But it might seem that since objects in the upper regions of the air part with their radiant heat more freely than objects on the ground, the former should be more copiously moistened with dew than the latter. That the fact is exactly the reverse is thus explained. The cold which is produced by the radiation of heat from objects high in the air is communicated to the surrounding air, which, growing heavier, descends towards the ground, its place being supplied by warmer air. Thus the object is prevented from reducing the air in its immediate neighbourhood to so low a temperature as would be attained if this process of circulation were checked. Hence, a concave vessel placed below an object high in air, would serve to increase the deposition of dew by preventing the transfer of the refrigerated air. We are not aware that the experiment has ever been tried, but undoubtedly it would have the effect we have described. An object on the ground grows cold more rapidly, because the neighbouring air cannot descend after being chilled, but continues in contact with the object; also cold air is continually descending from the neighbourhood of objects higher in air which are parting with their radiant heat, and the cold air thus descending takes the place of warmer air, whose neighbourhood might otherwise tend to check the loss of heat in objects on the ground. Here, also, we recognize the cause of the second peculiarity detected by Aristotle—namely, that dew is only formed copiously in serene weather. When there is wind, it is impossible that the refrigerated air around an object which is parting with its radiant heat, can remain long in contact with the object. Fresh air is continually supplying the place of the refrigerated air, and thus the object is prevented from growing so cold as it otherwise would.