microscope. When excessive dilutions are employed, the reaction may escape macroscopic observation; whereas, under the microscope, after a time (from one minute to two hours), agglutination may yet be observed.

In addition to typhoid-cultures, cholera-cultures, and coli-cultures, the reaction of Gruber has been demonstrated with numerous other, and also nonpathogenic, bacteria. The agglutinating property of the serum is, further, not present immediately after introduction of the bacteria into. the animal body; at least three and a half, and generally even five, days must elapse before this appears. Besides the blood-serum, the reaction of Gruber may be obtained in an intense degree with the serous contents of blisters induced by vesication, in lesser degree with milk, and in still less degree with urine, dropsical fluid, exudates, bile, tears, and aqueous humor of the respective animals and human beings.

Gruber considered the phenomenon of agglutination as a reaction of immunity, and attempted to base upon it a new theory of immunity, to which we shall later refer. A not inconsiderable advance was made in this connection when Widal showed that agglutination, at least in human beings, represents a reaction of the period of infection. Widal demonstrated that the serum of typhoid-fever patients at the end of the first or at the beginning of the second week of the disease most distinctly yields the agglutination-phenomenon with typhoid-bacilli (Widal-Gruber reaction). This fact is of the greatest importance from a practical point of view. It renders possible the diagnosis of typhoid fever with the aid of the serum of a suspected patient. The same appears to be true for Asiatic cholera and other disWe shall fully discuss the so-called serum-diagnosis in the special section. (See Typhoid Fever.)

eases.

The nature of the agglutinating substance has not yet been determined despite numerous investigations, especially on the part of Widal. It appears to be quite resistant, as exposure of the serum to a temperature of 60° C. (140° F.) for an hour fails to abolish the agglutinating phenomenon; this result is obtained only after exposure to a temperature of 80° C. (176° F.). The reaction of Gruber may be obtained also with dead bacilli. The bacteria are best destroyed with formol, and they can then be kept for weeks without losing in the slightest degree their sensitiveness to the action of the serum. From this Widal arrives at the

conclusion that the agglutination is not a manifestation of the vital activity of the bacilli, but rather that it represents a passive reaction on the part of the protoplasmic substance. According to a more recent statement by Kraus, a mixture of immune serum with filtered cultures, after exposure for twenty-four hours to a temperature of 37° C. (98.6° F.), exhibits a precipitate, and sometimes even the formation of flocculi. According to Gruber, the agglutinating substances are derived from the bodily constituents of the bacteria. This statement, however, can not be sustained, as agglutinating properties may be observed in the serum of animals after the injection of soluble metabolic products, of filtered, quite young bouillon-cultures.

We have already stated that, as the result of his investigations of agglutination, Gruber attempted to establish a new theory of immunity. This assumes that the agglutinating substances cause the bacterial membrane to swell,' and thus render the bacterial protoplasm accessible to the alexins of Buchner, in consequence of which death of the microorganisms is brought about. This hypothesis of Gruber, however, is not supported by the facts elicited by further investigation. There need be no relation between agglutination and immunity. There are cases in which, in spite of the agglutinating property of the serum, immunity does not exist, and vice versa. Further, the assumed swelling of the bacilli can not be demonstrated microscopically.

4. ANTITOXINS.-The anti-substances of the bodily fluids, and especially of the blood-serum, thus far described, manifest their activity, on the whole, directly against the bacteria themselves. The conditions are quite otherwise with that class of anti-substances that are the last to be discussed, and are, probably, the most important. We have reference to the substances designated antitoxins, which derive their name from the fact that their energy is directed not so much against the microorganisms, but rather against the metabolic products (toxins) generated by them. Their discovery in the case of diphtheria and of tetanus is the fundamental work of Behring, and the stimulus to their further investigation was given especially by Ehrlich. The latter found in ricin and in abrin-two albuminoid vegetable poisons-substances presenting numerous points of resemblance to bacterial poisons, and with which some of the laws of immunity can be readily studied. As a result of his

experiments with these substances, Ehrlich recognized that in the process of immunization through the activity of the disease-poisons in the body of the patient, substances are formed that he designates anti-bodies (antitoxins). These are, in a certain sense, antidotes, inasmuch as they neutralize or prevent the toxic action of the disease-poisons. When these antitoxins are present in sufficient amount, immunity exists. In the process of immunization with the aid of bacteria or their poisons (direct, active, immediate immunization), the antitoxins are formed either from the bacterial products themselves, or, under their influence, from substances that exist in the body preformed.

The disease continues until an adequate amount of antitoxin has been formed. For this reason the older methods of immunization are effective only after disease of greater or less severity, and after the lapse of a certain time. Immunization by means of serum (passive, mediate, indirect immunization) represents the transmission of preformed antitoxins. Therefore, this method, on the one hand, induces no disease, and, on the other hand, it establishes immunity at once; and this is, for the same reason, also, more transitory, lasting only a few weeks.

