The Mechanistic Conception of Life

and this explanation casts light on the nature of the protective or antagonistic action of salts. For the antagonistic action of a salt of lead or zinc against the toxic action of sodium chloride can only consist in the lead salt protecting the embryo against the toxic action of the NaCl. But how is this protective action possible? We have mentioned that if we put the young fish, immediately after hatching, into a pure m/2 solution of sodium chloride the animals die very quickly, but that they live indefinitely in the sodium chloride solution if we add both CaCl., and KCl. How does it happen that for the embryo, as long as it is in the egg shell, the addition of CaCl, to the NaCl solution suffices, while if the fish is out of the shell the addition of CaCl., alone is no longer sufficient and the addition of KCl also becomes necessary ? Moreover, if we try to preserve the life of the fish after it is taken out of the egg in an m/2 sodium chloride solution by adding ZnSO^, or lead acetate, to the solution we find that the fish die even much more quickly than w^ithout the addition.^

If we look for the cause of this difference our attention is called to the fact that the fish, as long as it is in the egg, is separated from the surrounding solution by the egg membrane. This egg membrane possesses a small opening, the so-called micropyle, through which the spermatozoon enters into the egg. I have gained the impression that this micropyle is not closed as tightly immediately after fertilization as later on, since the newly fertilized egg is killed more rapidly by an m/2 solution of NaCl than it is killed by the same solution one or two days after fertihzation. One can imagine that the micropyle contains a wad of a colloidal substance which is hardened gradually to a leathery consistency if the egg remains in the sea-water.

1 R. Lillie has found that in the larvae of Arenicola a slight antagonism between NaCl and ZnS04 can be proved. This shows that the general laws of antagonism between two salts differ in degree but not in principle in the living organism and the dead envelop of the fish egg. With the process of hardening, or tanning, it becomes more impermeable for the NaCl solution. This process of hardening is brought about apparently very rapidly if we add to the m/2 NaCl solution a trace of a salt of a bivalent metal like Ca, Sr, Ba, Zn, Pb, Mn, Ko, and Xi, etc. It is also possible that similar changes take place in the whole membrane. The process of rendering the m/2 Na solution harmless for the embryo of the fish, therefore, depends apparently upon the fact that the addition of the bivalent metals renders the micropyle or perhapsthe whole membrane of the egg more impermeable to NaCl than was the case before.

But these are only one part of the facts which throw a light upon the protective or antagonistic action of salts. Further data are furnished by experiments which I made together with Professor Gies, also on the eggs of Fundulus. Gies and I were able to show that not only are the bivalent metals able to render the sodium chloride solution harmless, but that the reverse is also the case, namely, that NaCl is required to render the solutions of many of the bivalent metals, for instance ZnSO^, harmless. (That the SO^ ion has nothing to do with the result wasshown before by experiments with Na^SO^.)

If the eggs of Fundulus are put immediately after fertilization into distilled w^ater, a large percentage of the eggs develop^ often as many as 100 per cent, and the larvae and embryos formed in the distilled water are able to hatch. If we add^ however, to 100 c.c. of distilled water that quantity of ZnSO^ which is required to render the NaCl solution harmless, all the eggs are killed rapidly and not a single one is able to form an embryo. If we add varjdng amounts of NaCl we find that, beginning with, a certain concentration of NaCl, this salt inhibits the toxic effects of ZnSO^ and many eggs are able to form an embryo. This can be illustrated by the foUo^Wng table :

This table shows that the addition of NaCl, if its concentration exceeds a certain limit, namely, m/8, is able to render the ZnSO^ in the solution comparatively harmless. If we now assume that ZnSO^ renders the 5/8 m NaCl solution harmless by rendering the egg membrane comparatively impermeable for NaCl we must also draw the opposite conclusion, namely, that NaCl renders the egg membrane comparatively impermeable for ZnSO^. We therefore arrive at a new conception of the mutual antagonism of two salts, namely, that this antagonism depends, in this case at least, upon a common, cooperative action of both salts on the egg membrane, by which action this membrane becomes completely or comparatively impermeable for both salts. And from this we must draw the further conclusion that the fact that each of these salts, if it is alone in the solution, is toxic, is due to its comparatively rapid diffusion through the membrane, so that it comes into direct contact with the protoplasm of the germ.

