in order to admit of comparison with other experiments, it ought further to be determined to what portion of the whole heavens, and in what direction, the instrument was exposed. It may be remarked that the method of measuring the intensity of terrestrial radiation by the maximum difference between the shaded and exposed thermometers is subject even to greater error than in the case of solar radiation, for it is evident, from the extremely heavy dew on a clear night, that the whole surface of the ground attains the temperature of the dew-point, and hence a detached thermometer must sink below it. But as soon as this happens, the decrement of temperature can no longer be a measure of the intensity of radiation in the same sense it was before, for there is now a continual influx of heat to the bulb from the condensing vapour. It might easily be shewn by the same process of reasoning employed to establish the dew-point formula, that if the tension of vapour corresponding to the minimum temperature of the exposed thermometer, be subtracted from the tension of vapour at the dew-point, and the difference be multiplied by 87, this product will express the additional number of degrees, on Fahrenheit's scale, which the exposed thermometer would sink, were it not for the dew.
P. S.—I venture to add, in a postscript, a few remarks which, while they are not altogether foreign to the subject of this paper, are not worthy of a separate place.
It is sometimes required to determine the internal diameter of a barometer tube when circumstances prevent its direct measurement. As this method of doing so, though easily derived from an elementary optical proposition, may not immediately suggest itself to those having only accidental occasion for it, it is here annexed. Apply the points of a pair of compasses to the external surface of the tube, and holding the eye opposite to them, and as far removed as is consistent with distinct vision, measure the least distance between the parallel sides of the bore. Two-thirds of this quantity is nearly equal to the true internal diameter. A nearer value is obtained by the following rule: add to the external diameter of the tube twice the distance of the eye from one of the points of the compasses, and divide by twice the distance of the eye; the product of this quantity into the approximate value already found, is very nearly equal to the internal diameter of the tube. When the external diameter is .333 of an inch, the apparent internal diameter .12 of an inch, and the distance of the eye 4 inches, the results of the two foregoing approximations, and of an exact calculation give, respectively, as the values of the internal diameter, .08 of an inch, .0833 and .08367.
