Gustatory sensitivities of the hamster's soft palate

The response properties of taste receptors distributed on thesoft palate of the hamster were studied by recording integratedresponses from the greater superficial petrosal (GSP) nerveStimuli were concentration series of sucrose, NaCl, HCl andquinine hydrochloride (QHCl), and several other 0.1 M saltsand 0.5 M sugars. For comparison, integrated responses wererecorded from the chorda tympani (CT) nerve in many of the sameanimals from which recordings were made from the GSP. Responsesin each preparation were scaled relative to the phasic responseto 0.1 M NaCl and were then expressed for each nerve as a proportionof the total response magnitude (TRM)—the sum of all theresponses to the four concentration series. In this way, therelative response of each nerve to all of the stimuli couldbe evaluated. There were significant differences between theGSP and CT nerves in the responses to NaCl, QHCl and sucrose.Both the phasic and tonic responses to sucrose were larger inthe GSP than in the CT, whereas the tonic responses to NaCland QHCl were smaller. The slopes of the concentration-responsefunctions for NaCl, HCl and sucrose were significantly differentbetween the two nerves. The responses to 0.1 M sodium and lithiumsalts were significantly greater in the CT than in the GSP;whereas the 0.5 M sugars elicited responses in the GSP thatwere 2–3 times greater than in the CT nerve. A comparisonof the relative responsiveness to 0.3M sucrose, 0 3 M NaCl,0.01 M QHCl, 0.01 M HCl and distilled water among the GSP, CT,glossopharyngeal (IXth) nerve and superior laryngeal nerve (SLN)indicated that the vast majority of information about sucroseand NaCl is transmitted to the brainstem by the VIIth nerve. ¹Present address: Department of Oral Physiology, Kagoshima UniversityDental School, Kagoshima 890, Japan     

Taste responses of the gerbil IXth nerve

Summated taste responses to 12 taste solutions were recordedfrom the IXth (glossopharyngeal) nerve of 38 Mongolian gerbils(Meriones unguiculatus). 0.3 M NH₄Cl was the most effectivestimulant. The relative magnitude of the peak summated responsewas a positively accelerated function of log molar concentration.Absolute thresholds were determined for three chemicals: 0.002M NaCl, 0.0003 M HCl, and 0.002 M sucrose. The relative magnitudesof the responses to quinine, NH₄Cl, and KCl were greater forthe IXth nerve than for the chorda tympani nerve, whereas NaClwas more effective for the chorda tympani. A similar patternis seen in the rat. Acetic and citric acid may bind to commonreceptor sites. NH₄Cl, KCl, and HCl may also have receptor sitesin common.     

On the gustatory effect of amiloride in the monkey (Macaca mulatto)

Amiloride suppressed the summated chorda tympani proper nerveresponses to NaCl and LiCl in rhesus monkeys but did not suppressthe responses to sucrose, acesulfame-K, aspartame, Na-saccharin,Na-cyclamate, xylitol, D-glucose, D-fructose, glycine, KCl andNH₄Cl.     

Chemosensitive responses from the cat epiglottis

Responses were recorded from single and few-fiber preparationsof the cat superior laryngeal nerve during stimulation of theepiglottis with 0.5 M KCl, NH⁴Cl, NaCl, and LiCl; distilledwater, 0.005 M citric acid, 0.01 N HCl; and light touch. Epiglottalreceptive fields were mapped. The order of effective stimuliis KCl > HCl > NH⁴Cl > distilled water > citricacid; NaCl and LiCl are generally not effective. Since (a) similarstimuli elicit responses from the cat chorda tympani nerve duringstimulation of tongue taste buds, and (b) receptive fields ofsuperior laryngeal nerve fibers compare well with a map of epiglottaltaste buds, we conclude that the chemosensitive responses fromthe epiglottis are probably mediated by taste buds.     

