The First Section of this paper contains an account of a variety of experiments undertaken with the thermo-multiplier of Nobili and Melloni, the instrument exclusively employed in the subsequent researches. By a comparison of its sensibility with that of air-thermometers, the author concludes that one degree of deviation of the needle of the multiplier corresponds to an effect indicated by about one-fiftieth of a centigrade degree on the others. Without increasing the dimensions of the multiplier, by which its sensibility would be impaired, he has been enabled, by an optical contrivance, readily to measure one-tenth of one of its degrees, corresponding to one-five hundredth of a centigrade degree. From an experiment intended to detect the heat of the lunar rays, concentrated by a polyzonal lens, thirty-two inches in diameter, and acting upon this instrument, he concludes that the direct effect of the moon upon an air-thermometer probably does not amount to one-three hundred thousandth part of a centigrade degree.