But before we proceed to this second part of our task, it will be necessary previously to consider, first, how small a degree of heat is required to detach or raise the vapour of water from its parent-fluid. In the coldest day, I might say the coldest night, of winter, when the weather is not frosty or very damp, wet linen or paper will become dry in the course of a few hours. A greater degree of heat must indeed cause a quicker evaporation. But I am persuaded, that was it not for the pressure of the superincumbent fluid, greatly less than that of boiling water would instantly disperse the whole so heated into vapour. Secondly, That in the opinion of Sir Isaac Newton, well illustrated by the late lamented Mr. Melvil, the sun-beams appear only to communicate heat to bodies by which they are refracted, reflected, or obstructed; whence, by their impulse, a reaction or vibration is caused in the parts of such impacted bodies. This is supported by the experiment of approaching some light body, or blowing smoke near the focus of the largest glasses; and from observing, that these do not ascend, it is evident the air is not so much as warmed by the passage of those beams thro’ it, yet would instantly calcine or vitrify every opake body in nature. And from this we may collect, that transparent bodies are only heated at their surfaces, and that perhaps in proportion to their quantity of refraction: which will further give and receive illustration from those very curious experiments, of producing cold by the evaporation of liquors, published by the learned Dr. Cullen, in the late volume of Essays Physical and Literary, at Edinburgh. In these experiments a spirit-thermometer was immersed in spirit of wine, and being suddenly retracted, was again exposed to the air; and as the spirit of wine adhering to the glass evaporated, the spirit contained within the thermometer was observed to subside. Now as the difference of the refraction of spirit of wine and glass is exceedingly minute, compared with the difference of refraction of spirit of wine and air; we may consider, in the above experiment, the heat to be communicated to the thermometer only at its surface: but here the adherent fluid escapes as soon as heated; by which means the glass, and its contents, are deprived of that constant addition of heat, which other bodies perpetually enjoy either from the sun-beams immediately, or from the emanations of other contiguous warmer bodies; and must thence, in a few minutes, become colder than before.