As the thermometer descended, it passed through a succession of strata of ever-increasing temperature. Consequently the mercury, which, it will be remembered, had completely filled the instrument when it was at the surface, began to expand according as it was exposed to greater temperatures. As the mercury expanded, it must, of course, flow out of the tube and be lost, because the tube had been already full. So long as the mercury was gaining in temperature, more and more of it escaped from the top of the tube, and the flow only ceased when the instrument was resting at the bottom of the hole, and the mercury became as hot as the surrounding rocks. No more mercury was then expelled, the tube, however, remaining full to the brim. After allowing a sufficient time for the temperature to settle definitely, the thermometer was raised to the surface. As it ascended through the long bore the temperature surrounding it steadily declined. With the fall in the temperature of the mercury the volume of that liquid began to shrink; but the mercury already expelled could not be recalled. When at last the instrument had safely reached the surface, after its long journey down and up, and when the mercury had regained the temperature of the air, the lessened quantity that remained told the tale of the changes of temperature. It is now easy to see how, even in the absence of an engraved scale on the instrument, it is possible to determine, from the amount of mercury remaining, the temperature to which the thermometer has been subjected at the bottom of the boring. It is only necessary to place this thermometer in a basin of cold water, and then gradually increase the temperature by adding hot water. As the temperature increases the mercury will, of course, rise, and the hotter the water the more nearly will the mercury approach the top of the tube. At last, when the mercury has just reached the top of the tube, and when it is just on the point of overflowing, we may feel certain that the temperature of the water in the basin has been raised to the same temperature as that to which the instrument was subjected at the bottom of the boring. In each case the temperature is just sufficient to expand the quantity of mercury remaining in the instrument so as to make it fill precisely both bulb and stem. When this critical condition is reached, it only remains to dip a standard thermometer, furnished with the ordinary graduation, into the hot water of the basin. Thus we learn the temperature of the basin, thus we learn the temperature of the mercury in the thermometer, and thus we determine the temperature at the bottom of the boring over a mile deep.