The fluctuations in signal strength of KDKA, Pittsburgh, Pa., were observed through the sunset period of Morgantown, W. Va. Since the two cities are approximately on the same meridian, any variation at sunset should be due to changes in the Heaviside-Kennelly layer and not to refraction through the earth's shadow. Observations were made on twenty-one days; it was found that on bright clear days, the curve fluctuated considerably, while on cloudy days, the curve was fairly steady. The fluctuations can be accounted for by assuming a reflected wave on clear days.