The capacity of a broad scale virus to induce an invasive tumor in a short time through plural activity at different levels has been related to its richness in lipids. Its analysis makes us suspect the presence of several parts in the virus, each one able to act at a different level, as in the case of active chemical carcinogens. A similar plural influence can be seen exerted by viruses other than those with carcinogenic activity. The study of bacterial viruses has shown the existence of such plural parts. (Note 1)

Luria (74) has shown, after irradiating a bacteriophage with ultraviolet light, that if lytic activity can no longer be obtained by the intervention of a single one of these particles, it can be induced with two or more of them. They act as though several parts, which usually are present in the virus but which were unequally inactivated by the irradiation, would be necessary in order to induce the process of lysis. This agrees with the experiments of Debruck and Hershey (75), which have shown that new types of viruses with new properties can be obtained when units of the same phage strain or related strains are mixed together. The new properties are combinations of those of the mixed units. (298)

Similar changes in the plural constitution of the viruses would explain other peculiarities observed in bacterial phages. Bacteria can carry phages for generations before any lytic activity occurs. The lysogenic strains (76) of bacteria are examples. It is possible that a virus may undergo temporary changes under certain circumstances; this would explain the frequent impossibility of finding a virus immediately after it infects a bacterium, and the "disappearance" of some viruses in animals immediately after infection. Since, in both cases, the virus is found later, a change which makes it unable to act and thereby be detected is plausible. The "masked" virus would be one with only some of its plural properties present. The possibility of recovering a lost property was demonstrated in the Berry Dietrich phenomenon, when a heat inactivated myxoma virus recovered its lethal capacity if inoculated along with a fibroma virus.

This concept of plural activity finds further application in the explanation of many phenomena observed in viruses in general and in variations in carcinogenesis. The "self sterilization of the neuroinfections" described by Levaditi has to be regarded rather as partial inactivation of the viruses especially if the viruses can be reactivated. This occurrence must be separated from cases where a total destruction of the virus can be supposed to have taken place.

The lethal infection induced in mice injected intraperitoneally with salivary gland viruses of certain strains is an example of the latter. The presence of the inclusion bodies in liver and other organs, and the total inability to produce the disease in other mice (77) can be interpreted as a sign of a destruction of the viruses in the organism. The inclusion bodies can be interpreted as resulting from an agglutination of the viruses themselves as shown by Nicolau in herpes. (78) In other cases, such as protracted herpes infection in rabbits (79), or vaccinal infection in rabbits (80), only partial inactivation can be considered to occur since electrophoresis, repeated passage, or even dilution restores pathogenicity. The restorative factor can be of varied nature. Cases in which pathogenicity is restored by a nonviral agent—activation of the virus of swine influenza in the presence of Hemophilus influenza (81), for example—are most revealing.