value of these findings and assuming that, in tuberculosis and in typhoid, neutral sulfur is usually present in the urine, we examined the histories of all the cases in which positive tests were obtained by either method and thus eliminated these two diseases as the cause of the appearance of neutral sulfur in the urine, with the following results:

As this summary shows, of 122 cases in which the history of tuberculosis and typhoid were ruled out, 113 gave negative tests by both methods. The 4 cases in which both diazo and Weiss tests were positive were clinically cases of diabetes, cholecystitis, gangrene and pleural effusion. Whether in these conditions, the findings for neutral sulfur are constant we cannot tell, for we had only one case of each in this series.

Comparisons of urinary findings with the results of the serum test. Having thus ascertained the degree of efficiency of the tests as compared with each other, and having found that the tests are usually negative for cases where both typhoid and tuberculosis are excluded, we proceeded to compare the findings of the urinary examination with those of the serum test.

TABLE 2

Data pertaining to comparative urinary findings and results of the serum test

The serum test was that described in detail before. The anti

9

gen was that of Besredka. The total number of cases compared

It was previously shown by one of us1o that, in cases of certain tuberculosis, only 90-95 percent of the cases give positive serum reactions. It was also shown that, in the advanced cases, the reaction is usually absent, so that it was even suggested that a failure of the serum test in an advanced case of tuberculosis may be taken as a bad prognostic sign.

Comparing the earlier findings with those of the present series, we notice the same phenomenon, namely, in 8 out of 100 cases the serum reaction is negative in spite of the fact that the cases are undoubtedly tuberculous. The value attributed previously to these negative findings seems to be justified by the present results, in spite of the fact that the urinary findings appear to contradict the serological, inasmuch as out of 92 cases of positive serum results the urinary findings in 42 cases, at least, were negative by both methods, and in only 40 cases, or less than 50 percent, the urinary findings confirmed, by both methods, the serum findings. On the other hand, out of 8 cases of tuberculosis in which serum findings were negative, every case gave a positive Weiss test and 6 cases also gave positive diazo tests.

Remembering that the presence of neutral sulfur, according to Weiss, is due to the destruction of tissue, and that the intensity and frequency of the occurrence of the reaction run parallel with the progress of the disease, the findings above may be of great value in confirming the opinion, stated earlier, that cases of tuberculosis in which there is no circulating antibody are cases in which there is considerable destruction of tissues, as indicated by the excess of sulfur in the urine.

The finding of II positive serum reactions among the cases which do not present any symptoms of tuberculosis clinically, should not be attributed, as was explained before, 10 to non-specificity of serum diagnosis, but rather to the fact that in its earlier stages, the tuberculous process does not induce symptoms enough for clinical diagnosis. A positive serum test in such cases may indicate its extreme diagnostic value. Although urinary findings in these II cases were all negative, with the exception of one, which was a case of typhoid, they indirectly explain the failure of clinicians promptly to discover the tuberculous process, since in incipient cases the destruction is so insignificant, that no increase of sulfur can be detected.11

SUMMARY. Comparisons of the diagnostic values of the urinary findings for neutral sulfur with those for the serum reaction in tuberculosis reveal the following facts:

1. The diazo or Weiss test in tuberculosis is less constant, in general, than the serum reaction.

2. Positive results with the diazo or the Weiss test are of value only if typhoid is excluded. There are also a few other pathological conditions in which these tests are positive, but the data at hand are inadequate for conclusions regarding the constancy of these findings.

3. The urinary findings are not sufficiently frequent in tuberculosis to be of special diagnostic value, even when other possible complications giving rise to positive tests can be excluded.

4. The occurrence of increased amounts of urinary neutral sulfur, in advanced stages of disease, is quite constant and may be of prognostic value, especially in connection with corresponding negative findings in the serum.

ABDERHALDEN TEST*

J. BRONFENBRENNER, W. J. MITCHELL, JR., AND PAUL TITUS (Pathological and Research Laboratories of the Western Pennsylvania Hospital, Pittsburgh, Pa.)

In previous communications (1, 2) we outlined the mechanism of the Abderhalden test as an autodigestion of serum protein of the patient due to the removal of antitryptic inhibition. This removal of anti-trypsin, although quite apparent in our experiments, has not thus far been demonstrated directly. In this preliminary report we wish to record the fact that actual measurement of the anti-trypsin in the serum, before and during the progress of the Abderhalden reaction, reveals the fact that the anti-tryptic titer of the serum is actually involved. The diminution of the anti-tryptic activity of the serum, as tested against trypsin solution, takes place in a specific manner, inasmuch as it occurs only in cases where the serum used is that of pregnant individuals and is parallel with the intensity of the Abderhalden test; so that the estimation of anti-trypsin in serum undergoing digestion, after its removal from contact with placenta, may be used as a method of diagnosis of pregnancy parallel with, and complementary to, that of the Abderhalden test. Moreover, it is evident that this inactivation of anti-trypsin takes place at ice-box temperature as well as at the temperature of the incubator. If the Abderhalden test is divided into two periods, as was shown before (3), over 30 percent of the anti-trypsin is removed during the first part of the reaction.

The comparison of the anti-tryptic index of the serum, before and during the Abderhalden test, with the index obtained by measuring the effect of kaolin and other substances capable of adsorbing anti-trypsin in a non-specific manner, confirms our contention (1) that the appearance of dialyzable cleavage products in serum may be determined by specific as well as by non-specific mechanisms, and that the essential part of this phenomenon is the removal of serum anti-trypsin, which in turn liberates the normal proteases of the serum, thus setting the serum into autodigestion.

ON THE NATURE OF THE ABDERHALDEN
REACTION1

J. BRONFENBRENNER

According to Ehrlich's theory the parenteral introduction of foreign protein causes the cells of the body to produce an excess of specific receptors, which, at a certain period of the process, circulate freely in the blood, and are known to the student of immunity under the name of amboceptors, antibodies, or substances sensibilisatrices. These specific antibodies are complex in character; and, although they are directly responsible for the specificity of the protective processes in the body, they are not of themselves active principles. It is to complement that Ehrlich and his school attribute the power of action on antigen.

The properties of the antibodies resemble those of enzymes in very few respects, while they differ from them at many points. According to Abderhalden, however, the parenteral introduction of foreign protein results in the production of specific enzymes capable of directly digesting antigen in vitro. I disagree with those who think that Abderhalden has proved that these substances are enzymic in character. On the one hand it is difficult to believe that the organism is able to supply so many specific enzymes; on the other, it is improbable that the enzymes circulating in the blood are strong enough to digest coagulated protein, as is the case in the Abderhalden test.

In our own experiments we tried to produce a specific enzyme by repeated inoculation of rabbits with egg-white; and, although the serum of these animals contained a very large amount of antibodies, a Mett-tube filled with coagulated egg-white failed to show even the slightest traces of digestion of the egg-white after suitable immersion in such serum. This and the results of other experiments led us to conclude that either the enzymes on which the supposed di