Klimmer and Haupt report that observations have been made in different places that perfectly healthy wives of farmers had miscarried without any evident cause. Later inquiries revealed that the cattle were infected with contagious abortion and that the women had been drinking the raw cow's milk.

According to Rebagliati, Malta fever exists in certain localities in Peru where infected goats can not be concerned in the etiology. He regards flies as possible carriers, but does not consider the possibility that cow's milk may be responsible for the disease. No statements are made as to whether the Peruvian patients had been drinking cow's milk.

PREVALENCE OF BR. MELITENSIS IN COW'S MILK

In 1911 Schroeder and Cotton reported that they had found the infectious abortion organism in 8 of 77 samples of market milk tested (over 10 per cent) and in the milk distributed by 6 of 31 dairies (over 19 per cent). Schroeder afterwards reported a higher percentage of milk samples infected with Bacillus abortus. He injected into guinea pigs 516 samples of milk from 90 dairies, and 103 of the animals developed the abortion disease. The results showed that the milk from 29 of the dairies was infected from time to time with the abortion organism.

The findings of Schroeder and Cotton have been confirmed by other investigators in other sections of the United States. Fabyan, working in Massachusetts, examined the milk from 12 cows of a thoroughbred Guernsey herd. He found the abortion organism in two of the samples. Huddleson has noted its prevalence in cow's milk in Michigan, and Fleischner and Meyer have noted its prevalence in California. There is no doubt, therefore, that a large percentage of our population at some time or other have ingested living Br. melitensis of the bovine type.

That this organism is also present in cow's milk in Europe is evidenced by the findings of Zwick and Krage, Kennedy, and Winkler. Zwick and Krage, working in Germany, cultivated the contagious abortion organism on agar directly from the milk of three cows. Their report does not indicate how large a percentage of samples of milk contained this organism. Winkler also investigated the milk of cows in Germany. By inoculation of guinea pigs he demonstrated the specific organism in the milk from 13, or 41 per cent, of 32 cows which had recently aborted. He was able to obtain cultures directly from 3 of the samples of milk. He also demonstrated B. abortus in 7, or 32 per cent, of 22 samples of market milk.

In 1914 Kennedy, working in England, was testing goat's milk for agglutinins specific to Micrococcus melitensis, and found, to his

surprise, that the control cow's milk gave a positive result. Following up this observation, he noted that 5 out of 13 samples of mixed cow's milk from 13 different dairies in London contained these agglutinins. The milk of 22 individual cows was then tested, and a positive reaction was obtained from 3 samples. Kennedy's observations were confirmed by Bassett-Smith and by Cummings, Coppinger, and Urquhart, who also tested the milk of individual cows or the mixed milk of London dairies and found agglutinins specific to Micrococcus melitensis in about the same percentage of samples of milk as Kennedy. The English investigators did not know how to interpret their results, but they mentioned the possibility that some of the continued pyrexias in the country might be explained by the infected milk.

The investigators of Malta fever in Mediterranean countries. found that cows, as well as other domestic animals, might be infected with M. melitensis. Shaw tested the serums of 33 cows in Malta and found specific agglutinins in 10 of them. He cultivated the organism from the milk of 2 of the cows. Horrocks examined the cows of a dairy of 12 in Gibraltar and found that the milk of 1 cow agglutinated M. melitensis in a dilution of 1 to 100.

INFECTIOUSNESS OF THE ABORTUS VARIETY OF BR. MELITENSIS FOR MAN

Mohler and Traum obtained serums promiscuously from 42 human beings. They tested the serums with B. abortus antigen and obtained no positive results by either the complement fixation or the agglutination tests. They also inoculated 56 tonsils and adenoids into guinea pigs and obtained the abortion organism from the organs of one of the inoculated animals.

Larson and Sedgwick reported that in systematically testing by complement fixation the serum of women who had aborted they found a larger number giving a positive reaction when the contagious abortion organism was used as an antigen than when the usual antigen of the Wassermann test was used. These investigators also tested the blood of 425 children for antibodies against B. abortus. They found 72 (17 per cent) which gave positive results with the agglutinin and complement fixation tests. It was shown that true antibodies were being dealt with, for a positive serum could be rendered negative by absorption with the abortion organism. Differing proportions of positive reactions were found in different groups of children. In one group 48 per cent of the serums gave positive reactions. A group of children supplied with milk from a herd which had never been affected with contagious abortion did not give a positive reaction. The authors considered it probable that the positive reactions indicated an active immunity, the result of repeatedly receiving the organism in the milk, though the individuals might not

have suffered any notable illness. Sedgwick and Larson tested the serum of four children with clinically demonstrable enlargement of the spleen. Two gave a positive and two gave a negative complement fixation reaction. These authors also reported instances of women aborting when there was an epidemic of abortion among the cattle on their farms. No definite cause of the abortion in the women could be found.

Ramsey examined the serum from 116 children, and 7 samples were found which gave a positive complement fixation reaction with abortus antigen.

Nicoll and Pratt carried out the agglutination tests with B abortus antigen and the serum of infants and children who were inmates of a foundling asylum. With few exceptions they found the serums negative in 1:10 and 1:50 dilutions. Two samples of serum from one child, undoubtedly rachitic, gave good reactions as high as 1:100. The serum of one child having enlarged tonsils gave complete agglutination in a dilution of 1:200. The serums from five other children with enlarged tonsils and adenoids gave a moderate or marked reaction in a dilution of 1:10. Guinea pigs inoculated with these tonsils failed to develop abortus infections. The serums of two children with no noteworthy clinical abnormalities gave slight reactions in the 1:10 dilution. These authors also investigated the case of premature birth at seven months in a woman who gave a negative Wassermann reaction; they found that the serum of the mother gave an agglutinin reaction in dilutions up to 1:300. The serum of the child was also positive. The vaginal discharge and milk of the mother and the feces of the child were injected into guinea pigs with negative results. Two other cases of miscarriage gave negative agglutinin reactions. The authors state that the presence of serum reactions are suggestive but not conclusive.