The significance of apical K+ channels in mudpuppy feeding behavior

Behavioral experiments were conducted to determine the effectsof known potassium channel blockes, minnow extract (a naturalfood substance), and otber taste stimuli on feeding behaviorin the mudpuppy (Necturus maculosus). Groups of 10–15mudpuppies were presented with one or more of the followingchemicals dissolved in gelatin cubes: H₂O (control), NaCl (1.0M; 0.2 M), crude minnow extract (0.1 g/ml), CaCl₂ (1.0 M), citricacid (0.1 M; pH2, 0.01 M; pH 3), quinine HCl (0.1 M), tetraemylammoninmchloride (TEA; 0.1 M), KCl (1.0 M) and an amino acid mixturecontaining L-phenylalanine, L-arginine, L-histidine, L-valine,L-glutamic acid and L-lysinc at concentrations of 0.02 M and0.04 M each. Crude minnow extract was centrifuged and separatedinto components consisting of the following fractions: >500D, >1000 D and >14 000 D, each of which was presentedto mudpuppies in cubes. Results were expressed as the percentageof animals rejecting the cube within 5 min of taking the cubeinto the oral cavity. Statistical analysis using the Fisherexact test indicated that cubes containing chemicals known toblock (CaCl₂, TEA, citric acid and quinine) or permeate (KCl)apical K⁺ channels in mudpuppy taste cells were more aversivethan control cubes, and aversion was concentration-dependentIn contrast, cubes containing minnow extract and componentsof minnow extract < 14 000 D were significantly preferredover control cubes. Cubes containing a mixture of minnow extractand an aversive chemical were significantly less aversive thancubes containing only an aversive chemical. These findings suggestthat activation of the apical K⁺ conductance in mudpuppy tastecells triggers an aversive response, but the aversion can beameliorated by low molecular weight compounds present in minnowextract.     

PLASMALEMMA POTENTIAL IN NITELLA

A considerable part of the response of the vacuolar potentialof Nitella flexilis to the change of external KCl, NaCl, RbCl,LiCl, or CaCl₂ concentration is caused by the response of thecell wall (a cation exchanger) to the external medium. The potentialswere measured on the internodes whose cell sap was exchangedfor simple salt solutions. The potential difference across theplasmalemma which is the internal potential measured againstthe cell wall phase changes largely with the change in concentrationof the external KCl, but also more or less with that of theexternal NaCl, LiCl or RbCl. CaCl₂ depolarizes the plasmalemmapotential by about 50 mv when the concentration is increasedfrom 10^–5 M to 10^–3 M, and hyperpolarizes it againby about 40 mv from 10^–3 M to 10^–1 M leaving thelevel of the peak of the action potential almost unchanged. ¹This work was supported by Research Grants from the Ministryof Education of Japan     

Different Characteristics of Gustatory Responses Between the Greater Superficial Petrosal and Chorda Tympani Nerves in the Rat

The integrated responses to gustatory stimuli applied to thesoft palate were recorded from the greater superficial petrosalnerve (GSP) and were compared with those from the chorda tympaninerve (CT) innervating the anterior part of the tongue in therat. Stimuli included various concentrations of NaCl, sucrose,HCl and quinine hydrochloride, and 0.5 M of six sugars. Theinhibitory effects of amiloride on the responses to sodium salts,including various concentration of NaCl, 0.1 M sodium acetateand 0.01 M sodium saccharin, were also tested. Both the phasicand tonic responses to sugars in the GSP were significantlylarger than those in the CT, whereas both responses to NaClin the GSP were significantly smaller than those in the CT.Although amiloride at 50 µM significantly depressed thephasic and tonic responses to NaCl with a wide range of concentrationin the CT, little inhibitory effect was observed in the GSP.The tonic response to sodium acetate, when dissolved in amiloridesolution, was depressed to 15% of the control in the CT, andslightly but significantly depressed to 70% in the GSP. Theseresponse characteristics of the GSP may play important rolesin the processing of gustatory information. Chem. Senses 22:133–140, 1997.     

Developmental changes in sugar responses of the chorda tympani nerve in preweanling rats

