Using inbred laboratory stocks of the locust Schistocerca gregaria, two experiments were carried out using composite heat-treatment programmes of 40 degrees C and 30 degrees C combinations. These were designed to superimpose different heat-induced increases and decreases in chiasma frequency, to determine what interactions would occur and which effects might be dominant. It was found that both of the heat-induced decreases in chiasma frequency which can be obtained (Effects 2 and 3), were dominant over both of the different types of increase (Effects 1 and 4), with the largest decrease (Effect 3) being dominant over all others.