Comparing the efficiency of insect traps

An experiment is described in which three light traps were tested for relative efficiency by having three locations in a wood and moving the traps each day so that over a period of three nights each trap had been once in each position. The actual number of moths caught were converted to a logarithmic scale and the mean log. (i.e., log. of the geometric mean catch) was used for comparison. By analysis of variance it is possible to show what portion of the differences between the catches is due to the difference between the locations, to the difference between the nights, and the difference between the traps.