Whence the antitoxins are derived has not yet been finally determined. The view that the antitoxin is formed directly from the toxin introduced for purposes of immunization appears gradually to be losing ground. It is more probable that in the course of every toxic disease the antitoxin is produced together with the toxin within the body. The antitoxic activity of the serum is abolished by exposure to a temperature of from 60° (140° F.) to 70° C. (158° F.). The manner in which antitoxins act upon the bacterial poisons has not yet been clearly determined. Originally, it was assumed that the antitoxins destroy the bacterial poiThe injection of a mixture of an antitoxin-containing serum and bacterial poison proved innocuous. From this it was concluded that the poison is destroyed by the antitoxin of the serum. It soon transpired, however, that destruction of the poison by the antitoxin does not take place, but that, to use the expression of Ehrlich, in the physiologically neutral mixtures of toxin and antitoxin both sets of constituents are yet present as such. Buchner and Roux assumed an action of the antitoxin upon the cells, as a result of which the latter are rendered immune to the in

sons.

toxication. In contradistinction from this cellular hypothesis is the chemic view of Behring and Ehrlich: that toxin and antitoxin undergo a sort of double combination, which proves innocuous for the tissues. The decision of this question appears finally to have been made in favor of the chemic theory by the more recent investigations of Ehrlich with regard to ricin and antiricin. Ricin has the property

of causing the red blood-corpuscles in defibrinated blood to collect together and to be precipitated to the bottom of the vessel-a process from which vital processes may with certainty be excluded. Ehrlich showed, then, that antiricin, which is present in the blood-serum of animals immunized to ricin, abolishes the activity of ricin in the test-tube, and that the peculiar coagulation following the addition of ricinserum no longer takes place. Ricin and antiricin must, in this instance, have directly influenced each other chemically. Ehrlich was further able to demonstrate that the combination of toxin and antitoxin takes place much more quickly in concentrated than in dilute solutions, that heat hastens and cold retards its occurrence. As similar manifestations are observed in chemistry in the formation of double salts, it would seem probable, according to Ehrlich, that also the neutralization of toxins by antitoxins represents the formation of a double salt.

The immunization of mammals against toxins is always attended with febrile reaction, and it, therefore, appeared that the formation of an antitoxin would not be possible in the absence of fever. Metschnikoff, however, found that of all animals the crocodile produces antitoxin most abundantly and most speedily, in spite of the fact that febrile movement does not take place.

According to the investigations of Behring and his collaborators, the antitoxin distributes itself throughout the organism in man and in animals in such a manner that after absorption of the injected serum, after passive immunization, the blood to a certain extent extracts the antitoxin from the tissues and stores it up. Twenty-four hours after subcutaneous injection of serum, and in a shorter time after intravenous or intraperitoneal injection, the maximum amount of antitoxin in the blood is demonstrable. The antitoxin is absorbed from stomach and bowel only when lesions of the mucous membrane exist. The maximum content of antitoxin in the blood persists for several days.

After this, the amount of antitoxin in the blood gradually diminishes. It now appears in the milk, in the urine, etc., until finally it is wholly swept out of the body. The rapidity with which this elimination of antitoxin takes place is most variable, in accordance with the different conditions present. It is the greater the larger the amount of immunizing serum injected. In the case of diphtheria, the protection in human beings following the usual immunization with 250 antitoxin normal units lasts about four weeks.

Success in the preparation of antitoxic serum may be hoped for only when the poison, the toxin of the respective species of bacteria, is known, and can be prepared of sufficient strength. It has thus far been possible to obtain a high degree of toxin-immunity only in the case of diphtheria, tetanus, botulism, snake-venom, ricin, and abrin.

It has already been mentioned that the lysogenic and the agglutinating substances of the serum are also present in the blood of normal individuals. The same statement applies also to the antitoxins. Attention has been called by a number of observers to a certain neutralizing activity on the part of the normal blood-serum of horses and of human beings against the poison of diphtheria and other similar poisons. The natural antitoxic power of the bloodserum is, however, only slight. It by no means attains the high degree of activity exhibited by the serum of animals artificially made poison-proof.

From the fact that immunity can be transmitted by means of serum containing antitoxin, the doctrine that protection against toxins is the cause of immunity has been brought forward. Behring, and also Ehrlich, ascribe the real cause of acquired immunity to the antitoxic property of the blood. However attractive this theory may be, all of the existing facts can not be brought in harmony with it. Reference may be made to the instances of a want of relation between the occurrence of antitoxin in the blood (immunizing capability of the blood-serum) and the presence of immunity (p. 55). It was mentioned that the occurrence of antitoxin in the blood of animals by no means establishes necessarily a condition of immunity in the latter, and that, under certain circumstances, the or ganism whose blood-serum possesses pronounced immunizing capability, exhibits not only not an increased, · but even a diminished, degree of resistance to the bacterial