54 PHILOSOPHICAL TRANSACTIONS. [aNNO 1791. 



utmost limit of regular divergency seems to be about 5 or near 6 inches; above 

 that they are unsteady and disorderly. The pith balls are near -^ of an inch in 

 diameter, suspended by very fine flaxen threads 5 inches long. 



This apparatus requires a constant attention, especially during a disturbed state 

 of the atmosphere. From the room in which the apparatus is placed, Mr. R. 

 was seldom absent 1 hour, excepting the time of sleep; but when he leaves it, 

 the last thing he does at night is to examine the state of the electricity, and, if 

 the rod is unelectrified, he then places the Leyden bottle on the table p, with its 

 knob nearly in contact with the ball l. The next morning, if he find this bottle 

 charged, he writes the kind of electricity it is charged with against the day in the 

 journal, and adds, by the night bottle. 



Lastly, it may be useful to observe, that he has always found the lower though 

 uninsulated part of the apparatus (viz. the metallic connection of the bell n 

 with the moist earth) to be in a contrary state of electricity to the upper and 

 insulated part, where the pith balls k are suspended. Having made a memoran- 

 dum of the several thunder-storms which have happened in divers parts of this 

 island, according to the information by letters, and from newspapers, he thought 

 it useful to insert them in this journal, to show whether some contemporaneous 

 appearances in his apparatus might not be attributed to them. This seems evi- 

 dently to have been the case on the 3d of September. 



Then follows the journal, registering in columns, for every day in the year, 

 the several phenomena, viz. the wind, the barometer, thermometer, the electric 

 sparks, the electricity positive and negative, with the state of the balls and of 

 the weather. 



After the journal is collected the following monthly account of electrical sparks, 

 and of positive and negative electricity, as indicated by the pith-ball electrometer, 

 and sometimes by only flaxen threads without balls to them. 



Number of clays in each 

 nioiitli ill which sparks 

 were perceived. 

 Times. Times. Days. 



23 days of May, 17S9, 1 p^^.^;^,^ ^ . 



8 days of May, l/yi»,J " •' 



June Positive 32 Negative 36 12 



July Positive 13 Negative 22 12 



August Positive 19 Negative IJ) p 



September . . Positive 9 Negative 23 7 



October ....Positive 17 Negative 7 7 



November . . Positive 12 Negative 8 S 



December . . Positi\ o 12 Negative (i 7 



January .... Positive 26 Negative 4 13 



Febiuary . . . Positive 26 Negative 3 



March Positive 30 Negative 1 3 



April Positive 28 Negative 12 3 



241 156 98 



