[Letters to Editor]

Abstract

DR. ALLEN'S letter is welcome because it emphasizes the fact that no simple transformation can act as a panacea. The choice of the appropriate technique for any given problem must always be based, if possible, on direct evidence of its suitability. Another type of distribution has been discovered by Bagnold¹, who studied the distributions of the sizes of particles of sand deposited by wind. They were not lognormal, but when the logarithm of the diameter was plotted against the logarithm of the frequency, the observations were closely fitted by two straight lines. This fact presumably depends on the physical factors governing the deposition of particles from moving air, and may have many applications in geology. Such distributions can be normalized, but no simple general formula can be given since the slopes of the two lines may vary independently, and it is doubtful whether normalization would serve any good purpose. The logarithmic transformation is a useful tool; but, as Dr. Allen has emphasized., it must be used with care.