“…Only one specimen was introduced into the chamber at one time, and numerical values of the intensity of reaction to the gradient were based on the respective times spent by the animal in each half of the apparatus. In most of tl,1eprevious work on the behaviour of animals in humidity gradients using an 'alternative' chamber, several specimens have been introduced at once, and measures of intensity of reaction based on readings of the number of animals in each half (Gunn, 1937;Gunn & Cosway, 1938;Pielou & Gunn, 1940;Lees, 1943). This method gives greater uniformity of results, since each result is an average for several specimens, but it cannot be applied to Talitridae because of their high rate of waterloss.…”