1863.] 693 



tures beside a thermometer which has been tested at Kew Observa- 

 tory ; and as the errors at other points of the scale were found not 

 to differ much from those at 32, I have applied a uniform correction 

 of *4 to the readings of the dry-bulb, and of '7 to those of the 

 wet-bulb. In the tabulated observations 1 have entered the cor- 

 rected reading of the dry-bulb thermometer, and the corrected 

 difference between dry and wet. 



My observations of cloud are recorded in the usual way, the figure 

 in the column " Amount " denoting the number of tenths of the sky 

 that are covered with cloud. In the column "Kind" the abbrevia- 

 tions ci., cu., st., nim., are used to denote cirrus, cumulus, stratus, 

 nimbus. 



The direction of the wind has been inferred from observations of 

 smoke and clouds, or from other obvious sources. The force of the 

 wind has generally been set down by estimation, 1 denoting a light 

 breath, 2 a moderate wind, 3 a rather high wind, 4 a gale, 5 a vio- 

 lent gale. Where the velocity is given in inches per hour in the 

 column " Remarks," it has been observed with a hemispherical cup- 

 anemometer. 



20. The entries of electricity are in three columns. The first 

 contains the mean of all the observations which compose the group, 

 these observations being generally taken at intervals of a minute ; the 

 second column contains the highest potential observed (that is, the 

 strongest positive or the weakest negative) ; and the third column 

 contains the lowest potential observed (that is, the strongest negative 

 or weakest positive). 



When a greater number of observations have been taken consecu- 

 tively, they have been broken up into groups ; and in grouping I 

 have been careful, as far as practicable, to avoid including positive 

 and negative in the same group ; but sometimes, when the electricity 

 was weak and oscillating, I have allowed them to enter the same 

 group ; and in these instances I have obtained the mean by dividing 

 the algebraic sum by the number of observations. As regards the 

 number of observations to be combined in one group, my arrange- 

 ment has been somewhat irregular, depending generally on conve- 

 nience as regards the pages of my observation-book. 



The time entered corresponds nearly to the centre of each group. 

 It is sometimes given in hours and quarters, sometimes in hours and 

 minutes. 