As long as we assumed that each of the two antagonistic salts acted, if applied singly, in the opposite way from its antagonist, it was impossible to understand these experiments or find an analogue for them in colloid chemistry. But if we realize that NaCl alone is toxic because it is not able to render the egg membrane impermeable; and that ZnSO^ if alone in solution is toxic for the same reason; while both combined are harmless (since for the 'banning" of the membrane the action of the two salts is required) these experiments become clear.

We may, for the sake of completeness, still mention that salts alone have such antagonistic effects; glycerin, urea, and alcohol have no such action. On the other hand, ZnSO^ was not only able to render NaCl harmless, but also LiCl, NH^Cl, CaCl2, and others; and vice versa. These experiments on the egg of Fundulus are theoretically of importance, since they leave no doubt that in this case at least the "antagonistic" action of salts consists in a modification of the egg membrane by a combined action of two salts, whereby the membrane becomes less permeable for both salts.

It is not easy to find examples of experiments in the literature which are equally unequivocal in regard to the character of antagonistic salt action ; but I think that some recent experiments by Osterhout satisfy this demand. It has long been a question whether or not cells are at all permeable for salts. Nobody denies that salts diffuse much more slowly into the cells than water; but some authors, especially Overton and Hoeber, deny categorically that they can diffuse at all into the cells. Overton's view is based partly on experiments on plasmolysis in the cells of plants. If the cells of plants, for example, those of Spirogyra, are put into a solution of NaCl or some other salt of sufficiently high osmotic pressure, the volume of the contents of the cell decreases through loss of water and the protoplasm retracts, especially from corners of the rigid cellulose walls. Overton maintains that this plasmolysis is permanent, and concludes from this

that only water but no salt can diffuse through the cell-wall; since otherwise salt should gradually diffuse from the solution into the cell, and through this increase in the osmotic pressure of the cell the water should finally diffuse back into the cell and restitute the normal volume of the cell. According to Overton this does not happen. Osterhout has recently shown that Overton's observations were incomplete in a very essential point and that in reality the plasmolysis, which occurs in this case when the cell is put into the h^q^ertonic solution, disappears again in a time which varies wdth the nature of the salt in solution. This stage of reversion of plasmolysis had been overlooked by Overton. If the cell, however, remains permanently in the h^^jertonic sodium chloride solution, a shrinking of the contents of the cell takes place again, which superficially resembles plasmolysis, but which in reality has nothing to do with plasmolysis, but is a phenomenon of death. That this second ''false plasmolysis," as Osterhout calls it, has nothing to do with the hypertonic character of the solution was proved by the fact that hypotonic solutions of toxic substances may produce the same phenomenon.

a portion of a Spirogyra filament was plasmolyzed in .2 m CaCl2, but not in . 195 m CaClo. A .29 m NaCl solution has approximately the same osmotic pressure as a .2m CaClo solution. But on placing another portion of the same Spirogyra filament in a . 29 m NaCl solution the expected plasmolysis does not occur and it is impossible to plasmolyze the cells until they are placed in . 4 m NaCl. Osterhout explains this difference in the concentration of the two salts required for plasmolysis by the assumption that NaCl diffuses more rapidly into the cell than CaCl.,, a conclusion which I reached also on the basis of my earlier experiments on animals. Osterhout's experiments also show that the antagonism of NaCl and CaClg depends partly on the facts that the two salts

inhibit each other from diffusing into the cells, and this conclusion is based among others upon the following experiment. By dividing a Spirogyra filament into several portions it was found that it was plasmolyzed in .2 m CaCL and in .38 m XaCl, but neither in .195 m CaCla nor in .375 m NaCl. On mixing 100 c.c. .375 m XaCl with 10 c.c. . 195 m CaClo and placing other portions of the same filament in it, prompt and very marked plasmolysis occurred.

