Bacteriophage particles are called active when they are able to give rise to production of active phage following adsorption on their specific host bacterium. This is generally detected by the formation of a plaque when plated on nutrient agar by standard techniques (Adams, 1950). If the particles are exposed to ultraviolet light, a fraction of them loses the ability to give rise to production of active phage. These particles are called inactive. This is a somewhat unfortunate term because the inactive particles are still able to perform functions characteristic of the active ones, in that they adsorb on the host bacteria and kill them. Moreover, the inability to give rise to production of active phage is not necessarily permanent because it can be overcome by at least two means, namely, by multiplicity reactivation (Luria, 1947) and by photoreactivation (Dulbecco, 1950). In multiplicity reactivation, production of active phage takes place in bacteria infected simultaneously with more than one inactive particle, in photoreactivation, it takes place in bacteria infected with one or more inactive particles and subsequently irradiated with visible light. Luria (1947) formulated an interesting theory of multiplicity reactivation, suggested by analogy with the findings of others on the recombination of genetic markers occurring in bacteria mixedly infected with several active phage particles. Luria assumed that each phage particle consists of a certain number of indispensable genetic units, which are sensitive to ultraviolet light. Each of these units is inactivated independently of the others. When one or more of them are inactivated, the whole particle becomes inactive. However, in bacteria infected with more than one such inactive phage particles, the remaining active units multiply and recombine; and this may lead to the reconstitution of a complete active phage particle under suitable conditions. We call this theory the "recombination theory" of multiplicity reactivation. Luria and Dulbecco (1949) have presented experiments in support of this theory, based on systematic measurements of the probablility that bacteria infected with more than one inactive phage particle (multicomplexes) yield active phage. We call this probability the survival of multicomplexes. The results were in fair agreement with the prediction of the theory. These experiments, however, had been limited for technical reasons to the determination of survivals generally higher than 10-.7, and therefore they did not constitute a critical test of the recombination theory; in fact, the most characteristic features of the theory come to light only at considerably lower survivals, as shown by the following considerations.