The use of logarithms in the interpretation of certain entomological problems

SUMMARY Evidence is brought forward that in comparing the numbers of insects caught under varying conditions, with particular reference to captures in a light trap, more consistent results are obtained if the geometric means are compared than by the use of arithmetic means. This is most conveniently done by summing the logarithms of the numbers instead of the numbers themselves. If any of the numbers in the series is zero it has been found practical to add one unit to all the captures in the series and so deal with log (n+1) instead of log n. The use of the logarithms prevents the swamping of the results in a series of observations by very high numbers on a single night. It also gives a more normal distribution of departures from a mean. As a result of the latter it is possible to apply the statistical formulae for standard deviation etc. which are not applicable to the skew curve obtained by the use of the departures of the numbers themselves from an arithmetic mean.