milk small rod forms resembling Bact, abortus more or less in morphological, cultural, and biochemical characteristics, but generally lacking in pathogenicity. By inoculating guinea pigs with samples of milk, Schroeder and Cotton had shown in 1911 that virulent Bact. abortus is present in about 11 per cent of samples of cow's milk. Similar results have been obtained by several other investigators in various parts of this country. In spite of their prevalence in cow's milk, however, virulent strains could not be readily demonstrated by plating methods. I never succeeded in obtaining them by the plate imethod from the milk of spontaneously infected cows, but did obtain them from the milk of two experimentally infected animals. Later, Huddleson succeeded in plating Bact, abortus from the milk of spontaneously infected cows. He agrees, however, that Bact. abortus exists in cow's milk only in small numbers. This point is of some significance, and will be referred to again.

The finding of abortus-like organisms commonly in cow's milk led me to consult the investigators of the Pathological Division of the Bureau of Animal Industry, who had had clinical experience with Bact. abortus. During a conversation with Doctor Eichhorn, then chief of the Pathological Division, he asked me if I had ever compared abortus with the organism causing Malta fever. He did not say why he asked that question, but since he had had experience with Malta fever in goats, as well as with contagious abortion in cattle, he had presumably noted the similarity of the clinical aspects of the disease as it existed in the two species of milk animals.

Following Doctor Eichhorn's suggestion, I secured a half dozen available strains of melitensis and compared them with bovine strains, with the astonishing result that the bovine and human strains were indistinguishable morphologically and culturally: they gave the same biochemical reactions in various media: inoculation into guinea pigs gave identical results, and the serum of the animals inoculated with either, agglutinated as well with the other as with the homologous type of organism. Only by agglutinin absorption tests could the strains be differentiated.

These observations have been confirmed by Meyer and his associates in this country; by Zeller, Jaffé, and Skarić, in Europe; by Khaled, in Egypt; and by Burnet, in Tunisia. Some of these investigators have made further observations which have added substantially to the evidence of the close relationship of the organisms from the two sources.

Fleischner and Meyer in California, found that the intradermal test, which they regard as highly specific, can not differentiate between an abortus and a melitensis infection in guinea pigs. Burnet, in Tunisia, found that human Malta fever patients show the same cutaneous hypersensitiveness whether melitensis or abortus is used as the antigen. Khaled immunized a monkey by treating with abortus, and found a mild reaction to an infecting dose of melitensis. as compared with an untreated control monkey. Several investigators have compared the disease in guinea pigs produced by experimental inoculation with abortus and melitensis. They all agree that the anatomic and histologic changes are so nearly alike for the two infections that they are indistinguishable.

In a recent experiment I was able to show that inoculation of cows with strains from human cases of Malta fever results in abortion.

Within recent years attention has been directed to contagious abortion in swine, a disease which is of increasing economic importance in this country. The causal organism has been found to be the same as that responsible for the disease in cows. The organisms of the melitensis-abortus group are therefore responsible for Malta fever in man and in goats; and for contagious abortion in cattle and swine and other animals. It has also been noted that abortion is one of the symptoms of the infection in goats.

It is important that a comparative study be made of a large number of strains from every species of animal in which the disease spontaneously occurs. During the present year I have collected a large number of strains of the abortus-melitensis group from human, bovine, porcine, caprine, and equine sources, and classified them according to the agglutinin-absorption test, with Feusier and Meyer's three type strains as the starting point for the classification. The strains came from various parts of the United States and Europe. They were found to fall into seven distinct types. Five of these types included only one, two, or three strains each, and they will be disregarded in this discussion. The great majority of strains fell into two large groups. One of these groups (or types) includes the majority of bovine strains and the majority of porcine strains, but it also includes two human strains. The other of the chief serological types includes the majority of human strains; it also includes one bovine strain and the only caprine strain and the only equine strain in the collection.

Summarizing the serological investigation it may be stated that the majority of strains of the melitensis-abortus group isolated from five species of animals fall into two distinct serological types. These two types are more closely related serologically than types I and II meningococcus or types I and II pneumococcus for the melitensisabortus types can not be distinguished by the simple agglutination test. One type, which may be called the abortus type, is characteristically bovine and porcine; the other, which may be called the melitensis type, is characteristically human and caprine. The types are not restricted, however, to the host species for which they are characteristic.

