From the results of the foregoing tests it would appear that: (a) Mengo virus is serologically closely related to Col. SK, M.M. and EMC viruses, (b) Mengo and Col. SK viruses bear no antigenic relationship to the Lansing, Yale-SK (from Yale laboratory) or MEF-1 strains of rodent-adapted poliomyelitis, (c) Mengo virus is not neutralized by convalescent serum from monkeys immunized against monkey-adapted strains of poliomyelitis, (d) Mengo virus is not serologically related to any of the strains of Theiler's encephalomyelitis virus of mice, (e) there is no essential difference in the pathological findings in the experimental disease induced in small laboratory animals by Mengo, EMC and Col. SK viruses, and (f) the Col. SK and M.M. strains of virus are serologically identical. (The similarity of Col. SK and M.M. viruses has already been reported by Schatz and Plagger (31), and this observation has been confirmed.) Since Col. SK virus bears no antigenic relationship to any of the poliomyelitis strains tested, it is reasonable to suggest that M.M. virus is also not serologically related to poliomyelitis virus. The evidence does not support the contention that Columbia-SK and M.M. viruses are rodent-adapted strains of poliomyelitis virus.

There is, therefore, a group of viruses, Columbia-SK, M.M., encephalomyocarditis and Mengo encephalomyelitis, which appear to be closely related not only serologically but also in their pathogenicity and in the pathological conditions which they produce in experimentally infected animals. Evidence is at hand to show that Mengo, EMC and Col. SK viruses are also immunologically closely related. The complete results of these cross-immunity experiments will be reported at a later date.

This group of viruses differs from the virus of poliomyelitis in (a) its pathogenicity for laboratory animals, (b) its wide host range, (c) its antigenic characteristics, (d) the lesions which it produces in small experimental animals, (e) the readiness with which all members of this group may be cultivated in chick embryos (32–35).8 Furthermore, since Mengo virus was isolated from Taeniorhynchus (C.) fuscopennatus Theobald and from a mixed batch of Taeniorhynchus spp. (1), there is circumstantial evidence that this group may be mosquito borne.

Jungeblut and Sanders (9) in discussing the Columbia strain of SK virus (Col. SK) state that the cotton rat seems to be able to carry this virus in a state of complete latency or subclinical infection. Powell et al. (34) have recently demonstrated that the Columbia-SK and M.M. strains of virus may be transmitted to white rats; the infected animals exhibit no symptoms but have large amounts of virus in their brains in the first passage. This is suggestive of the possible place of origin of the Columbia-SK strain and would offer an explanation for the differences which were shown to exist between the Yale-SK strain of poliomyelitis virus obtained from the Yale laboratories and the Columbia-SK strain, the latter being regarded by Jungeblut and Sanders (9) as a mutant of poliomyelitis.

That EMC and Mengo viruses infect man has been shown by Smadel and Warren (4) and by Dick et al. (38). Smadel described cases of “3-day fever” in Manila; he was able to demonstrate high antibody levels (neutralization indices above 1,000) to EMC virus in 7 persons who had suffered from “3-day fever” with neurological signs and symptoms. A comparison of the neutralizing capacity of early and late sera from three of these persons showed a very significant rise in antibody. Furthermore, smaller amounts of antibody were found in samples of sera from 10 other persons. Dick et al. described a case of encephalomyelitis occurring in Uganda (Central Africa) from which a strain of Mengo virus was isolated. Convalescent serum from the patient showed a significant rise in antibody titer to Mengo virus (38). No specific studies on the distribution of immunity to Mengo virus have as yet been made in Africa, but 2 of 143 sera examined contained antibodies in high titer. (Both of these sera were from children in the Budongo forest area of Western Uganda.) The original isolations of Mengo virus have already been described (1). The facts that (a) the human patient and the pyrexial rhesus monkey (from which 2 of the strains of Mengo virus were isolated) developed antibodies to Mengo virus during convalescence and (b) unused monkeys have developed antibody to this virus while caged in the compound at Entebbe (14), are sufficient evidence that these strains of Mengo virus did not originate in the inoculated mice.

In view of the suggestion that has been made that Col. SK virus probably originated in cotton rats, the fact that neutralizing antibody to EMC virus has been found among trapped wild rats (39), the observation that rats are not uncommon in the vicinity of the laboratory at Entebbe where Mengo virus was isolated, the possibility must be considered that not only the mongoose (1) but also the rat may act as a host of this virus. Such a possibility might be an epidemiological link between the isolation of Col. SK virus in New York and EMC virus in Florida For the cotton rats Sigmodon hispidus litoralis employed in the Col. SK isolations (9) have as their habitat the “eastern part of the peninsula of Florida, from Lake Harney to the Everglades” (Bailey, 40). Dania (Florida), where EMC virus was isolated (3), lies within that area.

The isolation of Columbia SK and M. M. viruses in New York, encephalomyocarditis virus in Florida and Mengo encephalomyelitis virus in Uganda would indicate a wide geographical distribution of this group of agents.

Much confusion has been caused by the use of the names “Columbia-SK and M. M. murine strains of poliomyelitis.” Since it has been shown that these viruses bear no antigenic relationship to the strains of poliomyelitis virus employed in this study, and since they are closely related serologically to Mengo and EMC viruses, it is suggested that the words “of poliomyelitis” be dropped from the names. Since EMC virus is indistinguishable from Mengo virus and, as we have pointed out, myocarditis is not a specific lesion characteristic only for this group, the name encephalomyocarditis is also not satisfactory. A common name is indicated for the four viruses of this group.


The experiments upon which this paper is based were performed at the International Health Division Laboratories of The Rockefeller Foundation, New York, during leave from the Yellow Fever Research Institute, Entebbe, Uganda, and while on a Rockefeller Foundation fellowship.


The behavior of Mengo virus in chicken embryos resembles that described for eastern, western (36) and Venezuelan (37) equine encephalomyelitis viruses in its rate of multiplication and mode of propagation (35).

This content is only available via PDF.