CHAPTER VIII.

_PHYSIOLOGY OF POISONING_ (_continued_).

ACTION OF VENOMS ON THE BLOOD.

On making an autopsy of an animal which has succumbed to intoxication by snake-venom, we find that the blood in the heart and large vessels is sometimes coagulated into a mass, sometimes entirely fluid, and that, in certain cases, it is as black as prune-juice, while in others it is of a fine transparent red colour.

These differences in the effects of venom upon the blood are due to the fact that the various venoms contain in variable proportions, besides the _neurotoxic_ substance which represents the true venomous _toxin_, other substances which act, some upon the plasmasia or fibrin-ferment, or upon the fibrin, others upon the red corpuscles, others on the leucocytes, and others again on the endothelium of the blood-vessels.

A.--EFFECTS OF VENOM ON THE COAGULATION OF THE BLOOD.

It was observed long ago by Fontana[29] that after viper-bites the blood remains fluid, and Brainard[30] on the contrary, pointed out that, in the case of animals that succumb very rapidly after having been bitten by a _Crotalus_, the blood was always found coagulated into a mass, while, when a certain interval of time had elapsed since the bite, it remained fluid. Weir Mitchell[31] explained these differences by the hypothesis that, in cases of rapid death, the blood had not had time to become modified by the venom.

Later on it was found by Sir Joseph Fayrer, and subsequently by Halford,[32] in Melbourne, C. J. Martin,[33] in Sydney, G. Lamb,[34] in Bombay, and recently by Noc, in my laboratory, that the venoms of COLUBRIDÆ, especially those of _Naja tripudians_ and AUSTRALIAN species of this family, always leave the blood fluid after death, while the venoms of VIPERIDÆ, on the contrary, are usually coagulant.

On the other hand, it was observed by Phisalix,[35] and at an earlier date by Mosso, of Turin, that the venom of _Vipera berus_ causes the blood of the dog to lose its coagulability, while, on the contrary, the same venom is actively coagulant as regards the blood of the rabbit.

How are these differences of action to be explained? It was found by Delezenne,[36] who made an excellent study of the phenomena following the injection of peptone, extracts of organs, and other anticoagulant substances into the organism, that those of these substances that render the blood non-coagulable always dissolve the leucocytes, and thus set at liberty two antagonistic bodies which they contain. One of these substances is coagulant and is found retained by the liver, while the other remains in solution in the plasma, and keeps the blood fluid after issuing from the vessels.

Now, certain extracts of organs, ricin, abrin and certain venoms in weak doses, retard coagulation, while in large doses, on the contrary, they produce partial or general intravascular clotting.

It is believed by Delezenne that the explanation of this phenomenon may be that the doses, which are weak but sufficient to produce the disintegration of the leucocytes, injure the red corpuscle in only a slight degree, while the stronger doses are equally destructive to the two kinds of blood corpuscles.

It follows that we must understand that there are two phases in the action of venoms: one _negative_, when the dose absorbed does not injure the leucocytes; the other _positive_, when the leucocytes are destroyed.

If the blood of the dog remains non-coagulable when mixed with doses of venom which, on the contrary, are actively coagulant for the blood of the rabbit, the reason would be that the leucocytes of these animals are not equally resistant to venom.

This conception, however, does not conform to the facts that I have myself observed. I have always found that viper-venom, mixed with citrate- or oxalate-plasma of the dog, rabbit, or horse, coagulates these various plasmas when the venom is in weak doses, while with strong doses coagulation is not produced. To be quite accurate, it should be stated that the quantity of venom necessary to render the plasma of the dog, or of the horse, non-coagulable is less than that which must be employed in the case of the plasma of the rabbit.

I have caused Noc to take up anew the study of this question in my laboratory, with venoms of nine different origins, and I here give a _résumé_ of the results of his researches.[37]

I. COAGULANT VENOMS.

The venoms of VIPERIDÆ studied range themselves as follows according to their coagulant power:--

CROTALINÆ: _Lachesis lanceolatus_ (Fer-de-lance, Martinique). _Lachesis neuwiedii_ (Urutù, Brazil). _Lachesis mutus_ (Bushmaster, or Surucucu, Brazil). _Lachesis flavoviridis_ (Japan). VIPERINÆ: _Vipera russellii_ (Daboia, India).

The venoms of _Ancistrodon contortrix_ and _A. piscivorus_ (CROTALINÆ) proved entirely inactive.