To clarify developmental changes in the gustatory system of the rat, integrated taste responses from the chorda tympani (CT) nerve were recorded and analyzed at different postnatal ages. The response magnitude was calculated relative to the response to the standard, 0.1 M NH4Cl. Even at 1 week of age, the CT responded well to all tested 0.1 M chloride salts (NH4Cl, NaCl, LiCl, KCl, RbCl and CsCl). The responses to 0.1 M NaCl and LiCl increased with increasing age of the rat while response magnitudes to KCl, RbCl and CsCl did not change up to 8 weeks. At 1 week, the integrated response pattern was quite similar to that in adult rats for NaCl, HCl and quinine hydrochloride (QHCl). The concentration-response functions for NaCl, HCl, QHCl and sucrose at 2 weeks were essentially the same as those at 8 weeks. These results suggest that taste buds in the 2-week-old rat are functionally mature for the detection of the four basic taste stimuli. The relative magnitude of the responses to the various sugars was smaller at 1 week compared to the adult rat and reached a maximum at weeks 3-4, then decreased gradually with age. Among the six sugars, sucrose was the most effective followed by lactose. From weeks 1-4, the magnitude of the integrated taste response to fructose was smaller than that to lactose except at 3 weeks of age. Maltose, galactose and glucose were less potent stimuli than the other sugars tested. The response magnitude to lactose at 4 weeks had decreased compared to that for the other sugars. Taste responses to the sugars in preweanling and adult rats were not cross-adapted by the individual sugars. These results suggest that after 1 week of age during postnatal development in the rat, taste information from the CT rapidly increases in its importance for feeding behavior.     

Amiloride does not block taste transduction in the mudpuppy, Necturus maculosus

Activity of the glossopharyngeal nerve was recorded with bipolarsilver wire electrodes while taste stimuli were applied to thelingual surface in anesthetized mudpuppies. Taste stimuli wereinjected into a continuous stream of distilled water which wasrunning over the tongue, KCl, CaCl₂ and LiCl₂ at 0.4 M elicitedbrisk responses, as did HCl at 0.2 M and quinine at 6 x 10^–4M. Sucrose, glucose and saccharin did not elicit responses.Twenty amino acids were surveyed for their ability to evokea response at 0.04 M: 1-arginine, 1-valine, 1-phenylalanine,1-tryptophan, 1-tyrosine, 1-glutamic acid, 1-lysine and histidinealways evoked responses, whereas other amino acids either didnot evoke responses or only occasionally evoked responses. Thesupernatants from solutions of minced worms and minnows andPurina Trout Chow were effective taste stimuli. Pre-adaptingthe tongue to Ringer's solution by running a continuous streamof Ringer's solution over it eliminated responses to quinineand decreased responses to NaCl. Pre-adapting the tongue to10^–4 to 10^–3 M amiloride, a potent sodium channelblocker, did not alter the responses to NaCl, LiCl, or othertaste stimuli.     

Effects of Calcium and Potassium Ions on Phototaxis in Cryptomonas

Effects on positive phototaxis and the cell motility of 7 cationsin 5mM MOPS (morpholinopropane sulfonic acid) buffer (pH 7.0)containing 0.16 mM NaCl, 0.68 mM KCl, 0.5 mM CaCl₂ and 0.16mM MgCl₂ were studied in the unicellular flagellate Cryptomonaswith a photoelectrical measuring apparatus and photomicrography.When calcium ion was removed from the medium by adding 1 mMEGTA (ethylene glycol-bis-(ß-amino-ethylether)-N,N'-tetraaceticacid), the phototactic response was totally inhibited, but theswimming rate was not much affected. The effect of EGTA waspartially reversed by the addition of 1 mM CaCl₂. When 15mMKCl or RbCl was added to the medium, phototaxis was greatlyinhibited, but there was no significant influence on the swimmingrate. Similar but less inhibitory effects were induced in thepresence of NaCl, LiCl and CsCl. KCl-induced inhibition waspartially removed by the addition of 15 mM CaCl₂ or MgCl₂. (Received June 25, 1982; Accepted September 27, 1982)     

Electrophysiological responses of galeal contact chemoreceptors of Apis mellifera to selected sugars and electrolytes

Using conventional electrophysiological methods, the galeal sensilla chaetica of the honey bee, Apis mellifera, responded linearly to the log of solute concentrations of sucrose, glucose, fructose, NaCl, KCl, and LiCl but not to CaCl₂ or MgCl₂, which failed to give consistent responses. These sensillae had much higher firing rates for sugar than salt solutions; their relative responses to lower concentrations being NaCl < KCl < LiCl ? fructose < glucose ? sucrose. At higher concentrations NaCl < LiCl < KCl ? glucose < fructose ? sucrose. Four different spike types were seen. The first type had the highest amplitude and resulted from sugar stimulation. The second type had a lower height and occurred in the first 30 sec of salt stimulation. A third type with the lowest height appeared with those of the second type after prolonged stimulation with KCl. A fourth type with a high amplitude resulted from mechanical stimulation. The sensilla adapted to sugar solutions linearly to the logarithm of time and non-linearly to the log of salt concentrations. Glucose-fructose mixed-sugar solution effected synergism of response while sucrose solutions caused inhibition when mixed with glucose and/or fructose. Responses of the sensilla to mechanical stimulation showed phasic-tonic characteristics. None of the sensilla tested responded to water.     

