O. N. Rood—Duration of Flashes of Lightning. 167 
finally been found preferable. The rate of rotation was kept 
nearly constant by winding up. 
The flashes were usually multiple, and the duration of the 
components was often or generally quite long, being as great as 
zx of a second if not longer; the brilliancy of the ring of light 
was considerable, and showed no signs of falling off throughout 
its whole extent. It was again noticed that when the duration 
of the earlier components had thus been considerable, the last 
act (or certain acts) were instantaneous. Light or ten times it 
was fairly noticed that the components of certain flashes were 
to all appearance instantaneous, there being no distortion in the 
sely of the square. If its area had been increased by one 
, this could not have escaped my attention, and would have 
implied a duration of ;'.- of a second. A number of uncer- 
ponents, in some cases, of from ;1, to zt; of a nck 
finally in two cases the breadth of the square was distinctly 
doubled, giving a duration of about ;1, of a second. 
To the above, I must add several observations made by the 
naked eye on normal zig-zag flashes, when on five or six occa- 
slons the duration of the direct flash was estimated at not less 
than one second, the light seeming to pour steadily in a stream 
from the cloud to the earth. They correspond to that of Fara- 
day, referred to in the earlier part of this article. 
Tangi 
Up to others at least as great as ;!, of a second, and furthermore, 
the prolonged discharges of the induction co L 
appearance. The ‘ae was placed in front of the rotation 
