122 Proceedings of Royal Society of Edinburgh. [mar. 7 ri 
deposited. Though in a general way this may be so, yet there are 
certain differences in the conditions, and the manner in which the 
vapour is condensed at the different temperatures, which seem worth 
referring to. 
If we examine a surface, such as a sheet of glass, exposed hori- 
zontally near the ground on a dewy night, we shall generally find 
that the windward edges are dry. This indicates that the air itself 
is not cooled to the dew-point, though the surfaces of bodies exposed 
to radiation are, and the air has to travel some distance over the 
cold surface before its temperature is reduced to the dew-point. If, 
however, we examine this same surface when the temperature is low 
enough to cause the deposited moisture to form hoar-frost, we shall 
frequently find a marked difference. The sheet of glass is generally 
not only covered with the deposited vapour up to the windward 
edges, but the deposit is heaviest along these edges, the ice 
crystals growing furthest out in that direction. This peculiarity in 
the deposition of the hoar-frost may also be observed on almost all 
objects — such as branches of trees, iron fences, &c. The heaviest 
deposit will often he found on the windward side and not on the 
top, where we might expect to find it, owing to the stronger radia- 
tion from that surface. 
The question then is, What is the cause of this difference ? Why 
should no vapour be deposited along the windward edges of surfaces 
when the temperature is above 32°, while the heaviest deposit is 
formed on these edges when the temperature is below the freezing- 
point ? The dryness of that part of the dewed surface where the 
air first touches it is caused by the air not being saturated, and 
requiring to travel some distance over the cold surface before it is 
cooled below its dew-point. When, however, hoar-frost is forming, 
the air seems generally to act as if it were supersaturated : the 
crystals growing most towards the wind seems to indicate that the 
air does not require to be cooled before it deposits its moisture. 
But is it possible for the air to he supersaturated 1 Under ordinary 
conditions we know this is impossible. Owing to the vast amount 
of dust in the air there is always plenty of free-surface present to 
prevent this happening so long as the temperature is above the 
freezing-point. When, however, the temperature falls below this 
point, we have a much more complicated condition of matters. 
