Fig. 2. Diagram of experimental arrangement as viewed from above (top) and potential landing points as viewed laterally and perpendicularly to the array of sheets (bottom). The null frequency of contact with a sheet is a function of distance from the drop point (P) to its landing position (A–E). Relative to the most direct path (i.e. from P to C), the distance from P to A or to E is 5.1% greater, and from P to B or to D is 1.3% greater. Thus, the proportions used to calculate null frequencies for G-tests were: 0.195 (landing points A and E), 0.203 (landing points B and D) and 0.205 (landing point C).