DIURNAL RANGE OF THE BAROMETER IN FINE AND CLOUDY WEATHER. 453 



midnight, except at the tropical stations San Jose" and Hong Kong, where the minimum 

 is in the afternoon. At Ben Nevis and Jan Mayeii the maximum is delayed til] noon, 

 but the minimum is still at midnight. On cloudy days there is superimposed on the 

 ordinary diurnal curve another single period curve, the converse of the fine day one, 

 with its maximum at midnight, and its minimum at 8 or 9 a.m., but at San Jose and 

 Hong Kong the maximum is in the afternoon. At Ben Nevis and Jan Mayen the 

 minimum is delayed just as the maximum of the fine days is. Sodankyla on cloudy 

 days is an exception to this rule. 



The main difference between the weather conditions on fine and on cloudy days is, 

 that in the former, radiation acts to and from the surface of the ground, but on the 

 latter chiefly to and from the upper surfaces of the clouds, radiation also taking place 

 from the earth to the lower surfaces of the clouds. In other words, the radiant heat 

 of the sun on a fine day is mainly absorbed by the air near the surface of the ground, 

 and the heat so stored up is got rid of at night by radiation from the ground. But on 

 a cloudy day the sun's radiant heat is largely absorbed by the atmosphere in the upper 

 part of the cloud layer, and probably no small part of it is spent in evaporating water 

 from the cloud surface, thus diminishing the rise of temperature in that stratum of air. 

 Night radiation takes place from the same stratum, the upper cloud surface acting as a 

 radiating body, and doubtless also recondensing into itself the water evaporated during 

 the day. If, then, the changes on the diurnal barometric curve due to fine and cloudy 

 days respectively are to be explained as temperature effects, we must connect them, on 

 fine clays, with a large diurnal range of temperature in the lower air, while the clear 

 upper atmosphere has little change of temperature. 



On cloudy days, on the contrary, the range of temperature of the lower air is less 

 than the normal, and the changes in the upper air greater than is usual for this high 

 region. Therefore on fine days there is an expansion of the lower air while temperature 

 is rising, the effect of which is to lift the quiescent mass of air above it, thus giving 

 rise to a small increase of pressure in the expanding air ; but while temperature is 

 falling, and the lower stratum of air contracting, there is a decrease of pressure till the 

 upper mass of air has again sunk. On cloudy days, on the contrary, the lower stratum 

 of air is much less affected by the sun's heat. The rise of the temperature of the 

 ground and of the air immediately resting on it shows a heating effect of the sun, 

 either directly through the intervening cloud, or from the heated cloud itself, the chief 

 absorption of the radiant heat of the sun, however, takes place in the upper part of the 

 cloud stratum. The pressure of the upper air is thus lightened during the day by 

 the rise of temperature and dilution with water vapour from the underlying clouds ; 

 and as there is less pressure above to check its expansion, this decrease of density 

 appears as a diminution of pressure at the earth's surface. Conversely the cooling 

 and condensation of water vapour at night increase the density of this upper stratum 

 and add to the pressure below, thus causing the evening maximum of pressure on 

 cloudy days. 