The explanation for this observation lies in the fact that in the mixture of NaCl and CaCl, the two salts render their diffusion into the cell mutually more difficult. After a longer period of time the plasmolyzed cells can expand again in a mixture of NaCl and CaCl,, but that occurs much later than if they are in the pure NaCl solution. These experiments are the analogue of the observation on the embryo of the eggs of Fundulus in which a pure solution of ZnSO^ diffused rapidly through the membrane or micropyle, while, if both salts were present, the diffusion was inhibited or considerably retarded.

While the observations of Osterhout show that Overton was not justified in using the experiments on plasmolj^sis to prove that the neutral salts cannot diffuse into the cells, yet they do not prove that these salts diffuse into the cell under normal conditions. In Osterhout 's experiments the cells are in strongly h^T^ertonic solutions and it does not follow that such solutions act like isotonic, perfectly balanced solutions. Wasteneys and I have recently sho^Mi that the toxic action of acids upon Fundulus can be annihilated by salts. If we add 0.5 c.c. n/10 butyTic acid to 100 c.c. of distilled water these fish die in 2| hours or less. In solutions which contain 0.4 c.c. or less acid they can live for a week or more. If we add, however, 0 . 5 c.c. of butyric acid to 100 c.c. of solutions of NaCl of various concentration, we find that above a certain limit

the NaCl can render the acid harmless. It is needless to say that the NaCl used in these experiments was strictly neutral and that the amount of acid present in the mixture of acid and salt was measured. The following experiment may serve as an example. Six fish were put into 500 c.c. of each of the following seven mixtures, namely, After certain intervals the number of surviving fish was ascertained. The result is given in Table II. If the amount of acid was increased, the amount of NaCl also had to be increased to render the acid harmless. In order to render 0.5 c.c. n/10 but^Tic acid pro 100 c.c. solution harmless, 10 c.c. m/2 NaCl had to be added; while 0.8 c.c. butyric acid required 20 c.c. and 1.0 c.c. butyric acid required about 28 c.c. m/2 NaCl in 100 c.c. of the solution.

Not only butyric acid, but any kind of acid, could be rendered harmless by neutral salts, e.g., HCl by NaCl. Wasteneys and I could show that the rate of the absorption of acid by the fish is the same in solutions with and without salt. This proves that the action of the salts consisted in this case not in preventing the diffusion or absorption of the acid, but in modifying the deleterious effect of the absorbed acid. We can state a little more definitely the cause of death by acid. If we put the fish into a weak acid solution in distilled water just strong enough to kill the fish in from one to two hours (e.g., 500 c.c. H,O+2.0 c.c. n/10 HCl), we notice that the acid very soon makes the normally transparent epidermis of the fish opaque, and a little later the epidermis falls off in pieces and shreds. This, however, is probably not the direct cause of the death, but I am inclined to assume that the fish die from suffocation caused by a similar action of the acid upon the gills.

The action of the acid upon the epidermis of the body as well as upon the gills is prevented through the addition of neutral salts. It is well known that the action of acids upon proteins can be inhibited by neutral salts. ^ Thus the internal friction of certain protein solutions is increased by acids while the addition of neutral salts inhibits this effect (Pauli). The swelling of gelatin caused by acid is inhibited by salts (Procter). ^ It is possible that in the experiments with acid the fish is killed in the following way. The acid causes certain proteins in the surface layer of the epithelial cells of the gills and of the skin to swell, whereby this surface layer becomes more permeable for the acid. The acid can now diffuse into the epithelial cells and act on the protoplasm, whereby the cells are killed. If salts are present in the right concentration, the combined action of acid and salt causes a dehydration of the surface film

1 It seems that the first, experiments on the antagonism between acids and salts were published by the author in PflUgers Archiv, Vol. LXXV, p. 308, 1899. 2 The beautiful osmometric experiments of R. Lillie should also be mentioned in this connection. of these cells, as it does in the experiments on gelatin or as in the cases of tanning of hides by the combined action of acids and salt solutions. This combined dehydrating or "tanning" action of acid and salts on the surface of the epithelial cells of the gills diminishes the permeability of this layer for the acids and prevents them from diffusing into the cells and thus destroying the protoplasm. In this way the gills are kept intact and the life of the fish is saved.

