From the commencement of the researches in these laboratories the aim has been disinfection as opposed to sterilization, because the breathing of air devoid of its normal, saprophytic flora over a long period might eventually lead to a lowering of the powers of resistance of the respiratory epithelium to the invasion of bacteria in general, and, therefore, was to be avoided as far as possible. The ideal conditions for controlling cross-infection in respiratory diseases would appear to be those in which attenuation of the causal organism results, death then being more easily accomplished within the animal body. By this means immunity of the animal to a subsequent infection would possibly be higher than if there had been inhalation of organisms which were already dead.Although our chief consideration of the effect of germicides on airborne organisms has been from the bactericidal aspect, bacteriostatic effects have not been overlooked. Absence of colonies on the culture media is taken as evidence of the former and size of colonies of the latter. Absence of colonies does not necessarily imply death of the organism, as can be demonstrated by using richer culture medium than usual. The fact that in air experiments, designed to test the sensitivity of several organisms to a particular germicide, different types of culture media may be employed does not matter so much when the results are calculated on a perceiitage survival basis. Another factor is the possibility of effect of the germicide on the organism after it has settled on the medium. This effect may be a continuation of that initiated in the air or it may be an indirect one, resulting from an alteration of the constitution of the mnedium by the collection of mist particles or the absorption of vapours. These considerations hardly apply when using ordinary gravity plates, but it may be otherwise where suction apparatus is in use, involving concentration of germicide and bacteria on the culture medium.