Temperatures were measured as a function time and depth in a borehole before and after it had been filled with water. The hole is 25 cm in diameter, 340 m deep, and effectively sealed from influx of ground water. The measurements reveal that: (1) Temperature differences between the air-filled hole and the water-filled hole (after dissipation of the water injection disturbance) do not exceed 0.05°C at the same depth. (2) Temperatures in the water-filled hole exhibit oscillations at all depths where temperature increases downward but are stable at the bottom and near the surface where temperature decreases with depth. (3) The temperature oscillations have periods ranging from several minutes to several hours. (4) The amplitudes of the oscillations are several hundredths of a degree, and are roughly proportional to the geothermal gradient near the point of measurement, thus suggesting that the size of cells or eddies are rather constant and that the distance of water movement is no more than several times the inner diameter of the casing. (5) No oscillations were detected in the air-filled hole, but because the time constants of the instrument are long in an air-filled hole, the possibility of short period oscillations cannot be excluded.