86 MR. T. HOPKINS ON IMPROVEMENT 
district around to be converted into vapour. Yet at the 
same time at more than 15° farther south, in the moist 
countries of Bengal and Birmah, the temperature of the 
wet grouud was seldom so high as 80° even when the sky 
was unclouded. Here we see that the solar heat which 
raised the temperature of dry ground in one part to 150°, 
raised that of wet ground in warmer latitudes not much 
more than one-half; but this difference evidently arose 
from the circumstance that much heat was taken from 
the wet ground by vapour which passed into the air. 
The solar rays make the surfaces of dry deserts, in warm 
latitudes, very hot, partly because dry air is a bad con- 
ductor of heat, and partly because the air does not ascend, 
as newly-formed vapour does, through another and a colder 
medium. Ascending currents of warm air are never found 
over hot deserts. Such parts receive and accumulate heat 
in the day and part with it during the night without 
creating material disturbance in the atmosphere. The 
great cold of the night, as compared with the heat of the 
day, has been long noticed in the deserts of Arabia, but no 
such atmospheric disturbances arise from the heat of those 
parts as are often experienced over certain tropical oceans 
that have a temperature below 80°, or over the compara- 
tively cool but wet countries of tropical parts. In some 
deserts the cold of the surface at night is sufficient to 
condense a part of the little vapour that is in the air into 
water or dew, but this water is generally reconverted 
into vapour by the morning sun, without enough being 
sent into the atmosphere to produce dense cloud and heat 
the air ; therefore, although the barometer might be raised 
by the morning evaporation, very little mid-day vapour 
can rise from the already dried ground. In parts of the 
great African desert where little or no rain falls, much 
dew is formed on the ground at night by cold, and where 
the land is level, it is probable that the evaporation of this 
