Further observations and trials were conducted with insecticides against larvae of Culicoides impunctatus Goetgh. on Soutra Hill, Midlothian. The analysis of the data was complicated by a large natural reduction (76%) which occurred in the untreated plots. Nevertheless, it appears that the following dosages of insecticidal preparation were still active two years after application:—(a) 50 and 200 mg. p,p′DDT/sq. ft. applied as a dust, wettable powder or water-miscible concentrate (99% control), (b) 50 and 200 mg. γ BHC/sq. ft. as a wettable powder or 200 mg. γ BHC/sq. ft. as a water-miscible concentrate (95% control), (c) dieldrin at 25 mg./sq. ft., which gave 97 per cent, control and (d) chlordane at 50 mg./sq. ft., which gave 84 per cent. control.

The effect of spray volume on insecticidal effect was investigated by applying 25 mg. and 50 mg. p,p′DDT/sq. ft. in the following spray volumes:—5, 10, 15, 25, 50 and 80 gals./acre. The effect of 50 mg./sq. ft. at each volume was less marked than in the previous work, when only one concentration was used. Weather conditions in the present season were unusual, but there were indications that moderate spray volumes (15 and 25 gals./acre) give the best results at this dosage of insecticide. The combined results for 50 mg./sq. ft. at all volumes gave only 43 per cent, residual control, while those for the lower dosage showed no residual control.

The results are discussed with reference to the effect of rainfall on residual control. It is concluded that the volume of spray applied is unimportant provided the droplets of insecticide are closely and evenly distributed on the mossy vegetation that overlies the peat. Subsequent prolonged rain, of the order of 20 inches, is required to attain the most effective distribution of the insecticide, which is at the surface of the peat, where the eggs are laid and the newly hatched larvae will come in contact with the poison.