed fairly well when [A]o was 1 mM or 10 mM. 7. [A]i/[A]o varied from about 2.5 for 10 mM-isoleucine to 30 for 1 mM-2-aminoisobutyrate when delta microNa, expressed as a ratio, was ostensibly in the range 19-43. 8. The concentration of 2-aminoisobutyrate from a 0.1 mM solution in the presence or absence of ouabain was consistent with the model, whereas the concentration of isoleucine from a 0.1 mM solution exceeded the predicted values 2-5-fold. 9. The tumour cells concentrated 2-amino-bicyclo[2,2,1]heptane-2-carboxylic acid by a non-electrogenic mechanism, with which isoleucine may also interact.

1. The tumour cells were incubated at 37 degrees C in Ringer solutions containing glucose, 1 mM-methionine, various concentrations of Na+ and K+ and, in some instances, ouabain or valinomycin to lower the membrane potential generated by the Na+ pump. After about 30 min, when the system had reached a steady state, the ratio [extracellular Na+]/[cellular Na+] varied from about 0.6 to 3.2 with the ionic conditions. The membrane potential, determined by means of the fluorescent probe 3,3′-dipropyloxadicarbocyanine, also varied systemically from about zero to–55mV.2. the ratio [cellular methionine]/[extra-cellular methionine] varied from about 1 to 35 in these circumstances. The Na+ electrochemical gradient, measured in the same units, varied from about 1 to 30. Its magnitude in 46 assays was almost directly proportional to, though slightly smaller than, the methionine gradient. 3. A mathematical model was used to define the relation between these two gradients, which were not in equilibrium, owing to the presence of a leak pathway for the amino acid. On the assumption that the values of [cellular Na+] were correct, the methionine gradient was about 1.8 times larger than this version of the gradient hypothesis predicted.

1. The magnitude of the K+ gradient across the plasma membrane, which was in equilibrium with the membrane potential (E) of the tumour cells, was determined by the “null point” procedure of Hoffman & Laris (1974) [J. Physiol. (London) 239, 519–552] in which the fluorescence of the dye serves as an indicator of changes in the magnitude of E. 2. A mixture of oligomycin, 2,4-dinitrophenol and antimycin was used to stop the mitochondria from interfering with the fluorescence signal. Transport functions at the plasmalemma were maintained under these conditions in the presence of glucose. 3. Physiological circumstances were found in which incubation with glycine or with glucose changed the “null point” value of E within the range–20mV to–100mV. The fluorescence intensity at the “null point” was an approximately linear function of E over that range. The procedure enabled E to be inferred form the fluorescence intensity in circumstances where titration to the “null point” was not feasible. 4. The rapid depolarization caused by l-methionine or glycine was shown in this way to have a maximum amplitude of about 60mV. A mathematical model of this process was devised. 5. The electrogenic Na+ pump hyperpolarized the cells up to about −80mV when the cellular and extracellular concentrations of K+ were roughly equal. 6. The observations show that the factors generating the membrane potential represent a major source of energy available for the transport of amino acids in this system.

1. Pigeon erythrocytes, resealed lysed erythrocytes or liposomes derived from erythrocyte lipids were suspended in solutions containing up to 2 micrometer-3,3′-dipropyloxadicarbocyanine iodide. Gramicidin, valinomycin, nigericin or carbonyl cyanide p-trifluoromethoxy-phenylhydrazone, or combinations of these, were used to induce electrical diffusion potentials dependent on Na+, K+ or protons. In each instance hyperpolarization of the cell membrane lowered the fluorescence of the cell suspension, a process that was completed in about 1 min. Subsequent depolarization caused an increase in fluorescence. 2. Quenching of the fluorescence of the cell suspension appeared to be due to the reversible binding of the dye to the cells. Much larger amounts of dye were bound, both to the intact and to the resealed erythrocytes, than would be expected if partitioning of the dye cation followed the Nernst equation. The dependence of the binding on the extracellular dye concentration was studied in the presence and absence of valinomycin. The results were consistent with the suggestion of Sims, Waggoner, Wang & Hoffman [(1974) Biochemistry 13, 3315-3330] that the dye was bound at both membrane surfaces and that, at low dye concentrations, hyperpolarizing the cells promoted dye binding at the inner membrane surface. 3. The applications of the technique are limited by the circumstance that the direct effect of the electric field on the uptake of the dye into the cells is amplified by a binding process that may be affected by other physiological variables.