As long as the amount of acid is small the amount absorbed is not essentially diminished by the presence of salts ; but while in the presence of salts the acid is consumed in the tanning action of the surface layer of the cells, or is absorbed in this layer; if no salt is present part of the acid diffuses into the epithelial cells and kills the latter. We have thus far considered the cases of antagonism between two electrolytes only. The case of the antagonism between three electrolytes is a little more complicated.

We choose as an example the antagonism between NaCl, KCl, and CaCl, — the antagonism which is most important in life phenomena. If the mechanism of the antagonism between NaCl, on the one hand, and KCl and CaCl,, on the other, is of the same nature as that between NaCl and ZnSO^ in the case of the eggs of Fundulus, it must be possible to show that not only is NaCl toxic if it is alone in solution, and that it is rendered harmless by the two other salts, but that the reverse is true also. This can be proved in the case of KCl. To demonstrate it, we have again to experiment on organisms which are, in wide limits, independent of the osmotic pressure of the surrounding solution since the concentration of the KCl in sea-water is very low. The experiments were carried out by Mr. Wasteneys and myself on Fundulus. The method consisted in putting six fish, after washing them twice with distilled water, into 500 c.c.

of the solution. It was ascertained from day to day how many fish survived. When the fish were put into pure solutions of KCl of the concentration in which this salt is contained in the sea-water (2.2 c.c. m/2 KCl in 100 c.c. of the solution) they died mostly in less than two days. This is not due to the low concentration of the KCl solution, which is only 1/50 of that of the sea-water, since the fish can live indefinitely in a pure NaCl solution of the same concentration as that in which the KCl exists in the seawater.

If we add to the toxic quantities of KCl increasing quantities of NaCl, we find that as soon as the solution contains 17 or more molecules of NaCl to one molecule of KCl, the toxic action of KCl is considerably diminished, if not completely comiteracted. The following table may serve as an example : More accurate determinations showed that already a 3/16 m NaCl solution renders the solution of 2 . 2 c.c. m/2 KCl in 100 c.c. of the solution harmless.

It was next determined whether different concentrations of KCl required different concentrations of NaCl. It was found that the coefficient of antagonization KCl /NaCl has an approximately constant value, namely, about 1/17, as the following table shows. What happens if we vary this ratio ? If we add too little NaCl to the KCl solution, namely, only 1 to 10 molecules NaCl to 1 molecule of KCl, the solution becomes more harmful than if KCl is alone in solution; if we add considerably more than 17 molecules NaCl, e.g., 50 molecules to one molecule of KCl, the solution becomes toxic again; and the more so the higher the concentration of NaCl. This indicates that the antagonistic effect requires a rather definite ratio of the two salts. This furnishes the reason why an m/2 solution of NaCl can, as a rule, not be rendered completely harmless by the mere addition of KCl, but that in addition CaClg is needed.

If we add to 100 c.c. m/2 NaCl enough KCl to make the ratio KCl: NaCl = 1/17 we find that the antagonization of KCl: NaCl becomes incomplete. If the amount of KCl in 100 c.c. of the solution exceeds 2.2 c.c. m/2 KCl, antagonization is still to some extent possible, but it becomes more incomplete the higher the concentration of KCl. For this reason it is not possible to render an m/2 solution of NaCl harmless by the mere addition of KCl. CaClg acts upon KCl similarly as does NaCl, but it acts more powerfully; i.e., the coefficient of antagonization, KCl/CaCla, is several hundred or a thousand times as great as that of KCl/NaCl, as the following tables shows.

