252 O. N. Rood—Nature and Duration of 
second. This rate was accurately ascertained by causing the 
lowest wheel to wind up a fillet of paper, on which secon 
marks were at the time impressed. 
scribed were quite invisible. 
Micrometers.—The general mi tri 9 
in part second of this paper, under the head “ Micrometer,” was 
sometimes employed, the strip of white paper being replaced 
by a small cylindrical mirror, situated under the spark. A 
tube of glass filled with mercury was used for this purpose. 
On the other hand, sometimes an achromatized cale spar pmsm 
was employed, being located between the eye of the observer 
and the image on the ground-glass; but for actual work, an- 
other and much simpler proceeding was found to furnish results 
equally reliable, with an expenditure of less time and patience. 
his method, which at first sight may seem rude and unreliable, 
consisted merely in holding on the ground-glass along the 
£1 awe ee 
electrical phenomena of an unvarying nature, the other methods 
would probably have yielded more accurate results ; but 12 the 
present case, where the actual variations often rose as high as 
50 per cent of the quantity involved, this mode, as the follow 
ing experiment shows, is all that could be desired. 
x strip of paper was formed on a plate of grouse. 
the hands of the experimenter. The disc was 
now set in rapid rotation and illuminated with 
simple (not multiple) electric sparks, an i 
tempts were made to measure the length of the 
image of the strip of paper with a pair of ier 
passes, in the manner just indicated. The room was only sutl 
ciently light to enable me to perform this work, and the meas 
