The local twelve-hour change of temperature lapse rate from 5,000 to 10,000 feet is computed in 50 cases as a function of horizontal velocity divergence and vertical variation of horizontal temperature advection, and the result compared with the observed change of lapse rate. The divergence is measured from charts of horizontal velocity components; the vertical variation of horizontal temperature advection is measured by the rate of flow across the isotherms at two different levels and, as an alternate method, by the hodograph. The standard errors of estimate for these two methods are, respectively, 2.2 and 4.4 C km−1. The reliability of the measurements of divergence is studied by comparing the measured values with values computed from the difference between observed change of lapse rate and vertical variation of advection. The correlation coefficient relating these two methods is + 0.60 when advection is measured on constant level charts and + 0.40 when advection is measured by the hodograph.