The consequences of this property of _latent heat_ are very important. It is on this account that the changes now spoken of necessarily occupy a considerable time. Each part in succession must have a proper degree of heat applied to it. If it were otherwise, thaw and evaporation must be instantaneous: at the first touch of warmth, all the snow which lies on the roofs of our houses would descend like a waterspout into the streets: all that which rests on the ground would rush like an inundation into the water courses. The hut of the Esquimaux would vanish like a house in a pantomime: the icy floor of the river would be gone without giving any warning to the skaiter or the traveller: and when, in heating our water, we reached the boiling point, the whole fluid would “flash into steam,” (to use the expression of engineers,) and dissipate itself in the atmosphere, or settle in dew on the neighbouring objects. It is obviously necessary for the purposes of human life, that these changes should be of a more gradual and manageable kind than such as we have now described. Yet this gradual progress of freezing and thawing, of evaporation and condensation, is produced, so far as we can discover, by a particular contrivance. Like the freezing of water from the top, or the floating of ice, the moderation of the rate of these changes seems to be the result of a _violation_ of a law: that is, the simple rule regarding the effects of change of temperature, which at first sight appears to be the law, and which, from its simplicity, would seem to us the most obvious laws for these as well as other cases, is modified at certain critical points, _so as to_ produce these advantageous effects:--why may we not say _in order to_ produce such effects?