1. When yeast N.C.Y.C. 240 was grown with maltose in a complex medium based on yeast extract and peptone, washed cell preparations fermented alpha-methyl glucoside much more slowly than maltose. 2. The yeast absorbed alpha-methyl[14C]glucoside from a 10mM solution in the presence of antimycin and iodoacetamide, producing [14C]glucose, which accumulated outside the cells. The yeast itself contained hexose phosphates, trehalose, alpha-methyl glucoside and other products labelled with 14C, but no alpha-methyl glucoside phosphate. 3. About 1 equiv. of protons was absorbed with each equivalent of alpha-methylglucoside, and 1 equiv. of K+ ions left the yeast. 4. alpha-Thioethyl glucoside was also absorbed along with protons. Studies by g.l.c. showed that the yeast concentrated the compound without metabolizing it. 5. The presence of trehalose, sucrose, maltose, L-sorbose, glucose or alpha-phenyl glucoside in each case immediately stimulated proton uptake, whereas fructose, 3-O-methylglucose and 2-deoxyglucose failed to do so. 6. The observations support the conclusion that alpha-thioethyl glucoside, alpha-methyl glucoside and maltose are substrates of one or more proton symports, whereas they seem inconsistent with the notion that the absorption of alpha-methyl glucoside involves the phosphorylation of the carbohydrate [Van Stevenick (1970) Biochim. Biophys. Acta 203, 376-384].

1. At pH 4.5 and 30degreesC, yeast preparations depleted of ATP in the presence of antimycin and deoxyglucose spontaneously lost K + , gaining roughly an equivalent amount of H + . 2. Five proton conductors including azide and 2,4-dinitrophenol accelerated this process, as did [ 14 C]glycine, which was absorbed with two extra equivalents of H + . 3. The rate of glycine uptake at pH 4.5 diminished fourfold when cellular K + fell by 20%. 4. The distribution of [ 14 C]propionate indicated that the intracellular pH fell from 6.2 to 5.7 when the cellular content of K + fell by 30%. 5. Glycine uptake from a 5 muM solution was about 400 times faster at pH 4.5 than it was at pH 7.4 with 100mM-KCl present ostensibly to lower the membrane potential. 6. Yeast preparations containing 2mM-[ 14 C]glycine absorbed a further amount from a 0.1 muM solution at pH 4.5. After about 10 min a net movement of [ 14 C]glycine out of the yeast occurred. The ratio of the cellular [14Ia1glycine concentration to the concentration outside the yeast reached 4 × 10 4 in these assays, whereas at pH 7.4 in the presence of 100mM-KCl it did not exceed 15 in 3h. Dimitrophenol lowered the accumulation ratio at pH 4.5, apparently by causing proton conduction. 7. The observations are consistent with the notion that glycine uptake is driven by a proton symport mechanism. 8. Possible factors governing the strikingly low rate of glycine efflux as opposed to its optimum rate of influx are discussed.

1. A study was made of the pH changes occurring when 0.1-4 mumol of glutamate, phosphate and certain phosphate esters was added at about pH 4.8 to washed cell preparations (50 mg dry wt.) of strains of Saccharomyces. The system also contained deoxyglucose and antimycin to inhibit energy metabolism and so prevent proton ejection from the yeast. 2. A strain of Sacc. carlsbergensis was grown in a chemostat with a limiting supply of phosphate in order to enhance the subsequent rate of phosphate transfer into the yeast. These preparations absorbed 0.2 mumol of phosphate with about 3 equiv. of protons/mol of phosphate. The charge balance was maintained by the efflux of 2 equiv. of K-+ from the yeast. 3. Larger amounts of phosphate were absorbed with fewer proton equivalents. 4. Arsenate and phosphate caused similar pH changes. 5. Glucose 6-phosphate, ATP and certain order phosphate esters each initiated a rise in pH, possibly because hydrolytic extracellular enzymes released phosphate that was subsequently absorbed. 6. Four strains of yeast were grown with glutamate as principal source of nitrogen. Each absorbed extra protons in the presence of L-glutamate. 7. One of them, a strain of Sacc. cerevisiae, absorbed 0.2 mumol of glutamate with 3equiv. of protons/mol of glutamate, and in these circumstances 1-2 equiv. of K-+ left the yeast cells. 8. The role of ionic gradients in the transport of these anions is discussed.

