that the abortus variety may infect cattle, hogs, or man; and that the melitensis A variety may infect cattle, horses, goats, or man.

It is to be noted that there is a distinct zonal distribution of the various serologic groups of Br. melitensis. The abortus variety, which is widespread in this country and in northern and central Europe, was not found among the 23 strains of this collection which came from Mediterranean countries. On the other hand, three of the four serologic groups prevalent in Mediterranean countries appear to be confined to those countries, in so far as this study indicates, for they did not occur among the 33 strains isolated in this country nor among the 7 strains from northern and central Europe. Only one serologic group (melitensis A) appeared to be distributed in every geographical region from which strains of this collection were obtained.

It is of interest also that the serologic types prevalent in Mediterranean countries are grouped together in Charts 1 and 2, signifying a close relationship between these groups on the one hand, whereas the serologic types prevalent in the United States and in northern and central Europe are grouped together on the other end of the charts.

The zonal distribution of strains suggests that a difference in the virulence of the serologic groups may account to some extent for the virulent type of human Malta fever in the Mediterranean countries. This would offer an explanation as to why it is that in central Europe goat's milk is commonly used without the prevalence of Malta fever such as exists in the Mediterranean countries. The view that the varieties prevalent in Mediterranean countries may be characterized by high virulence is not at variance with that already suggested that passage through the bovine animal may lower virulence. The one serologic group which is prevalent in this country as well as in Mediterranean countries appears to be as virulent here as there, judging from the epidemic at Phoenix, Ariz., reported by Lake.

It should also be mentioned here that melitensis is present in great numbers in the milk of infected goats. This fact was established by Eyre, McNaught, Kennedy, and Zammit, and is confirmed in general experience by the readiness with which the organisms may be isolated from goat's milk. Burnet and Anderson have shown that abscesses are common in the udders of infected goats. On the other hand, it was pointed out by the writer, and later by Huddleson, that the organism of contagious abortion is not ordinarily excreted in large numbers in the milk of infected cows. Since one of the factors determining an infection is the number of organisms in the inoculum, the heavy contamination of goat's milk in Mediter

50587-25- -3

ranean countries as compared with cow's milk in this country must be to some extent responsible for the greater danger of contracting Malta fever from goat's milk.

SUMMARY

The agglutinin absorption tests with 68 strains of Brucella melitensis showed that the species may be differentiated into at least eight serologic groups. Three of these groups included only one strain each.

The majority of bovine and porcine strains fell into one large group (33 strains), which is designated variety abortus. Five strains of human origin were of this variety.

Another group important in this country includes strains of human, bovine, caprine, and equine origins (12 strains). It is designated variety melitensis A.

Three groups which were found to be prevalent in Mediterranean countries did not occur among the strains received from countries outside of those regions. One of these groups is designated variety melitensis B; another, which corresponds with the descriptions of the so-called paramelitensis is designated variety paramelitensis; another serologic group is designated para-abortus, because it is servlogically closely related to the abortus variety, and exhibits agglutination peculiarities like those of the variety paramelitensis.

Simple agglutination tests can not distinguish the varieties of Br. melitensis.

STUDIES ON BRUCELLA MELITENSIS

II. THE NOMENCLATURE OF THE MELITENSIS-ABORTUS GROUP OF BACTERIAL ORGANISMS

By ALICE C. EVANS, Associate Bacteriologist, Hygienic Laboratory, United States Public Health Service

There is so much confusion in the nomenclature of the melitensisabortus group of bacterial organisms that a brief consideration of the various names that have been applied, together with conclusions as to the generic and specific names which appear to be available and valid in the light of recent investigations, must necessarily accompany the publication of reports of further investigations of this subject.

The first report of the isolation of cultures of the group of organisms under consideration was by Bruce in 1887. He obtained the specific germ, upon autopsy, from the spleens of human cases which had died of undulant fever on the island of Malta. It was not until 1893, however, when he published a more complete description of his organism, that he named it "Micrococcus melitensis."

Durham (1898) appears to have been the first writer to note that a bacillary form also occurs. He believed that conditions of temperature and medium determined whether the forms would be coccoid or bacillary.

Preisz (1903) described diphtheroid forms of the contagious abortion organism, and gave it the name "Cornynebacterium abortus endemeci." Since then it has been occasionally referred to in German literature under that name.

ACKNOWLEDGMENT

The writer is indebted to Maj. J. F. Coupal of the Army Medical Museum for the photographs.

Eyre (1912) mentions the fact that some investigators consider the organism a real "bacillus." He explains the rodlike forms as involution forms, however, or as dividing cocci between which the separation is not yet complete.

The generic name "Bacillus" for the Malta fever organism has been used by Jordan in his "Textbook of General Bacteriology since the third edition, which appeared in 1912. The majority of American textbooks, however, have clung to the name "Micrococcus

27

melitensis," although all investigators who have studied the organism in recent years have agreed that it is a rod form.

Bruce's choice of the generic name "Micrococcus" is explainable. In the Hygienic Laboratory collection of 23 strains from human sources there are three, isolated years ago, for which the geographical source is not accurately known, but circumstantial evidence indicates that two of them, possibly all three, were originally obtained on the island of Malta. The history of these strains is as follows: Strain 102 was obtained in December, 1907, from England. According to the record it was labeled "M. melitensis, Dr. Annett." Strain 103 was obtained from the Royal Army Medical Corps of London, England, in January, 1908. It was labeled "M. melitensis, R. A. M. C." Colonel Bruce, R. A. M. C., the discoverer of the Malta fever organism, was the chairman of the British Commission for the Investigation of Mediterranean Fever, which made its report of the investigation on the island of Malta during the years 1905-1907. It appears very likely that these two cultures received from England a few months subsequent to the publication of the final report of the commission were obtained on the island of Malta. There is no record of the origin of the third strain, No. 104, other than that it was obtained from the United States Naval Medical School in 1909. Since there had been no isolations of melitensis in this country as early as 1909, it seems probable that strain 104 also came to this country via England from Malta. Indeed, it may be the same strain as 102 or 103. The fact that all three strains under consideration belong to a serologic type which appears to be confined to Mediterranean countries (see Table 3 of the preceding paper) furnishes further evidence that they may have originated in Malta.

The cells of the three strains are almost coccoid. The smears present a slightly different appearance from that of the strains of other serologic types in which distinct rod forms are more evident, scattered among the coccoid cells. Bruce's description of his organism and the illustration accompanying it correspond with the morphology of the three strains. A photomicrograph of strain 104 is given in figure 3, Plate I, where its morphology may be compared with that of strains of other serologic types. The photographs do not, however, emphasize the distinction in morphology which impresses one when several fields of the slides prepared with

EXPLANATION OF PLATE

The films were prepared from 48-hour cultures on agar slopes, and were stained with carbol-fuchsin. X2420.

Fig. 1. Strain 466 (abortus variety).

Fig. 2. Strain 428 (Melitensis A variety).

Fig. 3. Strain 104 (melitensis B variety).