The coefficients are not as regular as in the case of antagonization of KCl by NaCl. This is due to the fact that the minimal value of CaCl^ at which it renders the KCl harmless cannot be determined as sharply as the limit for NaCl. Why is less CaCl, required than NaCl ? We can only answer with a suggestion first offered by T. B. Robertson, namely, that CaCIg produces its protective effect through the formation of a comparatively insoluble compound (in this case on the gills or the rest of the surface of the animal) while NaCl acts through the formation of a compound which is more soluble. This view is corroborated by the observation which we made, that Sr is just as effective to antagonize KCl as CaCl,, but that Mg is much less efficient. This would correspond with the wellknown fact that many strontium salts are just as insoluble, if not more insoluble, than the calcium salts, while the magnesium salts are often incomparably more soluble, for instance, in the case of the sulphates. BaCl, antagonizes KCl also powerfully, but, probably, in consequence of the fact that the substances formed at the surface of the animal or the gills, diffuse slowly into the cells, the fish do not remain alive as long if Ba is used as if the more harmless Ca and Sr are used.

It is very remarkable that CaCl, renders harmless any given concentration of KCl below 6.6 c.c. m/2 KCl in 100 c.c. of the solution, but not above this limit. This limit is exactly the same which we found in the case of antagonization of KCl by NaCl. Even the combination of NaCl and CaCl, does not permit us to render harmless more than 6.6 c.c. m/2 KCl in 100 c.c. of the solution. If we try to render NaCl harmless by KCl and CaCL, we find that CaCl, can antagonize even a 6/8 m and a 7/8 m solution of NaCl, while KCl ceases to show any antagonistic effect if the NaCl solution exceeds m/2 or 5/8 m.

Experiments with pure CaClg solutions give the result that this substance is harmless in a solution of that concentration in which this salt is contained in the sea-water. Fundulus can live indefinitely in a solution of 1.5 c.c. m/2 CaClg in 100 c.c. Botanists have also found that weak solutions of CaCl^ are comparatively little toxic. This gives us the impression that the effect upon the surface film of protoplasm produced by CaCla is especially important for the protection of the protoplasm. This conclusion receives an indirect support by the well-known experiments of Herbst, who found that in sea-water deprived of calcium the segmentation cells of a sea-urchin embryo fall apart through the disintegration or liquefaction of a film which surrounds the embryo and keeps the cells together. If such eggs are brought back into solution containing calcium the film is restored and the cells come into close contact again.

It is therefore not impossible that the mechanism of the antagonism between KCl and NaCl is similar to that found between NaCl and ZnSO^. It seems only due to the high concentration of the NaCl in the sea-water and in the blood that, in addition to KCl and NaCl, CaClj is needed. But the case is not so unequivocal as the previously mentioned cases of antagonism between only two electrolytes. It is necessary for our understanding of the life-preserving action of salts that we do not depend merely on conclusions drawn from experiments, but that we must be able to see directly in which way abnormal salt solutions cause the death of the cell. Such an opportunity is offered us through the

observation of the eggs of the sea-urchin. If we put the ferti- Hzed eggs of the sea-urchin into an abnormal salt solution, a destruction of the cell gradually takes place. The destruction, as a rule, begins on the surface of the protoplasm, and consists very often in the formation and falling off of small granules or droplets. This process gradually continues from the periphery toward the center until the whole egg is disintegrated. For different salt solutions the picture of the disintegration is a little different, but sufficiently characteristic for a given solution, so that if one become familiar with these pictures, one is able to diagnose to some extent the nature of the solution from the way in which the cell disintegrates.

This process of disintegration can be observed if the eggs are put into a pure solution of sodium chloride, or in a mixture of sodium chloride and calcium chloride, or in a mixture of sodium chloride and potassium chloride. If, however, all three salts are used in the proportion in which they occur in the sea-water no disintegration takes place and the surface of the egg remains perfectly smooth and normal. One gains the impression as if the protoplasm of the egg were held together by a continuous surface film of a definite texture. If we put the egg into an abnormal solution this surface film is modified and changed, and the change of the surface film is often followed by a gradual process of disintegration of the rest of the cell.