Mendeleev and the mathematical treatment of observations in natural science

Abstract

D. I. Mendeleev (1834-1907), the eminent chemist, rejected doubtful experiments and spoke out against amassing observations. He gave thought to eliminating systematic errors and offered a simple test of the "harmony" of observations. Modern statistics has recognized harmony as symmetry of the appropriate density function and has independently quantified asymmetry in accordance with Mendeleev's idea. Mendeleev made mistakes in estimating the plausibility of his data, and he hardly knew Gauss's second formulation of the method of least squares. An analysis of his work sheds light on the level of statistical knowledge in the natural sciences beyond astronomy and geodesy in the late 19th century. © 1996 Academic Press, Inc.