AND HORTICULTURE. 25 
contemporaneous on both fields, so that regarding these 
no inequality was observed. 
The influence machine was at work all the time it could 
be charged from June 2nd to 18th, and likewise from 
September 6th to 13th. During the rest of the time it 
was at work, with small changes, from 7 p.m. to 7 a.m. 
. Average No. of No. of 
Period. hae daily. - Length of spark in mm. days. 
Day. Night. 
June 2-18...) 17-0 day and night | 1°5 to 2*2 o's torg | 8 
Sept. 1-13...) 12°5 night and day 15 or, 03] Sil} 
@ 
9 19-30...] 10°5 night 0°55 15 o'2 ,, 08 = he 
July 1-15...) 86 ,, 15 seldom o°3 ,, 0o'5 | 4t z 
° 
a 
) 16-31... 10°7,, rg taoe oF q o8| at], 
seldom 3 
Aug. I-15...) 7°2  ,, Very short 18/3 
+ 
» 16-31...) 76 ,, — or, o5 | 3t]e 
d 
a 
The spark-length between the discharging balls was 
measured at the time of setting in motion, when it was 
stopped, and sometimes while it was going on. In general 
this length is very variable, as will be seen from the above. 
The spark-length is the measure of the potential to 
which the wire net can be charged, and depends on 
external meteorological conditions, the most significant 
of which is the humidity of the air. The faster the elec- 
tricity is streaming out from the wire net the lower is the 
potential to which the wire net can be loaded by means of 
a machine of given strength. The electricity meets in 
such a case with smaller resistance in the air-layer between 
* The sky often clear. t+ Clear and half-clear. { Cloudy and clear. 
§ Misty. {| Cloudy and half-clear. ‘J The length of spark was measured 
in millimetres =o'o4in. 
