This paper reports nonoptical height-gain measurements on a remarkably uniform desert link at frequencies of 25, 63, 170, 520, 1,000, 3,300, and 9,375 Mc using horizontal polarization. It has been shown that when the index of refraction of the atmosphere is a linear function of elevation, the problem of refraction and diffraction may be represented approximately by a problem in diffraction alone. Data taken under linear and slightly nonlinear index of refraction gradients are compared with fields predicted by standard diffraction theory. A mechanism other than diffraction is required to explain the higher-frequency fields observed under standard meteorological conditions.