The accumulating evidence of the close relationship between the strains of bovine and human origin has culminated in the conclusion of Burnet, that melitensis and abortus are not distinct bacteriological species, but merely distinct serological types of the same species. I concur in that conclusion.

Considering the array of facts which has established the close relationship between the causal organisms of Malta fever and of contagious abortion in cattle, the question arises with increasing emphasis why no recognized disease resembling Malta fever results from drinking raw cow's milk, for there is no doubt that a large part of our population has at some time or other ingested virulent abortus organisms.

As in every case of infection or immunity, there are three factors concerned the virulence of the microorganism, the number of bacteria which gain entrance to the tissues, and the resistance of the host. There is at. hand a certain amount of information, though meager, relating to these factors as applied to melitensis and abortus. The British investigators found by the plate culture method that about 10 per cent of goats on the Island of Malta were excreting melitensis in their milk. They found commonly 30.000 organisms per cubic centimeter, and they state that frequently a figure of 3,000,000 per cubic centimeter would be more nearly the correct number. In contrast to these figures for goat's milk, only two or three investigators have reported success in obtaining abortus from cow's milk by the plating method, and they agree that there are few of these organisms present in the milk of infected cows. It is only by inoculation of milk into guinea pigs which are susceptible to infection by a very few organisms, that the prevalence of abortus in cow's milk has been established.

The common use of goat's milk in central Europe without the prevalence of Malta fever has already been mentioned. It suggests a possible regional susceptibility to the disease, or a possible increase in the virulence of melitensis in subtropical regions.

The comparative virulence of bovine and human strains has been tested on monkeys by Meyer and his associates in California. In their feeding experiments they found that human strains were far more invasive, causing infection of the monkeys when small doses were used, as compared with the doses of abortus necessary to cause infection. These results, and the fact that there is no recognized acute disease in man caused by the ingestion of cow's milk infected with this species of microorganism, suggest that passage through generation after generation of cattle may have reduced the virulence of these organisms for man and monkeys.

It must be conceded that, in general, the bovine abortus does not cause in man such an acute disease as the Malta fever of subtropical regions. It would be a careless indifference to the prevalence of the organism, however, to conclude that it is harmless in temperate climates where cow's milk is more generally used.

In subtropical countries Malta fever manifests itself by such a great variety of symptoms that medical authorities are agreed that the only possible way to diagnose it with certainty is by laboratory methods, i. e., by the cultivation of the microorganism, or by serological tests. The disease also varies from the acute form, which occasionally ends fatally, to a mild, ambulatory type. The British commission found native Maltese "carriers" of the organism who asserted that they had never had Malta fever. It would appear to be a problem worthy of investigation to determine whether abortus infection might not be responsible for mild disabilities in our own localities for the cause is never determined of many cases of low fever, or bronchial cough, or rheumatism, or neuritis, or other mild symptoms of disease, such as are common in Malta fever.

During the present year I have made serological tests for melitensis or abortus infection with 500 samples of human serums. Most of these serums were sent to the Hygienic Laboratory from veterans' hospitals to be tested for the Wassermann reaction. A part of them.

were taken in the naval hospital at Washington, D. C., for the same purpose. The serums were therefore taken from people who were suffering with an ailment of some kind or other so that they had presented themselves at a hospital for diagnosis. Of the 500 serums, 57, or 11.4 per cent, gave a definitely positive agglutinin reaction. In seeking an interpretation for these results the opinion of investigators in countries where Malta fever is endemic must be taken as a guide. We have also the data showing the titer of agglutinins in 9 undoubted human cases of Malta fever contracted in Arizona in 1922 or 1923.

Fici, an Italian, has recently summarized all of the literature on the specificity of the Malta fever agglutinin reaction with the conclusion that it is a specific diagnostic test.

Nicolle believes that a positive reaction in a 1:10 dilution indicates a melitensis infection.

Bassett-Smith, who has had a wide experience with Malta fever in Mediterranean countries, is of the opinion that a positive agglutination in the 1:30 dilution of serum may be considered conclusive evidence of Malta fever, past or present. He adds that the serum of several of his patients who were undoubtedly suffering with Malta fever, as proved by the cultivation of the organism from the blood, gave a positive agglutination reaction in dilutions no higher

than 1:10.