Cooledge examined for abortus antibodies the serums of six persons who were drinking raw milk or cream, and found agglutinin and complement fixing bodies in three of them. Of four persons drinking pasteurized milk, one gave a reaction to the complement fixation test only. Four persons who had been drinking no milk gave negative reactions. Of seven persons who drank daily for experimental purposes a pint and a half of milk known to be infected with abortus, there was an increased agglutinin content in the serums of five, a titer of 1:40 being the highest reported. The individuals were apparently in perfect health throughout the experiment. The author therefore concludes that the antibodies indicate a passive immunity due to the absorption of antibodies from the milk. That conclusion does not appear reasonable; for the milk consumed had an agglutinin titer of 1:40, and in one case the serum of the subject showed an aglutinin titer of 1:40.

By means of the complement fixation reaction, Williams and Kolmer examined the serums of 50 women who had recently had interrupted pregnancies. They found no evidence of abortus infections. The serums of 12 aborting women gave negative agglutinin reactions.

Fleischner and Meyer tested 75 infants for cutaneous hypersensitiveness to abortus antigen and found no specific reactions.

Nicolle, Burnet, and Conseil inoculated five human subjects with laboratory strains of abortus-two with a strain of bovine origin and three with a strain of porcine origin. These subjects suffered no particular ill effects from the experiment, although the serums of three of them acquired specific agglutinating properties-two in a 1:20 dilution and one in a 1:250 dilution. One subject showed a slight febrile reaction (maximum 38.3 degrees) on the two days following the day of inoculation. These investigators conclude from their experiments that "Bacillus abortus" is not pathogenic for the human species, a conclusion contrary to the facts which have been revealed since their experiments were carried out.

The preceding references furnish presumptive evidence that the abortus variety of Br. melitensis may infect man. In not a single case, however, is the evidence conclusive. The first human case of unquestionable abortus infection was reported by Keefer in January, 1924. The clinical picture and the course of the disease were characteristic of Malta fever as it has been commonly observed in regions where it is endemic. The source and mode of infection could not be determined. There was no history of the patient having ingested goat's milk or products of goat's milk, but he was in the habit of drinking large quantities of raw cow's milk. A sample of the patient's serum and a culture of Br. melitensis isolated from his blood were submitted to the Hygienic Laboratory through the courtesy of Dr. H. L. Amoss. A study of these was made by the writer, and it was found that the infecting organism was of the abortus variety. The data will be presented further on. Shortly after the publication of Keefer's paper it was reported from South Africa that the abortus variety was responsible for Malta fever in Rhodesia. In 1921 Bevan had called attention to the fact that in that country a number of cases had occurred in which the patients showed clinical manifestations of Malta fever without having imbibed goat's milk. Moreover, the patients resided on farms where infectious abortion of cattle was known to exist. By agglutinin absorption tests Orpen proved that Bevan's suspicion that Malta fever in Rhodesia was caused by infection with the abortus type of the organism was founded on fact. Strains

from four human cases were compared with two strains of melitensis, one strain of paramelitensis, and one strain of abortus from Europe, and two strains of abortus from Rhodesia. The human strains from Rhodesia agreed with the abortus strains and differed from the strains of melitensis and paramelitensis.

Later De Korte reported from South Africa a case of Malta fever in a man who apparently contracted the infection through the removal of the placenta from a cow suffering from infectious abortion, and Duncan reported the case of a butcher who was in the habit of drinking large quantities of cow's milk. Agglutinin absorption tests showed that the organism cultivated from this patient's blood resembled the abortus variety more closely than it did the melitensis variety.

In a recent publication Ficai and Alessandrini present evidence that certain cases diagnosed as Malta fever in Italy are in reality due to infection from bovine sources. In these cases illness followed about 10 days after handling an aborted bovine fetus, or after assisting in the treatment of a cow which had aborted.

Viviani has also reported from Italy a case of melitensis infection traced to a bovine source. The patient denied the possibility that he could have received his infection from goats. He owned a herd of cows in which contagious abortion existed, and he and his family had been using milk from a cow which had recently aborted. The infection could not be attributed to the milk, however, for they never drank it raw. About 15 days before the onset of illness the patient had butchered a sick calf belonging to his father, whose herd was badly affected with contagious abortion. The calf was considered the probable source of infection.

The Italian writers did not determine the variety of Br. melitensis with which they were dealing. In countries where the true melitensis varieties or the paramelitensis or para-abortus varieties are common, cattle infections can not be assumed to be of the abortus variety.

Within recent months there have come to the Hygienic Laboratory for identification three strains of the abortus variety from human cases in South Dakota, Connecticut, and New York. The serum from another case of Malta fever in Baltimore has also been received. The histories in these cases exclude infection from goats, and agglutinin absorption tests identified all the infections as the bovine type. In two of these cases, however, the histories suggest infection from hogs. They are discussed in the following paper. In the remaining two cases the patients had been drinking raw milk from herds known to be infected with the organism of contagious abortion.