Abstract
Measurements must be outside the limits of experimental error.
Clinically it is seldom possible to make a series of measurements of the same material.
Hence the accuracy of a method must be previously determined in the laboratory.
Small degrees of physical abnormality have an importance in mental disease which they may not possess in the wards of a general hospital.
Examples discussed: Latent jaundice; mild degrees of uræmia; lymphocytosis.