The taste characteristics of sodium-saccharin in the rat: a re-examination of the dual taste hypothesis

A conditioned flavor aversion (CFA) was used to measure thetaste characteristics of 0.005 and 0.025 M Na-saccharin in therat. At neither concentration did the rats show generalizationto NaCl, quinine HCl or SO₄, sucrose octa-acetate or HCl, whereasgeneralization to sucrose was found at both concentrations.When the CFA was formed to 0.0001 M quinine HCl, no direct evidencewas found for generalization of the aversion to either saccharinconcentration. Thus at the two concentrations used Na-saccharinwould appear to be primarily sucrose-like in the rat taste systemwhen behavioral criteria are used.     

Asymmetrical neural cross-adaptation of hamster chorda tympani responses to sodium and chloride salts

Cross-adaptation has occurred when exposure to an adapting chemicalstimulus (A) reduces the response to a subsequent test stimulus(B). The degree of cross-adaptation between two stimuli is thoughtto reflect the overlap of their ‘neural activation processes’.We measured self- (A—A) and reciprocal crossadaptation(A—B, B—A) of the response of the hamster chordatympani nerve with lingual presentations of stimuli elicitingequal unadapted transient responses. Adapting and test stimuliwere 0.1 M NaCl, 0.1 M NaNO₃, 0.1 M NaBr, 0.4 M Na acetate (NaAc),0.09 M LiCl and 0.4 M NH₄Cl. Nearly complete and symmetricalcross-adaptation was seen for NaCl, NaNO₃ and NaBr. Those Nasalts paired with LiCl showed strong but asymmetrical cross-adaptation.Exposure to sodium completely eliminated the response to LiClbut not vice versa, suggesting that lithium and sodium are notcompletely interchangeable taste stimuli for the hamster chordatympani. Relatively little cross-adaptation between NH₄Cl andother salts suggested relatively separate neural activationprocesses. Strongly asymmetrical cross-adaptation was foundbetween NaAc and the other sodium salts. Responses to NaCl,NaNO₃ or NaBr were eliminated after adaptation to NaAc whereasthe response to NaAc during the reciprocal cross was strong.Asymmetries are discussed in reference to sensitivities of singlenerve fibers for the chorda tympani, effects of adaptation andthe concept of anion inhibition.     

Responses of single taste fibers and whole chorda tympani and glossopharyngeal nerve in the domestic pig, Sus scrofa.

Whole nerve, as well as single fiber, responses in the chorda tympani proper (CT) and glossopharyngeal (NG) nerves of 1- to 7-week-old pigs were recorded during taste stimulation. In the CT acids and in the NG bitter compounds gave the largest responses. Both nerves exhibited large responses to monosodium glutamate (MSG), MSG with guanosine 5'-monophosphate (GMP) and MSG with inositine 5'-monophosphate (IMP) as well as to glycine, xylitol, sucrose, fructose and glucose. Alitame, aspartame, betaine, neohesperedin dihydrochalcone (NHDHC), super-aspartame, saccharin and thaumatin elicited no or little response. Hierarchical cluster analysis of 49 CT fibers separated four major clusters. The M cluster, comprising 28.5% of all fibers, is characterized by strong responses to MSG, KCl, LiCl and NaCl. The responses to NaCl and LiCl were unaffected by amiloride. The H cluster (24.5%) includes units responding principally to acids. The Q cluster (18.5%) responds to quinine hydrochloride (QHCl), sucrose octaacetate (SOA) and salts with amiloride. The S cluster (28.5%) exhibits strong responses to xylitol, glycine and the carbohydrates as well as to MSG alone and to MSG with GMP or IMP. In 31 NG fibers, hierarchical cluster analysis revealed four clusters: the M cluster (10%), responding to MSG and MSG with GMP or IMP; the H cluster (13%), responding to acids; the Q cluster (29%), responding strongly to QHCl, SOA and tilmicosinR; and the S cluster (48%), responding best to xylitol, carbohydrates and glycine but also to the umami compounds. Multidimensional scaling analysis across fiber responses to all stimuli showed the best separation between compounds with different taste qualities when information from both nerves was utilized.     