1. Preparations of mouse ascites-tumour cells depleted of ATP and Na + ions accumulated l -methionine, in the presence of cyanide and deoxyglucose, from a 1m m solution containing 80mequiv. of Na + /l and about 5mequiv. of K + /l. Valinomycin increased, from about 4 to 16, the maximum value of the ratio of the cellular to extracellular concentrations of methionine formed under these conditions without markedly affecting the distributions of Na + and of K + . Similar observations were made with 2-aminoisobutyrate, glycine and l -leucine. Increasing the extracellular concentration of K + progressively decreased the accumulation of methionine in the presence of valinomycin. Over the physiological range of ionic gradients, the system behaved as though the absorption of methionine with Na + was closely coupled to the electrogenic efflux of K + through the ionophore. The process was insensitive to ouabain and so the sodium pump was probably not involved. 2. The amount of methionine accumulated during energy metabolism was similar to the optimal accumulation in the presence of valinomycin when ATP was lacking. It was also similarly affected by increasing the methionine concentration. 3. A mixture of nigericin and tetrachlorosalicylanilide mimicked the action of valinomycin. The anilide derivative inhibited the absorption of 2-aminoisobutyrate in the presence of valinomycin, but not in its absence. 4. Gramicidin inhibited methionine absorption and caused the preparations to absorb Na + and lose K + . 5. The observations appear to verify the principle underlying the gradient hypothesis by showing that the tumour cells can efficiently couple the electrochemical gradient of Na + to the amino acid gradient.

1. Proton uptake in the presence of various amino acids was studied in washed yeast suspensions containing deoxyglucose and antimycin to inhibit energy metabolism. A series of mutant strains of Saccharomyces cerevisiae with defective amino acid permeases was used. The fast absorption of glycine, l -citrulline and l -methionine through the general amino acid permease was associated with the uptake of about 2 extra equivalents of protons per mol of amino acid absorbed, whereas the slower absorption of l -methionine, l -proline and, possibly, l -arginine through their specific permeases was associated with about 1 proton equivalent. l -Canavanine and l -lysine were also absorbed with 1–2 equivalents of protons. 2. A strain of Saccharomyces carlsbergensis behaved similarly with these amino acids. 3. Preparations of the latter yeast grown with maltose subsequently absorbed it with 2–3 equivalents of protons. The accelerated rate of proton uptake increased up to a maximum value with the maltose concentration ( K m =1.6m m ). The uptake of protons was also faster in the presence of α-methylglucoside and sucrose, but not in the presence of glucose, galactose or 2-deoxyglucose. All of these compounds except the last could cause acid formation. The uptake of protons induced by maltose, α-methylglucoside and sucrose was not observed when the yeast was grown with glucose, although acid was then formed both from sucrose and glucose. 4. A strain of Saccharomyces fragilis that both fermented and formed acid from lactose absorbed extra protons in the presence of lactose. 5. The observations show that protons were co-substrates in the systems transporting the amino acids and certain of the carbohydrates.

1. Despite the depletion of both their content of exchangeable endogenous amino acids and reserves of ATP, starved hypo-osmotically shocked preparations of the tumour cells accumulated relatively large amounts of 14 C-labelled 2-aminoisobutyrate, l -alanine, glycine, l -leucine, l -methionine, l -phenylalanine and l -serine, against their respective concentration gradients, by a process apparently driven by the spontaneous flow of Na + ions into the cellular phase. Dependent on ( a ) which compound was used, ( b ) its concentration and ( c ) the direction of the Na + ion gradient, the peak value of the ratio of the cellular to extracellular amino acid concentration varied from about 0.4 to 7. 2. The extent to which ATP increased the ratio was defined for l -methionine. 3. Chemical analysis of the cellular amino acid content showed that this increased in parallel with the absorption of 14 C. 4. The accumulation of l -methionine and of glycine, against their own concentration gradients, continued in the presence of either 0.3m m -ouabain or 10μg of oligomycin/ml. Thus the sodium pump was probably not involved in the process when ATP was lacking. 5. l -Leucine caused 0.72±0.12 ( s.e.m. ; 6) extra equivalents of Na + to enter the shocked starved tumour cells in parallel with the uptake of leucine itself. Only a small loss of K + was induced. 6. The influx and efflux of l -methionine in preparations depleted of ATP were both markedly accelerated by the presence of Na + ions. 7. The observations provide further examples of the application of the ion-gradient hypothesis, according to which Na + ions act as co-substrates of the amino acid pump. The quantitative importance of parallel Na + -independent systems was studied with a new mathematical model.

