The term drift spraying denotes a method of applying an insecticidal spray whereby it is released at a height above the ground and allowed to drift with the wind while it gradually sinks; if the source of emission is moved in a line across the direction of the wind, insecticide will be deposited at a distance downwind of that line. The method has been used for vegetation baiting, a term meaning the application of stomach poisons to natural vegetation for control of locusts; this may be more economical than the application of contact insecticides to the insect itself.

Spray particles emitted from a point source have a mean trajectory downwind inclined to the horizontal at an angle of tan−1 (f/V), where f is their terminal velocity and V the windspeed, and they are dispersed by wind eddies to all sides of this mean axis so as to form a conical plume; if the source is moved across the wind, horizontal dispersion normal to the wind tends to cancel out and only the vertical component of the dispersion need be considered. The concentration of spray drops is a maximum on the plume axis and decreases with distance from it. The spray cloud from such a source thus occupies a wedge-shaped space as it drifts downwind, and it reaches the ground over a band whose width depends on the height of the source and the degree of dispersion. The point at which the plume axis strikes the ground is determined by its declination and by the height of the source; the density of the spray deposit, whose overall value is determined by width of the band sprayed and by the quantity of spray emitted from the source per unit distance travelled across wind, is a maximum upwind from this point and diminishes more steeply towards the upwind edge of the band than towards the downwind one.

Large drops will fall faster and hence be deposited nearer the line of emission than will small drops, but in practice, in day-time, the effects of dispersion by wind eddy tend to exceed and mask those due to the range of drop size and it is thus a valid simplification to consider the spray cloud as composed of drops all of one size. The behaviour of heterogeneous sprays is best characterised by the assumption that all drops are of a diameter equal to the surface-volume mean diameter. At moderate wind speeds, the drop diameter used should be of the order of 100μ, and an oil of low vapour pressure is needed if the tendency of such small drops to evaporate before reaching the target is to be prevented.

The angle of dispersion probably depends on the degree of wind turbulence. Experimental measurements of the concentration of drops within a spray plume, relative to that on the axis and at different angular deviations from the latter, enable the fraction within each sectpr to be known, and suggest that for normal atmospheric conditions, when the lapse rate is adiabatic, 95 per cent. of the spray is contained within ± 7° of the plume axis in the vertical plane.

The calculated densities of deposits on a horizontal surface at varying distances downwind, related to a source whose height and rate of emission are unity, are shown as graphs for different values of the declination of the plume axis (predictable from the wind speed and drop size) and are compared with experimental measurements. Using these data, the conditions that will afford any desired deposit density can be selected. Since the interest of drift spraying for locust control lies mainly in the deposits on the nearly vertical stems of grass, similar data are given for values of the deposit density on an exposed vertical surface, which can be shown to be n times that on a horizontal one at the same distance downwind, when the latter is expressed as a multiple, n, of the source height.