Condensation of Atmospheric Humidity on Solid Surfaces . 85 
degree of cold, previous to any deposition of moisture, in the 
form of dew. When it was ascertained that the wool had not 
acquired any additional weight, no farther attention seems to 
have been paid to it ; and as the experimenter had a previous 
conviction, that evaporation had no influence in producing the 
diminished temperature of the wool, the loss of weight must 
have been very apparent indeed, that would have arrested his 
attention. But it is well known, that the evaporation of a 
very minute quantity of moisture is sufficient greatly to re- 
duce the temperature of the evaporating surface. And we may 
also remark, that the greatest degree of cold always takes place 
on those evenings when dew is latest in forming ; that is, when 
the air is driest, and, consequently, when evaporation is necessa- 
rily most active. It must also be observed, that, in performing 
experiments with a nicely adjusted balance, even in a close room, 
accurate results are not to be obtained without considerable 
trouble. If, then, such an instrument be employed in the open 
air, on a damp evening, or in a cold benumbing state of the at- 
mosphere, considerable inaccuracies must be almost inevitable. 
On another occasion, Dr Wells informs us, that , 66 on the 25th 
of January, the ground being covered with snow, during eight 
hours that I attended to my thermometers, the whole sky was 
constantly overcast with clouds. The atmosphere was, for the 
greater part of that time, very still ; and & thermometer on the 
snow was generally about 2° lower than another in the air. 
That this was not owing to evaporation, was proved by the ther- 
mometer on the snow always rising, from a half to a whole de- 
gree, whenever the air was a little moved, and falling the same 
quantity as soon as a great stillness again took place.” Far from 
proving, however, that the reduction of temperature was not the 
effect of evaporation, this observation will be found to furnish, 
if not a proof, at least a strong argument, in favour of that ex- 
planation. When the air was very still, that is, without a per- 
ceptible progressive or undulatory motion, the evaporation that 
was going forward at the surface of the snow carried off a 
greater quantity of heat than was communicated by the conti- 
guous air. Hence the snow became colder than the air a short 
distance above it ; an equilibrium being on this occasion esta- 
blished, when the temperature of the former was reduced about 