No COLUBRINE venom exhibited coagulant power, whatever the dose employed.

There is, therefore, a very decided difference between venoms of divers origins as regards their effects upon the coagulation of the blood.

Noc has determined more especially the coagulant action of the venom of _Lachesis lanceolatus_ (Fer-de-lance of Martinique) upon 1 per cent. citrate-plasmas, 1 per cent. oxalate-plasmas, 4 per cent. chloridate-plasmas, and upon blood rendered non-coagulable by extract of leeches’ heads. He found that, while weak doses of venom (1 milligramme per cubic centimetre of horse- or rabbit-plasma) produce coagulation in a few minutes in the citrate-plasmas, chloridate-plasmas, or those treated with extract of leeches, the doses of the same venom greater than 4 milligrammes on the contrary suppress the coagulability of these plasmas, even when there be added to them doses of chloride of calcium (for the citrate- and oxalate-plasmas), or of distilled water (for the chloridate-plasma), or of fibrin-ferment (for the plasma treated with leech-extract) sufficient to cause rapid coagulation in the control tubes that do not contain venom.

Noc also observed that the venom of the same species of snake (_Lachesis lanceolatus_), when heated to 75° C., entirely loses its coagulant properties; and that, with a temperature of 58° C., its coagulant power already commences to diminish. When heated for thirty minutes at a temperature of 65° C., a dose of 1 milligramme does not coagulate more than 1 c.c. of citrate-plasma in one hour. G. Lamb has likewise found that the venom of _Vipera russellii_ loses its coagulant power when heated to 75° C.

The coagulant substance in these venoms is precipitable by alcohol at the same time as the _neurotoxin_ and other active substances. The precipitate, when dissolved again in physiological water, preserves all the properties of the original solution.

Antivenomous _anticolubrine_ serum, that is to say, that furnished by horses vaccinated against the venoms of the _Cobra_ and the _Krait_, does not prevent coagulation by coagulant venoms. This need not surprise us, since the coagulant substances in venoms are destroyed by heating, and the animals vaccinated in order to obtain antitoxic serum are usually inoculated exclusively with heated venoms.

It is easy, however, to obtain active serums specific against the coagulant venoms; it is sufficient to treat these animals by inoculation with progressively increasing doses of the same venoms unheated. I have had no difficulty in achieving this result with small laboratory animals (guinea-pigs and rabbits) and also with the horse, but I have never had at my disposal a sufficient amount of the venoms of _Lachesis_ or _Vipera russellii_ to undertake with them the regular acquisition of large quantities of horse-serum, at once _antineurotoxic_ and _anticoagulant_. The preparation of such a serum, nevertheless, presents much interest for certain countries, such as Burma, where the Daboia (_Vipera russellii_) is almost as common as the Cobra, and Brazil, where nearly all the casualties due to venomous snakes are produced by _Lachesis_.[38]

II.--ANTICOAGULANT VENOMS.

Contrary to what is observed with the venoms of VIPERIDÆ in general, all the venoms of COLUBRIDÆ and, as exceptions to the rule, the venoms of some North American CROTALINÆ (_Ancistrodon contortrix_ and _A. piscivorus_) suppress the coagulability of the blood _in vivo_ and _in vitro_. It is, however, important to observe that, _in vivo_, the blood remains fluid after death only if the dose of venom absorbed has been sufficient. _In vitro_ this phenomenon is easier to study, and has been the subject of several important memoirs.

Halford,[39] Sir Joseph Fayrer,[40] C. J. Martin,[41] Delezenne,[42] Phisalix,[43] and lastly Noc,[44] have shown that the venoms of COLUBRIDÆ exert a manifestly anticoagulant action upon citrate-, chloridate-, or oxalate-plasmas, and also upon blood mixed with venom on issuing from the vessels.

On adding 1 milligramme of _Cobra_-, _Bungarus_-, Australian _Pseudechis_-, or _Ancistrodon_-venom to 1 c.c. of citrate-, oxalate-, or chloridate-plasma, and supplementing the mixture, after varying periods of contact, with a quantity of chloride of calcium (for the citrate- or oxalate-plasmas), or distilled water (for the saline plasma) sufficient to produce coagulation in a few minutes in the control tubes without venom, we find that coagulation no longer takes place after one hour in the tubes containing _Cobra_- or _Bungarus_-venom, and after ten minutes in those that contain the venom of _Ancistrodon_.

