Abstract. Alternating temperatures stimulate the germination of Rumex crispus L. and Rumex obtusifolius L. The optimum period spent at the lower temperature in a diurnal cycle is greater than that spent at the higher temperature. Under most conditions the optimum period at the upper temperatures is about 8 h but, as the upper temperature of a cycle is increased, the optimum period at the upper temperature becomes shorter and more critical. Thus when it is 35°C the optimum period is 2.5–4 h in the light, or about 1 h in the dark. The effect of alternating temperatures is much less in the dark than in the light and in general only extreme alternations with short periods at the higher temperature are effective in the dark. In the light any temperature alternation within the range 1–35°C is effective to at least some extent, providing the temperature difference is 5°C or more and providing the alternation includes one temperature which is above approximately 15°C and one which is below approximately 25°C. The optimum temperature difference is about 15°C. In the light, 4 to 10 cycles saturate the response, but in the dark, where the effect is much less, the response may not be saturated even by 16 cycles. KNO3 at 10−3 M has little effect on the response to alternating temperatures either in the light or the dark. The response to alternating temperature regimes does not appear to vary in quality, i.e., in terms of which particular treatments are best, but it varies in magnitude with site and year of seed collection; and it increases slowly during dry storage, even when stored at a temperature as low as 1.5°C.