On some Meteorological Conditions' controlling Nocturnal Badiation. 391 
passed the radiant energy from his source of heat through a given 
quantity of aqueous vapour (at a temperature maintained constant) con- 
fined, suppose, in a telescopic tube, he would, if my argument is sound, have 
found a greater or less effect upon the needle of his galvanometer accord- 
ing as the tube was lengthened or closed up ; since the relative humidity 
would be less or greater as the tube was made longer or shorter, while the 
actual quantity of moisture would have been unaltered. Tyndall, as is 
known, ascribed his results solely to the pressure of the water vapour in 
the given absorbing space," omitting to notice that the relative humidity 
was greater as the quantity of water vapour used was greater. 
Again, Magnus j placed his source of heat in contact with his absorbing 
material, thus raising its temperature and lowering its relative humidity 
in the considerable interval that was allowed to elapse before the galvano- 
meter needle attained to its stationary " position. 
Whenever dew begins to form, condensation invariably commences on 
the lower of the two thermometers. Because, therefore, of the latent heat 
liberated at the lower level, and because of the lesser radiation of water as 
compared with the ordinary solid matter of the earth's surface,! it might 
fairly be expected that the commencement of condensation in the form of 
dew would check the nocturnal cooling of the earth's surface, and hence 
that the radiation-temperature gradient in the space between the two 
thermometers would become less steep. For reasons now to be discussed, 
however, my observations show no easing of this gradient by dew. 
Throughout the whole period of observations there were 27 clear evenings 
in which dew began to form at some instant between XX. and midnight. 
[There were many other evenings in which dew was already formed at 
XX. o'clock, but these are not brought into account since the point to be 
tested was what reduction of the radiation took place when dew began 
to form.] 
Now these 27 evenings of incipient dew show the following hourly 
gradients : — 
XX. XXI. XXII. XXIII. Mid't. 
6-9 6-8 6-6 6-3 6-4 degrees F. 
* See, chiefly, his two memoirs contributed to the Royal Society in 1862: (1) " On 
the Absorption and Radiation of Heat by Gaseous Matter"; (2) "On the Relation of 
Radiant Heat to Aqueous Vapour." Poynting and Thomson {Heat, 1904, p. 236) take 
the very sanguine view that Tyndall's conclusion is "confirmed by the observations of 
meteorologists on the greater or slower rate of cooling of the earth at night when the air 
contains less or greater amounts of water vapour." 
t Phil. Mag., August, 1861. 
\ When dew is forming there is, moreover, no evaporation whereby the temperature of 
the wet surface can be lowered. 
