98 
L. Sehwendler— On 'Electric Light Measurements. [Makch, 
Hence if the two lights under comparison contain the different co¬ 
loured rajs in different proportions, it may be said that by adjusting balance 
by means of the red glass, the ratio must become erroneous. 
However, actual experiments proved that this is practically not the 
case. If there is a difference it is quite within the limits of errors caused 
by the considerable variation in the intensity of electric lights. 
Two electric lights compared with eacli other by means of the red 
glass and without it, gave practically the same ratio. 
Further, the same electric lights compared each with a standard 
light (either a standard-candle or a platinum-light standard, to be de¬ 
scribed further on), and by adjusting balance through the red glass gave 
a ratio sufficiently close to the one found by direct comparison of the 
two electric lights. I can therefore recommend this photometer for 
practical use. 
The Dynamometer, described in Appendix (IV). When transmitting 
force by means of a strap from a driving pulley to a driven pulley, it w T ill 
be noticed that that side of the strap which runs towards the driving 
pulley has always a greater tension than the other side of the strap which 
runs from the driving pulley. 
This difference of tension in the two sides of the same strap is pro¬ 
portional to the force transmitted, a well and long recognised fact. The 
difference of tension is measured either by a graduated spring or weight, 
and hence the force can be calculated if the constant of the dynamometer 
is known. 
From this and the speed the power consumed per second can be calcu¬ 
lated. Dr. C. W. Siemens called my attention to this very simple, con¬ 
venient, and accurate method of measuring force transmitted and power 
consumed. It is a method not to be found in any text-book, and was first 
suggested and employed by Mr. Hefner von Alteneck, of Berlin. 
A drawing to scale of the dynamometer, as actually made, is attached 
to Appendix IV. 
The measurement of speed. See Appendix V. In this investigation, 
where almost all results depend on the speed of the induction cylinder, 
it becomes a matter of importance to select the proper apparatus for 
measuring that speed accurately. 
I have employed three essentially different instruments :— 
The Pocket-Counter —an instrument well known and extensively used. 
The Mercury Speed Indicator by J. M. Hapier. The speed is mea¬ 
sured by a column of mercury, which increases its length when the speed 
increases. It is a most convenient instrument. 
The Cycloscope, this is another very ingenious instrument for mea- 