Of the nine serums from Malta fever patients who recently contracted the disease in Arizona, one had an agglutinin titer of 1:20. From this figure the titers varied to as high as 1:640 in one case.

Taking into consideration Bassett-Smith's and Nicolle's results and our own, it appeared that serums of an agglutination titer of 1:20 or higher should be investigated. There were five serums among the five hundred with a titer of 1:20 or higher. None of the patients. from whom these five serums were taken had had any chance of infection from goats so far as they were aware. Agglutinin absorption tests with the serums showed that in at least two of the five cases the infection had been with the abortus type. In the other three cases the type of infection could not be determined by a study of the serum. The laboratory findings for these five patients would have led to a diagnosis of Malta fever if they had lived in regions where the disease was known to be endemic.

The significance of the agglutinin reactions in dilutions of 1:5 or 1:10 in the remaining 52 of the 500 serums is not clear. They may indicate a present infection; or they may be the result of a former infection; or they may be the result of ingestion of virulent organisms which the host disposed of without becoming diseased. There has been considerable written during recent years about vaccination by mouth, and Nicolle and Conseil have demonstrated that it can be accomplished with killed cultures of melitensis.

Several investigators in this country have previously reported that a considerable percentage of human serums gives a serological reaction with abortus antigen. The explanation has been either that the reaction was probably not specific, or that the individuals whose serum gave a positive reaction had absorbed the agglutinins from the cow's milk which they had drunk. These explanations are not satisfactory in the face of known facts. The exact significance of

77612-24-37

positive serological reactions of the melitensis-abortus organisms with human serums in temperate regions offers a fertile field for investigation.

If there were no other reason for milk Pasteurization, it would appear to be folly to drink raw milk containing the abortus organism With this in view leaders in the dairy industry should anticipate public health authorities in urging with renewed emphasis the general use of heated milk in country districts.

Dairymen who live on farms where contagious abortion is prevalent can aid the investigations now being carried on at the Hygienic Laboratory if they will note coincidences of illness in their families with the disease in cattle, and in case of such coincidence have their physician take a sample of serum to be sent to the Hygienic Laboratory, Washington, D. C. A report of the results obtained would always be made to the sender. This would be valuable information to the sender in case the results were positive; and a study of a few hundred such samples would yield important data obtained from a group of people among whom this disease is most apt to exist.

Chairman WELD. We have been obliged to pass over several papers this morning in the absence of the authors. Is it your desire that the papers be read by title?

Mr. ERNEST KELLY (United States Department of Agriculture, Washington, D. C.). I move that the papers referred to be read by title.

(The motion was seconded and carried.) (Adjournment.)

(Papers read by title):

QUALITÄTSBEZAHLUNG UND QUALITÄTSBELOHNUNG IN DER FINNLÄNDISCHEN MILCHWIRTSCHAFT.

OTTO P. PEHKONEN, technischer Direktor der Firma Valio, Cooperative Butter Export Association, Helsingfors, Finnland.

Die Einführung und Fortentwicklung der Qualitätsbezahlung in der finnländischen Milchwirtschaft steht in unmittelbarem Zusammenhang mit der genossenschaftlichen Organisation und Konzentration des Molkereiwesens in Finnland. Solange das Molkereiwesen im Lande durch Private und Aktiengesellschaften vertreten waralso bis etwa 1900-wurde bei der Bestimmung des Milchpreises nicht einmal der Fettgehalt als wichtigste Grundlage der Preisfestsetzung betrachtet, sondern die Molkereien bemassen den Milchpreis oft mehr nach anderen geschäftlichen Gesichtspunkten. Erst als die Viehbesitzer nach 1900 zur Gründung von Genossenschaftsmolkereien schritten, wurde die Basis der Preisbestimmung bei der Abrechnung über die Milch eine andere, indem die Genossenschaftsmolkereien von Anfang an die Milch nach deren Fettgehalt liquidierten. Für die Bestimmung des Fettgehalts der Milch wurde allgemein die Gerber'sche Säuremethode in Gebrauch genommen und für die Milch nach dem Fettgehalt in 1 Kg. Milch sowie für die Sahne pro Fettkilo bezahlt. Aehnlich wie die privaten Molkereien und die