Decline of IXth nerve taste responses following nerve transection

Summated impulse discharges to taste solutions were recordedfrom intact and transected IXth nerves in the Mongolian gerbil(Meriones unguiculatus). Five taste stimuli were used: 0.3 MNH₄Cl, 0.3 M NaCl, 0.01 M HCl, 0.01 M quinine hydrochloride,and 0.5 M sucrose. 0.3 M NH₄Cl was the most effective stimulus.Taste responses from intact nerves were stable for more than10 hours. Following IXth nerve transection, the peak summatedresponse to 0.3 M NH₄Cl declined by 50% in a mean of 119 min.(Some animals failed to show this taste response decline inthe winter months.) The transected IXth nerve's spontaneousactivity and responses to other taste solutions also typicallydeclined. The continued presence of normal compound action potentialsindicated that the transection-induced decline in taste responsesdid not result from a failure of impulse propagation mechanismsin the nerve trunk. The results are consistent with the propositionthat transection interferes with axonal transport of materialsvital to the short-term maintenance of taste responses.     

THE VAPOUR PRESSURES OF AQUEOUS SOLUTIONS WITH SPECIAL REFERENCE TO THE PROBLEM OF THE STATE OF WATER IN BIOLOGICAL FLUIDS

Data for the depression of vapour pressure are presented for the following aqueous solutions: NaCl (0.03 to 0.1 molar), KCl (0.03 to 0.1 molar), urea (0.05 to 0.5 molar), sucrose (0.05 to 0.10 molar), lactic and succinic acids, creatine, CaCl₂ (0.05 molar), and mixtures of these substances with one another and with certain other solutions (gelatin, gum acacia, sea water, LiCl, etc.). The relation of the depression of vapour pressure of a mixed solution to that of solutions of the individual constituents was investigated in order to ascertain to what extent such studies may be used for the determination of the degree of hydration, or of the state of water, in solutions. Organic substances (urea, sucrose, etc.) showed anomalous results which were markedly affected and unpredictable in mixed solutions. They are, therefore, unsuited for the study of water binding. In the case of solutions of inorganic substances—LiCl and CaCl₂—the principle of the additive nature of colligative properties is also only approximately true—except perhaps in very dilute solutions. The limitations of the colligative method for determining the degree of hydration have been defined in accord with the above findings. Studies of the vapour pressures of mixtures of gelatin or gum acacia with NaCl or KCl demonstrated that hydration in gelatin is relatively small at pH = 7 and undetectable in gum acacia solutions. The view, therefore, that hydrophilic colloids are strongly hydrated has not been substantiated. The passage from the sol to the gel state also was not accompanied in gelatin or in blood by any appreciable change in the degree of hydration of the hydrophilic colloids present in these substances.     

Effect of Sudden Salt Stress on Ion Fluxes in Intact Wheat Suspension Cells

Although salinity is one of the major problems limiting agriculturalproduction around the world, the underlying mechanisms of highNaCl perception and tolerance are still poorly understood. Theeffects of different bathing solutions and fusicoccin (FC),a known activator of plasma membrane ATPase, on plasma membranepotential (E_m) and net fluxes of Na⁺, K⁺and H⁺were studied inwheat suspension cells (Triticum aestivum) in response to differentNaCl treatments. E_mof cells in Murashige and Skoog (MS) mediumwas less negative than in cells exposed to a medium containing10 mM KCl + 0.1 m M CaCl₂(KSM) and to a basic salt medium (BSM),containing 1 m M KCl and 0.1 m M CaCl₂. Multiphasic Na⁺accumulationin cells was observed, peaking at 13 min after addition of 120m M NaCl to MS medium. This time scale was in good agreementwith net Na⁺flux changes measured non-invasively by moving ion-selectivemicroelectrodes (the MIFE system). When 120 m M NaCl was addedto all media studied, a quick rise of Na⁺influx was reversedwithin the first 20 min. In both 120 and 20 m M NaCl treatmentsin MS medium, net Na⁺efflux was observed, indicating that activeNa⁺transporters function in the plant cell response to saltstress. Lower external K⁺concentrations (KSM and BSM) and FCpre-treatment caused shifts in Na⁺fluxes towards net influxat 120 m M NaCl stress. Copyright 2000 Annals of Botany Company Sodium, potassium, proton, membrane potential, fusicoccin, salt stress, wheat, Triticum aestivum     

