248 The Data on which we have to depend 
such lands often drain large tracts of country, but I cannot 
understand how Mr. Thom should imagine that fifty-four 
inches of rain fall in Paisley. This town is only seven miles 
distant from Glasgow, and, according to the meteorological 
tables, the rainfall for this city is twenty-one inches. 
Mr. Thom’s observations and experiments, therefore, could 
not have been conducted with much regard to scientific 
accuracy, when he computes the available rainfall at thirty- 
nine inches out of an annual rainfall of twenty-one inches, 
and they contrast rather singularly with Dr. Thomson’s 
observations and experiments, which give four inches of 
available rain out of twenty-one inches for the Clyde district. 
Since Dr. Wells published his well known Essay on Dew, 
his theory of its formation has been almost universally re-. 
ceived as correct. 
The production of dew occurs in the following manner. 
The quantity of aqueous vapour that can exist in the atmosphere 
depends entirely on temperature. 
During a clear calm night, all bodies that are fully exposed 
in the air become more or less rapidly cooled by radiation of 
heat from their surface. The air in contact with such bodies 
suffers a corresponding loss of heat, and, as soon as its 
temperature reaches the dew point, the moisture, which can 
no longer retain the form of vapour, is condensed in the form 
of dew. Thus, those bodies which radiate most heat and 
conduct least, condense most dew, and it is found that all 
bodies which are good conductors and good reflectors of heat 
from their surface, are bad radiators. 
The metals, therefore, condense dew very sparingly. Water, 
though a bad conductor of heat when applied to its surface, is 
from the extreme mobility of its particles, the most rapid 
conductor of heat and cold, when these are applied with due 
regard to its peculiar laws. 
Water in this sense may be regarded as strictly analogous 
to the metals, and, being a good conductor and reflector of 
heat, it is necessarily a bad radiator, and the dew is not formed 
on any surface whose temperature is not cooled by radiation 
below the dew point, which ranges from 5° to 20° below the 
temperature of the air. Unless, therefore, the surface of 
water be cooled by radiation below the dew point, it is quite 
clear that no dew can be condensed but its density, which 
varies with every change of temperature, and its fluidity 
operate to prevent any reduction of its temperature until the 
whole mass is similarly affected. 
The temperature of water is thus very slowly reduced by 