In doses less than 1 milligramme for 1 c.c. of plasma, these venoms by themselves never produce coagulation as do those of _Lachesis_ or _Vipera russellii_. They are thus sharply differentiated in this respect.

If fresh blood issuing from the arteries of an animal be received in a vessel containing a sufficient quantity of COLUBRINE-venom (that of the Cobra for example), and steps be immediately taken to ensure the perfect mixture of the venom and the blood, we find that the latter has entirely lost its coagulability, just as though it had been mixed with peptone or extract of leeches’ heads.

Again, if a mixture be made _in vitro_ of coagulant venoms, such as that of the _Lachesis_, with anticoagulant venoms such as that of the _Cobra_ or of _Ancistrodon_, it is found that these mixtures, when properly effected, become neutral, so that the respective effects of the component venoms are entirely destroyed. Assuming, for instance, that 1 milligramme of _Lachesis_-venom coagulates in two minutes 1 c.c. of 1 per cent. citrate rabbit-plasma, if we add to the plasma firstly 1 milligramme of _Ancistrodon_-, or 1 milligramme of _Cobra_-venom, and then 1 milligramme of _Lachesis_-venom, the plasma remains non-coagulated, yet coagulates perfectly on the subsequent addition of 1 c.c. of a ½ per cent. solution of chloride of calcium.

There is, therefore, a real antagonism between the actively coagulant substance contained in certain VIPERINE venoms and the anticoagulant substance comprised in the venoms of certain other VIPERIDÆ (_Ancistrodon_), belonging to the subfamily CROTALINÆ, and in those of all the COLUBRIDÆ.

The conclusion to be deduced from the foregoing facts is that the venoms of COLUBRIDÆ and those of certain VIPERIDÆ are decidedly _anticoagulant_, while the majority of the venoms of VIPERIDÆ, on the contrary, possess strong _coagulant_ properties, even when mixed with blood in infinitesimal doses.

The question therefore arises why these _coagulant_ VIPERINE venoms suppress the coagulability of the blood when mixed with it _in vitro_ in strong doses (for example, in doses beginning from 4 milligrammes of _Lachesis_-venom, or 7 milligrammes of the venom of _Vipera russellii_ for 1 c.c. of 1 per cent. citrate rabbit-plasma).

The explanation of this apparently contradictory phenomenon is furnished by the intense proteolysis that these VIPERINE venoms exert upon fibrin, in solution or coagulated. This proteolysis actually manifests itself with weak coagulant doses, for the compact clots formed at the outset soon become soft and then dissolve, like a cube of egg-albumen in an experiment in artificial digestion by trypsin. We shall revert to the subject later on.

III.--MECHANISM OF THE ANTICOAGULANT ACTION OF VENOMS ON THE BLOOD.

The anticoagulant action of the venoms of COLUBRIDÆ and of _Ancistrodon_ upon the blood appears to take effect in the first place upon the fibrin-ferment, and afterwards upon the fibrin by proteolysis. The action on the fibrin-ferment seems manifest when we experiment with anticoagulant venoms which are feebly proteolytic, like the venom of the _Cobra_.

I have already stated that a mixture of fresh blood with a sufficient dose of _Cobra_-venom is non-coagulable, as though the blood on issuing from the animal had been mixed with peptone or leech-extract. But, while blood when peptonised or mixed with leech-extract coagulates readily on the subsequent addition of fibrin-ferment, blood mixed with venom remains positively non-coagulable. It is the same with citrate- or oxalate-plasmas, which no longer coagulate when chloride of calcium is added to them, and with 4 per cent. saline plasma on the addition of distilled water.

The anticoagulant substance in the venoms of COLUBRIDÆ and _Ancistrodon_ is precipitable by alcohol, like the coagulant substance in the venoms of VIPERIDÆ and like the _neurotoxins_, from which it is difficult to separate them. The separation can nevertheless be effected by the aid of heat, if we make use of certain venoms that are particularly resistant to high temperatures, such as those of the _Cobra_ or the _Krait_. These latter venoms, when heated for one hour at 70° C., cease to be anticoagulant, and _preserve their toxicity unimpaired_. It is, however, impossible to suppress the toxicity without at the same time destroying the anticoagulant substance.