Taste responses of human chorda tympani nerve

Records from humans of summated action potential dischargesof the chorda tympani nerve were examined. The magnitudes ofneural and psychophysical responses were well related only whenthe comparison was made within a given taste quality. The responseto a mixture of 0 02 M citric acid and 0.5 M sucrose was lessthan the sum of the separate responses to the mixture components.Citric acid failed to cross-adapt the response to sucrose, implyingthe receptor sites for sucrose are independent of citric acid.The human chorda tympani nerve shows vigorous responses to mechanicalstimulation and cooling of the tongue that are maintained aftertreatment of the tongue with a water extract of the herb Gymnemasylvestre. Gymnema extract selectively suppressed the responseto all sweeteners tested (sucrose, fructose, saccharin and cyclamate)and also suppressed by – 50% the water-after-citric-acidresponse which has a predominantly sweet taste. Gymnema suppressedby 0 – 10% the water-rinse response following NaCl. fructoseand sucrose that have a predominantly bitter-sour taste. Water-rinseresponses were present even when mechanical and thermal stimulationof the tongue were minimized. The human chorda tympani nerveappears to have positive water-rinse taste responses. Theseare solute-specific off-responses that are probably mediatedby receptor sites independent of those responsible for the on-responseto the given solute.     

Effects of amiloride on gustatory neural responses to salts in the frog.

In frogs, the glossopharyngeal nerve (GL) innervates taste receptors on almost the entire tongue. The mandibular branch (MBF) and palatine branch (PN) of the facial nerve innervate taste receptors on a very small area at the base of the tongue and on the palate, respectively. In the present study, effects of amiloride, an epithelial sodium channel blocker, on the tonic responses of the GL, MBF and PN in frogs to NaCl, LiCl, KCl and CaCl(2) were investigated. In three nerves, amiloride at 0.5 mM, a relatively high concentration, did not affect the responses to 0.15 (concentration just above threshold)-0.5 M NaCl, 0.5 M LiCl and 0.3 M KCl, whereas it almost completely inhibited the response to 1.0 mM CaCl(2). Amiloride may exert an inhibitory action on the response to CaCl(2) by a competitive antagonism between Ca(2+) and a monovalent cation of amiloride, because the response to Ca(2+) is competitively inhibited by other cations such as Na(+) and Mg(2+). The lack of inhibitory effect of amiloride on the responses in the GL, MBF and PN to NaCl suggests that amiloride-sensitive sodium channels in the apical membrane of taste receptor cells are not involved in sodium taste transduction in frogs.     

MICRURGICAL STUDIES IN CELL PHYSIOLOGY : V. THE ANTAGONISM OF CATIONS IN THEIR ACTIONS ON THE PROTOPLASM OF AM?BA DUBIA.

I. The Plasmalemma. 1. On the plasmalemma of amebæ CaCl₂ antagonizes the toxic action of LiCl better than it does NaCl, and still better than it does KCl. MgCl₂ antagonizes the toxic action of NaCl better than it does LiCl and still better than it does KCl. 2. CaCl₂ antagonizes the toxic action of LiCl and of KCl better than does MgCl₂: MgCl₂ antagonizes NaCl better than does CaCl₂. II. The Internal Protoplasm. 3. The antagonizing efficiency of CaCl₂ and of MgCl₂ are highest against the toxic action of KCl on the internal protoplasm, less against that of NaCl, and least against that of LiCl. 4. CaCl₂ antagonizes the toxic action of LiCl better than does MgCl₂: MgCl₂ antagonizes the toxic action of NaCl and of KCl better than does CaCl₂. 5. LiCl antagonizes the toxic action of MgCl₂ on the internal protoplasm more effectively than do NaCl or KCl, which have an equal antagonizing effect on the MgCl₂ action. III. The Nature of Antagonism. 6. When the concentration of an antagonizing salt is increased to a toxic value, it acts synergistically with a toxic salt. 7. No case was found in which a potentially antagonistic salt abolishes the toxic action of a salt unless it is present at the site (surface or interior) of toxic action. 8. Antagonistic actions of the salts used in these experiments are of differing effectiveness on the internal protoplasm and on the surface membrane.