pendent solely, as was Jenner, upon crude, fortuitous, epidemiotherapeutic statistics for proof of the efficacy of the method, because there have been established precise indices of immunity by which can be determined not only the positivity or negativity of the results of treatment, but accurate quantitative calculations as well. These indices are, with reference to the measurement of the antitoxin content of the blood, the Römer 21 and the Schick 17, reactions.

Criteria of Efficiency.

24

Römer's method as applied to the determination of the amount of antitoxin in the blood, is that of injecting guinea-pigs intracutaneously with 0.1 ccm. of different dilutions of the unknown serum mixed with a certain amount of toxin. This amount of toxin is such as will, when injected in the same manner with a definite fraction of a unit of standard antitoxin, just produce noticeable necrosis of the superficial layer of skin by the fourth or fifth day. (Limes-necrosis dose.) From the grades of reaction produced by the mixture of this known amount of toxin with the unknown serum in graded dilutions, the amount of antitoxin in the latter can readily be calculated. Römer claims that unit can be recognized in this way. Park, on the other hand, states that it is impossible to detect less than 1 unit. However, for practical purposes it is rarely necessary to test for less than 1 unit, hence the method furnishes an extremely valuable criterion as regards the control of prophylactic treatment for diphtheria.

The Schick test, somewhat similar, consists in the intracutaneous injection, into the human subject, of of the minimal fatal dose of pure toxin contained in a volume of o.1 ccm. A positive reaction-redness and slight infiltration at the site of injection, appearing in 24 to 48 hours and leading to a fading brownish pigmentation, is a manifestation of the local irritative property of the toxin, and indicates the presence either of no antitoxin or an amount less than unit per ccm. of blood. A negative reaction is generally accepted to mean that there is over a unit present and that the subject is immune to diphtheria. This test as used is not so delicately quantitative 15 as Römer's but is of value in eliminating the necessity of immunizing a large percentage of those exposed to infection. Early researches showed that antitoxin production reached a protective height in about 21 days. With modification in dosage and methods of administration that height may sometimes be attained in 8 or 10 days; 14 days may be taken as a fair average.

It would naturally be expected that the amount of antitoxin in an individual is the factor which conditions his resistivity or his

susceptibility to infection with diphtheria. All the evidence so far collated substantiates the validity of the assumption that a moderate production of antitoxin protects against infection. The amount produced in consequence of injection of the vaccine seems to parallel the resisting powers and to decline as they run out. Direct proof of this assumption can only be secured by a comparison of the numbers of immunized and non-immunized subjects who contract the disease during a period of time considerably longer than has elapsed since the introduction of the vaccine. Indirect evidence, however, is derived from the observations of Hahn, Otto and others, that persons who contract the disease possess insignificant amounts of natural antitoxin; this fact likewise holds true in those who sustain repeated attacks.

In considering the question as to how much autogenous antitoxin is sufficient to protect against an ordinary infection with diphtheria, it may be stated at the outset that it is wholly unnecessary to force the immunization to the individual's limit of productivity of antitoxin. None of the investigators urges the need of accumulating more than 0.1 unit per ccm. of blood; many agree that 0.05 unit is enough, while v. Behring considers that even 0.01 unit is sufficient to give protection under ordinary circumstances of health. He reaches this conclusion from results obtained by passive immunization with antitoxin. Children of 25 kilos who receive 100 units show an immediate maximum titre of about 0.1 unit per ccm. of blood; this falls in ten days to slightly less than o.or unit, and the immunity ceases at about this time. Kleinschmidt and Viereck 12 report never having seen diphtheria develop in exposed subjects. possessing an antitoxin titre of 0.05 unit or more, and their estimate that this amount, which is almost always secured by two prophylactic injections, is probably the safest one to follow. Massive virulent infections, occurring when susceptibility is possibly heightened by other diseases, may break down this resistance. v. Behring claims that fever causes increased destruction and elimination of antitoxin. Park in reporting his series of scarlet fever patients makes no mention of any cases of diphtheria occurring subsequent to successful active immunization, although he does state that 20 per cent. of those in which less than 0.033 unit was formed -Schick reaction positive-did contract the disease. In Kissling's scarlet fever series five cases developed among the 89 who received one injection, and no case in the 109 who received two doses. Neither did diphtheria develop in any of the 32 immunized cases suffering with measles; this fact tends to substantiate Karasawa's and Schick's findings that the antitoxin content remains unchanged throughout the course of measles.

Dosage and Reaction.

The matter of dosage is also in a somewhat unsettled condition because different sized doses as well as different methods, and intervals of administration have been employed by the various investigators with the result that exact comparableness of findings is not yet possible. v. Behring insists that susceptibility is not a haphazard matter of individual response but is subject to laws almost mathematical in their precision. A certain dose of diphtheria toxin cails forth certain phenomena in a normal guinea-pig of a certain weight; if a tumor or pregnancy increases the bulk of the animal, in proportion to this increased weight it becomes less susceptible to the toxin. Although weight is only one of a complexity of factors, he is confident that we may yet discover a similar normal standard of susceptibility and thus be able accurately to apply the necessary dosage.

5

Certain experimental observations made by Südmersen and Glenny, 28 Brown and others,18 upon the immunizing effect on guinea-pigs of small doses of diphtheria toxin should be taken under advisement in this connection. They were able to demonstrate that the use of doses which produce profound general and local reactions, or by injecting before the pig recovers from the previous dose of toxin, there results serious interference with the progress of immunization. Contrarily, when perfect toleration is secured by the use of small enough doses, the formation of antitoxin proceeds without hindrance. Löwenstein,16 from a long and varied series of experiments upon rabbits, concludes that under-neutralized mixtures are no better than neutralized mixtures and that it takes longer to raise the resistance when a vaccine containing an excess of toxin is used.