1. Proton uptake into the yeast Saccharomyces carlsbergensis , was studied at pH4.5–5.5 in the presence of both antimycin and 2-deoxyglucose to inhibit energy metabolism. Previous work had shown that the cells then absorbed about 20nmol of glycine or l -phenylalanine against a considerable amino acid concentration gradient. The addition of the amino acid immediately stimulated the rate of uptake of protons two- to three-fold. About 2 extra equivalents of H + accompanied a given amount of the amino acids into the yeast preparations exposed to the metabolic inhibitors for 2–4min and about 1.2 equivalents after 20min exposure. 2. Analogous observations were made during serial additions of glycine, l -phenylalanine, l -leucine and l -lysine to preparations lacking the metabolic inhibitors and deficient in substrates needed for energy metabolism. In fresh cellular preparations the influx of glycine was then closely coupled to a stimulated flow of 2.1 equiv. of H + into the yeast. A similar number of K + ions left the cells. About 30% of the extra protons was subsequently ejected from the yeast. Deoxyglucose and antimycin together inhibited the ejection of protons. When the yeast had been fed with glucose energy metabolism was stimulated and almost as many protons as were absorbed with the amino acid were apparently ejected again. 3. Yeast preparations containing Na + , instead of K + , as the principal cation absorbed about 1 extra equivalent of H + after the addition of phenylalanine, glycine or leucine. This response was not observed in the presence of both deoxyglucose and antimycin. 4. The observations show that H + and, in certain circumstances, K + are co-substrates in the transport of the amino acids into the yeast. An analogy is drawn with the roles of Na + and K + as co-substrates in certain mammalian systems. The results lead to various models relating the physical flow of the co-substrate ions on the amino acid carrier to the transduction of chemical energy in an associated ion pump forming part of the mechanism for transporting amino acids into the yeast.

A study has been made of the effects of both varying the pH and extracellular [K + ] on the initial rate of uptake of glycine ( v ) by a strain of Saccharomyces carlsbergensis that concentrated the amino acid, with respect to the extracellular phase, by up to 1400 times. When no other substrate than glycine was provided and [glycine] was relatively small (≤0.2m m ) (1) v increased fivefold when the pH was lowered from 7 to 4.5; (2) v fell by up to about 80% as [K + ] rose, K + behaving as a non-competitive inhibitor of the system, with K i 0.33mequiv./l at pH7; (3) the absorption of glycine caused up to about 2 or 3 equiv. of K + to leave the yeast cells. These three phenomena were each less evident when glucose was present. An analogy is drawn between the respective interactions of H + and K + with the yeast system and the well recognized effects of Na + and K + on amino acid transport in certain mammalian systems.

1. The ATP content of preparations of a strain of Saccharomyces carlsbergensis was lowered below 0.3nmol/mg of yeast by starving the yeast cells in the presence of both antimycin and 5m m -deoxyglucose. 2. When the depleted cells were put at pH4.5 with glycine up to about 20nmol of the amino acid/mg of yeast was absorbed without being chemically modified. The mechanism did not depend on an exchange with endogenous amino acids. 3. The concentration of the absorbed glycine could apparently reach 100–200 times that outside the cells. 4. Replacement of the cellular K + by Na + almost stopped amino acid absorption in the presence of antimycin and deoxyglucose, but not in their absence. 5. It is suggested that, when energy metabolism itself had stopped, a purely physical process, namely the movements of H + and K + into and out of the yeast respectively, served to concentrate the amino acids in the cells. Both ionic species appear to be co-substrates of the system transporting amino acids.

1. To deplete them of ATP the tumour cells were starved at 37° in a Ringer solution containing 33m-equiv. of Na + /l., 131m-equiv. of Li + /l., 2mM-sodium cyanide and 0·1m m -ouabain. The cellular content of K + was largely replaced by Li + , but cellular [Na + ] remained near 33m-equiv./l. 2. The addition of 12m m -glycine to the system caused cellular [Na + ] to increase, during the next 4min., by about 4m-equiv./l., so that it slightly exceeded extracellular [Na + ]. This occurred in parallel with the absorption of glycine. 3. The cellular K + content fell by an amount representing about 10% of the amount of Na + absorbed. 4. The results provide a clear demonstration that the flow of glycine into the cells is linked to a parallel movement of Na + ; K + appears to play a facultative role in the carrier system, whereas Li + is almost inert. 5. The effects produced by glycine were not reproduced by l -arabinose.

