The liquid impinger was first described by Greenburg and Smith (1922) as a dust cloud sampler. More recently the device has come into wide use for bacterial aerosol sampling, as it is often very con venient for this purpose both in laboratory apparatus and in the field. Rosebury (1947) and Henderson (1952) describe the use of the Port?n develop ment of the original device. For estimating the number of viable organisms trapped in the impinger fluid they use the technique of Miles and Misra (1938) in which serial dilutions of the fluid are plated out for subsequent incubation and colony counting. Valuable features of the impinger when sampling airborne organisms are: (1) It is compact and inex pensive. (2) The sample fluid can be plated out simultaneously on differing media thus ensuring optimum growth conditions for organisms of interest. (3) An extreme range of airborne concentra tion can be accommodated by the serial dilution technique. (4) Virus aerosols can be estimated effectively, provided that they do not contain a high proportion of very small, single virus particles which are difficult to impinge. (5) The particle retention efficiency is very high, effectively all particles down to about 0-5/x being trapped in the impinger fluid. (6) It gives a measure of the number of individual viable organisms in an aerosol, not the number of infective particles or clusters of organisms. (7) It acts as its own constant-flow metering device. (8) It is unaffected by repeated autoclaving. Impingers similar to that shown in Fig. 1 were originally developed by Prof. J. H. Gaddum and have been in use in this establishment for many years. Fig. 1 shows the modern all-glass Port?n type. Suction is applied to the small side arm and draws air in through the curving intake tube and down through the impinging jet. The jet is a short length of capillary tubing and acts as a critical flow orifice, i.e., when the suction reaches about half an atmosphere the flow in the jet attains sonic velocity (Druett, 1955) and any subsequent increase in suction cannot increase the jet velocity further. Provided therefore that there is at least half an atmosphere of suction the impinger operates at constant flow and, once calibrated, does not require a flowmeter. The sonic velocity air jet strikes the flat base of the flask, throwing out its burden of particles which become suspended in the violently agitated circulating liquid. The apparatus shown in Fig. 1 has a jet diameter of 11 mm. giving a flow rate of 11 litres/min. and it holds 10 ml. of liquid in the flask, which has an internal diameter of li inches. The flask is made long in proportion to its width so that splashes of liquid cannot be entrained out of the suction tube (Rosebury used a conical flask which splashes less owing to the broader base, and which stands safely on the bench). The dis tance between the jet and flask base is 4 mm. It will readily be seen that the im pingement process is of great violence, for in a distance of about 1 cm. particles are accelerated up to sonic speed (about 760 m.p.h.) and are thrown against the hard glass base of the flask at approxi -All-glass Port?n impinger. 287