_Antivenomous serum_ completely protects citrate- or chloridate-plasmas against the anticoagulant action of venoms. It is sufficient to mix ½ c.c. of 4 per cent. saline antivenomous serum with 1 c.c. of 4 per cent. saline plasma to ensure that the subsequent addition of 1 milligramme of _Cobra_-venom to this mixture remains without effect upon the coagulability of the latter. If, after a contact of two hours or more, 2 c.c. of distilled water be added, coagulation is produced just as in saline plasma without venom.

B.--EFFECTS OF VENOM UPON THE RED CORPUSCLES.

(1) _Hæmolysis._--The hæmolytic properties of venoms, that is to say, their faculty of dissolving the red corpuscles, have been the subject of very important researches on the part of a number of investigators during the last few years (W. Stephens,[45] Flexner and Noguchi,[46] Calmette,[47] Phisalix,[48] Preston Kyes and Hans Sachs,[49] Noc[50]).

The different venoms are all hæmolytic, but in very variable doses. It is possible to make a very precise comparative study of them from this special point of view by taking as a base for each venom, as was done by Noc, the unital dose of 1 milligramme (or one-tenth of a cubic centimetre of a 1 per cent. solution freshly prepared and not filtered, the filtration through porcelain retaining an appreciable part of the active substance), and noting the time strictly necessary for this dose of 1 milligramme to dissolve completely, _in vitro_, 1 c.c. of a 5 per cent. dilution of red corpuscles of the horse in physiological saline solution.

It is very important, before allowing the venom to act on the red corpuscles, to first wash the latter by means of several successive centrifugings with 8 per 1,000 physiological saline solution.

It is also better to choose the corpuscles of the horse in preference to those of other species of animals, since they exhibit a nearly constant mean sensitivity. The corpuscles of the ox, goat, sheep, and rabbit are less sensitive. Those of man, the guinea-pig, and the rat, on the contrary, are more so.

On experimenting with _washed_ corpuscles, it is found that venom alone is incapable of dissolving them. In order that dissolution may take place, we are obliged to add to the mixture either a small quantity of normal horse-serum, preferably _heated_, and, consequently, deprived of alexin (Calmette), or ½ c.c. of a 1 in 10,000 solution of _lecithin_ in physiological saline water (P. Kyes).

Venom, therefore, is capable of hæmolysing red corpuscles only when it is _quickened_, either by heated normal serum, or by lecithin. The solution of lecithin employed for this purpose should be prepared by dissolving 1 gramme of lecithin in 100 grammes of pure methylic alcohol. Taking 1 c.c. of this dilution we add it to 9 c.c. of 8 in 1,000 saline solution, and make a second dilution of 1 c.c. of the foregoing mixture in 9 c.c. of saline water. This latter dilution of 1 in 10,000 is utilised as the reagent.

Let us now see how the serum or lecithin acts. It has been shown by P. Kyes that with either of these substances the mechanism of the hæmolytic action is the same, for the serum quickens the venom only through the agency of the free lecithin it contains. The lecithin takes part in the reaction by combining with the venom to form a hæmolysing _lecithide_ more resistant to heat than its two components, for it may be heated for several hours at a temperature of 100° C., without the loss of any of its properties.

When venom is brought into contact with certain kinds of highly sensitive red corpuscles, those of the rat for example, these corpuscles, although washed and freed from serum, may undergo hæmolysis. This result is due to the fact that these corpuscles contain sufficient quantities of lecithin, which becomes liberated from their protoplasm and, uniting with the venom, constitutes the active _lecithide_.

It was already known that lecithin is capable of combining with various albuminoid matters and with sugars to form _lecithides_. We must not, therefore, be surprised to find that such a combination may take place with the proteic substances in venom. The combination in this case is a truly chemical one. Lecithin in its natural state, or that which is normally found in serums which quicken venom, such as horse-serum, even when heated to 65° C., therefore plays the part of _complement_ according to Ehrlich’s theory, or that of _alexin_ according to the theory of Bordet, while venom itself would be an _amboceptor_ or _sensitiser_.

This is not, however, the way in which the phenomenon should be understood, for it is impossible to admit the identification of heated serum or lecithin with the complementary substances or alexins, seeing that the essential characteristic of the latter is that they are intolerant of heat and become entirely inactive on being raised to a temperature of 58° C., or even by simply being kept for a few days exposed to the air and light. We must therefore suppose, with P. Kyes and H. Sachs, that the red corpuscles themselves contain substances capable of playing the part of complements (_endo-complements_), and that it is with these that the venom combines when quickened by the presence of lecithin or heated serum, the latter only acting because it contains free lecithin.