1. The initial rate of uptake of glycine by the tumour cells was measured as a function of the Na + and K + concentrations in the solution in which the cells were suspended. When [Gly] was 1m m or 12m m , the rate in the absence of Na + was independent of [K + ] and about 3% or 10% respectively of the rate when [Na + ] was 150m-equiv./l. 2. The Na + -dependent glycine entry rate, v , at a given value of [Na + ] was successively lowered when [K + ] was increased from 8 to 47 to 96m-equiv./l. A kinetic analysis indicated that K + competitively inhibited the action of Na + . The results were in fair agreement with previous determinations of the kinetic parameters. 3. The presence of 2m m -sodium cyanide and 10m m -2-deoxyglucose lowered the cellular ATP content to less than 3% of the value in the respiring cells. Although v was then about 50% smaller, the relative effects of K + and Na + on the system were similar to those observed during respiration. 4. A theoretical analysis indicated that the variation of v with [K + ] is not a reliable guide to the extent to which the K + gradient between the cells and their environment may contribute to the net transport of glycine.

1. Tumour cells were starved to deplete them of ATP and transferred to 0·9m m -glycine in Ringer solutions containing 2m m -sodium cyanide and various Na + and K + concentrations. The uptake of glycine then usually reached a peak by about 10min. 2. When cellular [Na + ] and extracellular [Na + ] were each about 30m-equiv./l., the maximum amount of glycine absorbed increased between 1·2- and 3·0-fold on lowering extracellular [K + ] from 128 to 10m-equiv./l. 3. When extracellular [Na + ] was 150m-equiv./l., the ratio, R , of the cellular to extracellular glycine concentrations increased progressively, from near 1 to about 9, when cellular [Na + ] was lowered from 120 to 40m-equiv./l. 4. When cellular [Na + ] was almost constant, either at 45 or 70m-equiv./l., R fell about 14-fold when extracellular [Na + ] varied from 150 to 16m-equiv./l. 5. Values of R near 0·2 were found when cellular [Na + ] was about four times as large as extracellular [Na + ]. 6. R fell about threefold when the cells were put with 12m m - instead of 0·9m m -glycine. 7. The results were taken to imply that, under these conditions, the spontaneous movements of both Na + and K + across the cell membrane, down their respective concentration gradients, served to concentrate the glycine in the tumour cells (Christensen's hypothesis).

1. The tumour cells were starved in a solution lacking Na + and then transferred to a Ringer solution containing 2m m -sodium cyanide, 150m-equiv. of Na + /l. and 10m-equiv. of K + /l. Such cells were depleted of ATP and contained an endogenous pool of various amino acids equivalent to a 26m m solution. 2. At 4min. after the transfer the cellular Na + content had increased by about 100% and roughly an equivalent amount of K + had left the cells. 3. Under these conditions [ 14 C]glycine was absorbed from an 11m m solution and reached the same cellular concentration by about 4min. The pool size increased by approximately the same amount (ΔGly), so glycine did not simply exchange with the endogenous components. 4. After 4min. with glycine, the cells contained about 20% more Na + (ΔNa + ) than the control and about 10% less K + (ΔK + ). The mean values of ΔNa + /ΔGly and ΔK + /ΔGly from five experiments were respectively 0·90±0·11 and 0·62±0·11equiv./mole. 5. A further indication that these two ratios were not equal was that the cells absorbed more water than the movement of glycine itself required. The excess of water was osmotically equivalent to 0·95±0·16equiv. of solute/mole of glycine absorbed. 6. The variation of ΔNa + /ΔGly with the duration of the incubation was consistent with the stimulated uptake of Na + being linked to the actual transport of glycine. The same may apply to the movement of K + , though the time-dependence was not examined in that case. 7. The observations were analysed in terms of a model in which both K + and Na + moved with a glycine-carrier system without ATP being involved. The analysis supported the idea that the spontaneous movements of the ions through the system might concentrate glycine in the cells significantly by purely physical means (Christensen's hypothesis).