of its high-collection efficiency, especially for low values of K, was selected as the blade shape for rotating impactor experiments. The theoretical flow conditions of the analogue trajectory studies are quite different from those which may be encountered in actual sampling. However, reported experiments * using a suspension of 0.365 micron polystyrene spheres indicated general agreement with the predicted values. At 50 millibars pressure, where the outer 1-1/2 inches of the 4-1/2 inch rotating blade array was operating at an impaction parameter, K, greater than 0.2, a collection efficiency of 73 percent was measured. work can produce an array with higher efficiency and considerably reduced drag. It is expected that further This is especially noteworthy since the actual impactors will operate in the partial slip flow region with a mach numberclose to unity in the throat. The focusing impactor shown starts to choke at a mach numberof close to 0.2 and so should be im- proved for any final design. Difficulties have been experienced with rotary impactors when each impactor oper- ates in the wake of the preceeding one. with the rotorchute device. But, owing to the long spacing of the spiral, this should not be a problem To maximize sample size it is necessary to operate several impactors in parallel. Although the proposed rotorchute system will permit separation of the impactors, the general problem of parallel operation must be investigated before an optimum impactor design can be developed. This problem does not materially affect the feasibility of the system. In the region of interest the mean-free path of the air molecules is larger than the particle diameter with the result that the Cunningham Correction factor, ky in Equation (2) will reach very large values. This is advantageous since it permits the collection of small particles at low air velocities using large impactor surfaces. It appears that the Cunningham Correction factor can be used up to the region where the mean-free path approaches the impactor dimension. This occurs around 250, 000 feet with impactors sized for 100, 000 feet, and above 400, 000 feet for impactors sized for 200,000 feet. This is above the maximum altitude considered in this study. The minimum value for the inertia parameter in the sample design calculation is taken as K = 0.2 which corresponds to a collection efficiency of over 80 percent. remain constant. the impactor size, To maintain a constant collection efficiency, K must The velocity of the air relative to the blade Vp is a function of blade radius r. L, By making also a function of radius, the radius cancels in Equations (2) and the inertia parameter becomes a constant over the entire length. Substitution into Equation (2) for pressure and Cunningham Cor- rection factor as a function of height gives for the minimum particle size collected at 80 percent efficiency as a function of the maximum half-size of the focusing impactor array and altitude the relation d= 10.6 LYmax H/25, 700 P (3) This is plotted in Figure 3, which shows that the minimum particle size increases rapidly as the collector descends. Therefore, the lowest altitude of interest, 100, 000 feet in the design calculation, sets the size of the impactor. The maximum impactor size, Lax? at the tip of the rotorchute is shown for various altitude incre- ments of interest. The weight of air sampled as a function of altitude is equal to the volume of air swept out by the impactor array times the density of the air integrated over the altitude range sampled. The volume of air is equal to the volume per revolution times the revolutions per second times the vertical distance the rotorchute travels divided by the velocity of descent. Using a constant tip speed of 700 ft/sec and the velocity of descent from Equation (1), gives a sample size which is proportional to the integral of the air density to the 3/2 power. 10€