Abstract
It is shown that the results given by Lob concerning the distribution of FM discriminator click widths are incorrect. The error in Lob's analysis is indicated.