All substances that contain lecithin, such as bile, hot milk, or cephalin, are capable of exerting the same quickening action, but do not themselves possess any inherent hæmolytic power.

_Cholesterin_, on the contrary, represents a kind of antidote to lecithin, as also to normal serums. It prevents hæmolysis of the red corpuscles in a mixture of washed corpuscles and venom, yet it does not in any way modify the properties of true alexins or complements.

Moreover, no correlation exists between _lecithides_ and the _neurotoxin_ in venoms. The combination lecithin + venom possesses _hæmolytic_ action, but is in no way _neurotoxic_. Conversely, venom can be freed from its groups of molecules combinable with lecithin, and remain _neurotoxic_.

_Lecithide_ is insoluble in ether and acetone, but soluble in chloroform, alcohol, toluene, and water. Its properties are therefore entirely distinct from those of its two components. It precipitates slowly from its aqueous solutions, without losing its hæmolytic power; it does not show _biuret_-reaction; it dissolves with equal readiness the red corpuscles of all species of animals, and its effects, like those of venom, are impeded by cholesterin.

Kyes has succeeded in obtaining lecithides with all the hæmolytic venoms that he was able to study: thus he has prepared lecithides from _Lachesis lanceolatus_, _Naja haje_, _Bungarus_, _Lachesis flavoviridis_, and _Crotalus_. It is therefore probable that the _lecithinophile_ group exists in all venoms, even when these differ as regards their other properties.

A wide range of difference is exhibited by the various venoms, as regards their hæmolysing power in the presence of normal heated serum or lecithin. The venom of _Naja_ and that of _Bungarus_ are the most active. The action of the venoms of VIPERIDÆ, and especially of those of _Crotalus_, is very weak. For example, while 1 milligramme of _Cobra_-venom dissolves in from five to ten minutes 1 c.c. of a 5 per cent. dilution of red corpuscles in the presence of lecithin or normal heated serum, the same dose of the venom of _Vipera russellii_ takes thirty minutes to effect the dissolution, and the venom of _Lachesis lanceolatus_ takes three hours.

P. Kyes and H. Sachs have discovered the apparently paradoxical fact that, if to the red corpuscles of certain species of animals _Cobra_-venom be added in increasing doses, hæmolysis augments up to a certain point, beyond which the destruction of the corpuscles shows progressive diminution. In a large dose _Cobra_-venom no longer produces any effect upon the corpuscles of the horse, for example, even when the venom is added in presence of a great excess of lecithin or heated serum. It would seem, then, that, according to the theory of Ehrlich, under the influence of an exaggerated amount of venom-amboceptor there is produced a deviation on the part of the complement (serum or lecithin), and that the latter, instead of fixing itself upon the corpuscles, becomes united with the surplus fraction of the amboceptors, which has remained free in the liquid.

Noguchi,[51] resuming the study of this extremely curious action of strong doses of venom, observed that the red corpuscles of certain species of animals (such as the horse for example), when previously washed and held in suspension in a physiological solution of sea-salt containing 4 per cent. of _Cobra_-venom, acquire a considerable augmentation of resisting power with regard to various physical and chemical agents. In consequence of this they are no longer hæmolysed by distilled water, ether, or saponin.

Nevertheless, acids or alkalies, except ammonia, destroy corpuscles treated with venom more easily than those in their normal condition.

If corpuscles, previously treated with a strong dose of venom, are subjected to repeated washings in physiological saline solution, the special resistance acquired by them in the presence of the venom disappears; they even become more sensitive to the action of destructive agents, such as water, ether, or saponin.

The principle contained in venom, to which must be attributed the protective action, is not destroyed by heating to 95° C., although at this temperature _Cobra_-venom becomes partially coagulated. Moreover, the protective substance is contained in the coagulum, while the _hæmolysin_ remains entirely in the filtrate. The agglutinin of venom, on the other hand, is destroyed at a temperature of 75° C. The protective substance, therefore, can be identified neither with the hæmolysin nor with the agglutinin.

