MATERIALS AND METHODS Interfering virus. In most experiments the Melbourne strain of influenza virus A was used as freshly harvested allantoic fluid, mixed with a 2 per cent sodium citrate solution in normal saline and borate buffer pH 8-5, in the ratio 6 parts virus, 2 parts citrate-saline and 1 part borate buffer. This was heated at 56 for 1 hr., which abolishes the infectivity of the virus while retaining its interfering activity. The heated virus is referred to as heated MEL. In a few experiments the virus was inactivated with ultra-violet light. It was first partially purified by absorption from infected allantoic fluid on to chick red cells at 0 and elution at 370 into normal saline. The eluate was then irradiated in shallow layers with a mercury vapour germicidal lamp giving maximal emission at 2537 A. The time for irradiation was decided empirically by using doublethe minimal time necessary to inactivate the infectivity of the virus preparations used.

Other virus 8train8.-Other viruses used were the PR8 strain of influenza A, the haemagglutinating virus of Japan or Sendai virus (Kuroya, Ishida and Shiratori, 1953) and vaccinia virus. All strains were kept in capillary tubes at-70. Buffer8.-The buffer which we prefer is that described by Earle (see Parker, 1950), but at one point in these studies a bicarbonate buffer could not be used and a glucosol-phosphate buffer was used instead (see Fulton and Armitage, 1951). Chorio-allantoic membranes.-Membranes from 10-or 11-day eggs were used. If whole membranes were required the shell was removed from the pointed end of the eggs, the egg