614 
DR. J. T. BOTTOMLEY ON THERMAL 
emissivity at the temperatures and under the circumstances noted; but neither the 
absolute time nor the intervals of time appear in the final result. 
In order to prejoare the numbers thus recorded for calculation it tvas necessary to 
apply certain small corrections to the galvanometer readings. 
1. As the reflecting galvanometer and a plane scale were used, and as the angle 
through which the ray of liglit turns is double that through which the mirror and 
magnet are turned by the current force, a number proportional to n^jd^ must be 
subtracted from the galvanometer readings in order to obtain numbers proportional to 
the magnet deflections (or their sines or tangents); n being the observed deflection 
and d tlie distance of the scale from the mirror. To do this a table of values of 
#/{ 10152)® was prepared; but very convenient tables for this purpose were pub¬ 
lished in 1890 by Dr. Paul Czermak,"^ and ought now to form a part of the equip¬ 
ment of a physical laboratory. 
2. The second correction consists in the subtraction of a small quantity proportional 
to the square of the deflection. It is for the term in the thermo-junction formula 
which is proportional to the square of the temperature. This correction was also 
applied with the help of a table or with the slide rule. (See footnote to p. 603 of 
the paper.) 
These subtractions having been made, I obtain what I may call the “ proportional 
deflections ” and these stand opposite to the times observed. These numbers are 
proportional to the differences of the temperatures of the globe and the water-bath 
surrounding its enclosure at the corresponding times. The following copy from pages 
of my calculation book will show all details. 
* Footnote to jn 597 of paper. 
