AT NIGHT, FROM THE EARTH, ETC. 
165 
of the observations have been made, I shall proceed with the explanation of the 
process for the formation of the following - tables. 
As the variations of the excess of the readings of the thermometer in the air, above 
those otherwise placed, were thus found to depend on the variations in the amount 
and in the height of the clouds only, the next operation was dividing the results of 
each month’s observations into five groups as follow. 
The first group was formed from the observations taken on nights which were 
cloudy throughout. 
The second group from those taken on nights which were principally cloudy, but 
the clouds frequently broken. 
The third group from those taken on nights which were half-clear and half-cloudy. 
The fourth group from those taken on nights which were principally cloudless, 
but yet clouds were frequent. 
The fifth group from those taken on cloudless nights, and in this way Tables LL 
to XCIX. were formed ; the successive results in each class being arranged according 
to their dates of occurrence. 
The examination of the columns in each month as to the distribution of the 
numbers, gives a good knowledge of the distribution of the clouds at night during 
the month. 