Upon the size of the dose depends the intensity of the reaction. An examination of Smith's protocols of guinea-pig experiments does not reveal the existence of any direct relation between the reactions or the amount of free toxin in the mixture and the degree of passive immunity which was transmitted to the progeny; in the human subject, however, antitoxin production does seem to vary directly with the severity of the reaction. The reaction varies not only with the size of the dose but also with the route of administration and with the degree of previous sensitization toward the products of diphtheria bacilli. In equal dosage the intracutaneous injection gives rise to a more intense reaction than does the subcutaneous. Kissling, Hahn and Sommer, Kleinschmidt and Viereck, and Schreiber recommend the intracutaneous route exclusively, while v. Behring and Bauer favor the subcutis. Nearly every experimenter remarks that the amount of antitoxin produced seems to be directly proportionate to the intensity of the local reaction.

9

This is probably true in most cases and especially in those persons who already have some natural antitoxin but there are a few exceptions. Kassowitz and Schick studied the comparative effects of different mixtures of toxin-antitoxin when injected intracutaneously in guinea-pigs and in human beings. Their findings were correspondent in all but 10 per cent. in which there was discovered a hypersensitiveness to the mixture on the part of older children, especially young adult females. They suggest that this susceptibility may be due, first to an idiosyncrasy to a protein substance in the toxin-a substance which operates even after the total inactivation of the toxin by boiling; or second, the dissociation of the mixture within the tissues, or third, sensitiveness to horse-serum as such. v. Behring makes the more practical observation that a non-specific hyper susceptibility is frequently exhibited by tuberculous patients and leads to an intense local reaction and swelling of the proximate lymph-glands. This condition merits careful attention in reference to prophylactic vaccination against diphtheria because a positive reaction in these cases is not necessarily a sign of antibody production in adequate amounts for protection. Hence, for the purpose of a statistical investigation designed to show the protective action of the vaccine, Behring prefers to have the lymphatic constitution regarded as a contraindication for his method. He also would leave out results on atrophic children and infants under nine months old. We may conclude that the intensity of the local reaction is not in all cases an exact criterion of antitoxin production, but that with certain precautions it may serve as a practical guide. Particularly is it of value in subjects who already possess a small amount of antitoxin; in these, one reaction of the second grade almost invariably betokens the elaboration of a sufficient amount of antitoxin to protect again an ordinary epidemic infection.

Physicians and nurses who have much to do with diphtheria are usually found to possess exceptionally high antitoxin titres. So far as we know, antitoxin develops only in response to diphtheric infection. Otto 19 has demonstrated that a single attack of diphtheria does not cause the production of a large amount of antitoxin and he postulates that when there is no history of diphtheria the immune must have been subjected to repeated minute. infections which have passed over with minimal or misinterpreted symptoms. This hypothesis is borne out by the high antitoxin coefficients-even 2 or 3 units per ccm.-found in sera at the time and after the bacilli have been found in carriers. Those who, without having had the disease, have been in close contact with it, usually show more than .or unit: of six nurses who possessed less than that titre, two were attacked soon after being put on duty in

diphtheria wards. Kissling reports an instance of one interne, who had only .005 unit at first, but after two months' service on the ward had .or unit, and one treatment with the vaccine raised this amount to .05 unit. Kissling himself, who came less intimately and for shorter periods in contact with the cases, doubled his titre in less than two months without recourse to the vaccine. The development of immunity is of course much safer and more exact by the use of the T.-A. vaccine than by mere exposure to infection.

It is these subjects with natural antitoxin in which the vaccine. produces the stronger reaction both as regards the local effect and the formation of more antitoxin. This is explicable on the ground that the process of active immunization leads to a changed sensitiveness to the immunifying agent. Curiously enough this sensitiveness is concealed when pure toxin is injected but manifests itself upon the introduction of T.-A. mixtures.

The earlier workers followed no definite plan in the matter of intervals between doses. v. Behring has recently favored the giving of the second dose only after ten to fourteen days, since experience has shown that sensitization usually reaches an effective height only after this period. If the subject has been perviously sensitized he thinks that one injection is sufficient. As one means of determining previous sensitization the intracutaneous injection offers a safe and fairly reliable index. The reaction here is more easily judged than in the subcutaneous method where the thickness and condition of the skin and subcutis oppose obstacles to correct interpretations. Which method is the more effectual for immunizing purposes has not yet been decided. If rapidity of absorption and slight reaction are the important factors, the subcutaneous route would seem preferable. On the other hand, if antitoxin formation. is dependent more upon gradual dissemination of the mixture and the intensity of the local reaction, the intradermal method should be the one of election.

The grades of intracutaneous reactions have been formulated by Kleinschmidt and Viereck as follows:

1. Redness and infiltration i cm. to 2 cm. in diameter, lasting 2 or 3 days, with slight discoloration for a few days more.

More extensive redness and infiltration, over 2 cm. in diameter, lasting several days, moderate pain on pressure and subsequent mild scaling of the skin.

3. As above with painful, swollen lymph-glands.

4. Mild febrile phenomena with headache, anorexia and lassitude. Reactions of the third and fourth degrees are regarded as undesirable and may be avoided by the employment of a small intracutaneous trial test as mentioned above. The second grade