In order to examine certain possible irregularities in the loudness function, the method of adjustment was used to halve and double the loudness of a 1000‐cy tone at the levels 30, 60, 90, and 120 db re 0.0002 μbar. The results tend partially to confirm some of the irregularities found by Robinson. The results also indicate that a reasonable approximation to the loudness function is the rule that a change of 10 db corresponds to a loudness ratio of 2:1. This means that loudness grows approximately as a power function of intensity. It is noted that the power function also describes the relation between psychological magnitude and stimulus magnitude on a dozen other perceptual continua.