It follows that it is impossible to accept the hypothesis of the “deviation of the complement” suggested by Kyes and Sachs to explain the innocuousness of strong doses of venom. Besides, it would be difficult to reconcile this hypothesis with the fact, observed by Noguchi, that venom in a strong dose protects corpuscles, not only against the action of lecithin (complement), but also against distilled water, ether, &c.

Noguchi, seeking more thoroughly to elucidate the mechanism of this protective action, finds that _Cobra_-venom forms a precipitate with blood-serum, when the latter is relatively poor in salts or when it is dilated with water. It likewise forms a precipitate with the aqueous extract of red corpuscles, and precipitates the globulins, hæmoglobin, or globin of the corpuscle, when treated separately. The precipitates are insoluble in water, but dissolve with the assistance of a small quantity of acid or alkali, and also in a great excess of saline solution.

Noguchi supposes that red corpuscles, when treated with strong solutions of venom, are protected against destructive agents on account of the formation by the venom and certain constituents of the corpuscle (chiefly hæmoglobin) of a compound insoluble in water. When this compound is removed by repeated washings in physiological solution, the corpuscles can easily be hæmolysed afresh by the ordinary destructive agents. Venom, none the less, exerts a noxious influence upon the corpuscles in all cases; but when strong solutions are employed, this effect is masked by the protective action.

All kinds of red blood corpuscles are not equally sensitive to the protective action of strong doses of venom. In this respect all degrees are observed in the action of venom. Thus the corpuscles of the dog are not protected at all by _Cobra_-venom. But it is interesting to observe that this venom in no way precipitates either the aqueous extract of dog’s corpuscles, or the hæmoglobin, or the globin of this animal.

The venom of _Crotalus_ and that of _Ancistrodon_ likewise possess protective power, which is, however, less marked than in the case of _Cobra_-venom.

Noguchi finally points out that corpuscles treated with venom are not hæmolysed by fluorescent substances such as eosin. They are also refractory to the hæmolysing action of tetanolysin.

The resistance of the hæmolysins of venom to heat (which, according to Morgenroth, may extend to heating for thirty minutes at a temperature of 100° C.) explains how it is that the serum of horses immunised by means of venoms heated to 72° C. is distinctly antihæmolysing, and capable of perfectly protecting the red corpuscles _in vitro_ and _in vivo_.

I have been able to prove that the _antineurotoxic_ property of antitoxic serums with regard to the venoms of COLUBRIDÆ is pretty much on a par with their antihæmolysing property, so that it is possible to measure _in vitro_ the antitoxic activity of a serum by establishing the degree of its antihæmolysing activity. Thus we see that a serum, which is antitoxic and antihæmolytic with respect to the venom of _Naja_, is likewise antihæmolytic as regards the other COLUBRINE-venoms, and even certain venoms of VIPERIDÆ. Here we have a very important fact, for it enables us to measure _in vitro_ the activity of antivenomous serums.

(2) _Precipitins of Venoms._--The serum of rabbits treated with increasing doses of _Cobra_-venom precipitates the latter in more or less concentrated solution. It has no effect as regards other venoms. On the other hand, the serum of a strongly immunised horse, the antivenomous power of which was pretty considerable, gave no precipitate with _Cobra_-venom; the formation of precipitate is therefore entirely independent of that of antitoxins (G. Lamb).[52]

(3) _Agglutinins of Venoms._--Besides their hæmolytic action, it is easy to observe that certain venoms, especially those of VIPERIDÆ, agglutinate the red corpuscles, and that the substance that produces this agglutination is different from the hæmolysing substance; for it acts rapidly at a temperature of O° C., at which hæmolysin manifests its effects only with extreme slowness. Heating to 75° C. destroys this agglutinant property (Flexner and Noguchi).

C.--EFFECTS OF VENOM UPON THE WHITE CORPUSCLES: LEUCOLYSIN.

The white corpuscles themselves do not escape the action of venom. It is possible _in vitro_ to prove this action upon leucocytic exudations obtained, _e.g._, by injecting sterilised cultures of _Bacillus megatherium_ into the pleura or peritoneum of the rabbit. After a few hours this exudation is removed by means of capillary tubes, and, on mixing these with weak doses of venom, we see, in the course of a microscopic examination, that the large mononuclear cells are the first to be dissolved, then the polynuclears, and lastly the lymphocytes. The leucolysis is much more intense and more rapid with _Cobra_-venom than with that of _Crotalus_ (Flexner and Noguchi, Noc).