IMPACTION ON A ROTATING STICKY CYLINDER Rotating the cylinder at a peripheral speed comparable with the wind speed would be expected to reduce the thickness of the boundary layer on the surface, to induce the well-known Magnus effect, and to produce a local rotation of air round the cylinder itself. It was not obvious whether these effects would alter impaction efficiency.* Tests were accordingly made in the Rothamsted wind-tunnel, with a 15.5 cm.-square constriction upwind to generate turbulence. Lycopodium spores (32 µ diameter) were blown in at a point on the axis 18 cm. upwind of the constriction, and, after a diffusion path 63 cm. long, the spore-cloud reached the trapping section where a stationary and a rotating cylinder, each 0.5 cm. in diameter, were exposed simultaneously. The rotating cylinder, which was placed vertically across the axis of the tunnel, consisted of a steel rod mounted in a ball-race secured flush with the floor of the tunnel at one end, and connected by a sleeve of stout rubber tubing to the spindle of a 225-watt `Universal' motor mounted on the roof of the tunnel. The stationary cylinder was 2 cm. to one side and 2 cm. upwind of the rotatable cylinder, and both carried adhesive coatings of cellulose film which, after exposure, were removed and scanned under the microscope to measure the deposit. Speed of rotation of the cylinder was controlled by a `Variac' transformer and was measured by a strobo scopic lamp. The range of speeds that could be tested was limited by the steel rod which began to bend at speeds above 7,000 r.p.m.
With a wind of i • i metres per sec., efficiency was substantially un changed until the peripheral speed of rotation attained approximately the speed of the wind; at higher rates of rotation the efficiency decreased rapidly, reaching zero before the peripheral speed of the cylinder reached twice that of the wind. At a lower wind-speed of o•68 metre per sec., efficiency of the cylinder, when rotating at o•4 times the wind-speed, fell to 53 per cent of the stationary cylinder, and to 24 per cent at o•82 times the wind-speed. These results suggest that the centrifugal effect is suffi cient to decrease impaction to zero. The phenomenon would be interesting to explore at a wider range of cylinder diameters and wind-speeds (the work of Brun et al. refers to much higher wind-speeds than these).