SUMMARYThe contact poisons rotenone, 5,5‐dimethyldihydroresorcinol dimethylcarbamate (‘Dimetan’), 2‐bromomercurithiophen, 2‐isovaleryl‐1,3‐indandione (‘Valone’), α‐chlordane, toxaphene and DDT were tested, in probit assays, on as many as possible of four insect species (Oryzaephilus surinamensis (L.), Tribolium castaneum (Herbst), Tenebrio molitor L., and Musca domestica L.) by applying the poison so that there was no ‘pick‐up’ effect, and then keeping the insects at each of two post‐treatment temperatures (usually 10 and 28° C.) for as long as possible.The same insects were counted repeatedly throughout each test. At each counting, the two ED 50's were found and from these the temperature coefficient of toxic action (ratio of ED 50's) was calculated; temperature coefficients were ‘positive’ or ‘negative’, according to whether the toxic action was greater or less at the higher temperature. The time for each ED 50 to decrease to a steady value (the end‐point) was also found; the inverse of this time was the ‘speed of action’ of the poison. In some tests the end‐points were not reached, even though the insects were kept until the proportion dead in the control batches reached about 40%. Temperature coefficients, measured soon after treatment, were most probably temperature coefficients of paralysis only; and those at the end of the test, coefficients of kill, with continuous gradation between.The results were characteristic of the poison used, and not of the test species.Rotenone and ‘Dimetan’ each caused an initial paralysis, from which the insects temporarily recovered before dying. With rotenone, increase in post‐treatment temperature increased the speed of the sequence knockdown‐recovery‐death, and probably the speed of action; the initial temperature coefficient (of paralysis) was negative, but it changed to a positive coefficient as time passed; end‐points were not reached. Results with ‘Dimetan’ were somewhat similar, but the coefficients were very small and variable in sign.With 2‐bromomercurithiophen, ‘Valone’, α‐chlordane and toxaphene, the transition from the initial paralysis to death was not interrupted by a period of recovery; all the ED 50's decreased steadily in size as time passed, and the observed decrease was greater at the lower temperature. Increase in post‐treatment temperature nearly always increased the speed of action, which was greater with ‘Valone’ than with any of the other poisons. The temperature coefficients were initially positive, but became smaller or sometimes negative as time passed, so that an increase in post‐treatment temperature either did not affect the ultimate toxicity (2‐bromomercurithiophen and some tests with ‘Valone’) or decreased it (α‐chlordane, toxaphene and the other tests with ‘Valone’); increase in post‐treatment temperature also increased the curvature of the line relating ED 50 to time after treatment.In the tests with DDT, the ED50's also decreased steadily as time passed, but the temperature coefficients were consistently negative. Increase in post‐treatment temperature did not affect the shape of the ED50‐time curve; no general statement can be made about the effect of temperature on speed of action.Thus, a change in post‐treatment temperature can affect the course of poisoning of contact insecticides by affecting their speed of action (which usually increased with post‐treatment temperature), or the ultimate toxicity, or the shape of the ED50‐time curve, or in some combination of these ways.Some of the explanations of the negative post‐treatment temperature coefficient of DDT are discussed.