
ON ELECTRICAL OBSERVATIONS AT KEW. 129 
sions of the observations, all readings above and including it being regarded 
as high, and more or less measuring the electrical tension of aqueous vapour 
either invisible or condensed ; and all readings below it being regarded as 
more or less measuring the tension of “atmospheric electricity.” Of course 
this method is entirely tentative ; the separating point 60 div. has been arbi- 
trarily fixed, and, as before observed, it is not to be expected that the curves 
furnished will be true representatives of natural phenomena, when we come 
to contemplate the two different sources from which the conductor is sup- 
posed to be charged ; nevertheless it may not be without its use in assisting us 
to devise some mode by which the two tensions may be effectually separated, 
either by some subsidiary observations and computations by which the elec- 
trical tension of the aqueous vapour may be disengaged from the aggregate 
tension as exhibited by the electrometers, or by directly observing the elec- 
trical tension of the vapour itself. 
TaBLe XIX. 
Mean diurnal period of Jow tension 
= for the months of January, Fe- 
bruary, March and April 1845. 
4A 
10 A.M. 
10 P.M 
A 
Period. | Jan. | Feb. |March.| April. 
a | | | ee | ee 
Low. 
Feb. ae Qam......| 114 | 195 | 30:6 | 17-7 
March, Low. 8 am......./ 341 | 41-0 | 49-8 | 40°3 

April. Low. 2 p.m. ..e..| 411 | 57-2 | 49°7 | 37°3 
The above table and curves are 
Feb. 60 div. intended to illustrate the separation 
of the readings into those of high and 
Below low tensions. The table contains 
March. 6odiv. the diurnal periods for the first four 
months of the year 1845, and the 
. four upper curves are the projections 
Below of these periods on the same scale as 
the curves deduced from all the ob- 
servations. The four lower curves 
exhibit the diurnal period for the 
same months as deduced from the 
a readings below 60 div. The greater 
uniformity of the lower curves, especially of January and February as com- 
pared with the upper, is very apparent. The curves appear naturally to 
divide themselves into two sets, the greater uniformity appertaining to 
January and February below 60 div., and to March and April, higher 
tensions than 60 div. entering as elements into the discussion of low ten- 
1849. K 
4 A.M 
