Is self-estimated linear length linear?

There is wide evidence that magnitude estimates of apparent length are essentially linearly related to physical length, but it is yet unknown whether these self-estimates are linear, that is, whether they are linearly related to apparent length. Gage in 1934 and Pfanzagl in 1959 devised two tests that would help resolve this problem. In the study presented here, these tests were used to select subjects producing equidifferent apparent lengths. Mean magnitude estimates and mean ratings of equidifferent apparent lengths were found to be linearly related to these same lengths, thus showing that self-estimates of apparent length were linear.