R meters* were first described in 1910 by Bestelmeyer, who recognized the existence of errors due to rapidly changing conditions. However, little quantitative information is available even at the present time on the magnitude of the errors to be expected in particular maneuvers. These "dynamic errors" can be negligible or relatively large depending upon the time required for changes in the actual rate of climb as compared with a period called the "characteristic time" of a particular instrument. If the forcing changes occur in a time much less than the characteristic time, large dynamic errors will appear while slower changes have progressively less effect. The present paper outlines the general theory of the conventional " capillary leak" rate-of-climb meter and describes a laboratory method for determining the coefficients necessary to predict performance in flight. Theoretical and laboratory results in the form of curves are compared with data taken from flight tests. Special attention is paid to the case of sinusoidal changes, and non-dimensional curves applicable to any frequency are given.