RADIATION IN ABSOLUTE MEASURE. 
431 
sending a known current through the radiation wire, and simultaneously measuring 
the current and the electric resistance of the wire. This arrangement is one of two 
described in my first paper on this subject. It has proved entirely satisfactory; and 
X have used the other method but little, except in some experiments on incandescent 
lamps, and in a special experiment described at the end of this paper.* 
The arrangement shown in fig. 1 is a Wheatstone’s bridge adapted to suit the 
purpose in hand. The wire w under experiment is represented in this diagram as 
stretched between the points c and d. In reality it is surrounded with a tube which 
is immersed in a w'ater-jacket, as will be described presently, ab is a frame in which 
there are stretched a large number of somewhat fine copper wires, sometimes 12 or 
15 , or more. These wires are stretched between two very stout copper bars, shown 
cross-hatched at a and b. The multiple copper wires offer but little resistance to the 
battery currents, and present so large a surface to the air that the heating due to 
any current they are called upon to carry is absolutely insensible. At G is shown 
the Thomson’s current galvanometer (of extremely small resistance, 5-5-0 ohm) which 
measures the current flowing in the branch abed of the Wheatstone’s bridge—that 
is, the current passing though the wire w. 
The other branch of the Wheatstone’s bridge is a long fine platinum wire, 
specially drawn, and excellent as to uniformity. It is so long and fine that the 
minute proportion of current passing through it is quite insufficient to warm it 
sensibly. It is stretched backwards and forwards on a board of polished pine; and 
a sliding contact piece e , guided by a V-groove, runs along one length of the wire, the 
position of the slider being read off on the scale ss. One electrode from the testing 
galvanometer T of the bridge is attached to the slider, while the other is attached at 
c in the other branch of the bridge. The battery consists of six large secondary 
cells arranged in series; and a Thomson’s rheostat wound with thick platinoid wire 
is used to control the current. 
It will be seen that the arrangement here described gives a Wheatstone’s bridge 
in which the only part sensibly heated is the radiation wire w; and that its resist¬ 
ance at any temperature can be measured in terms of the other three parts of the 
bridge. Every connection in the four branches of the bridge is soldered. 
The current galvanometer used during my earlier experiments was one of Sir W . 
Thomson’s graded current galvanometers with a magnet of low power to give 
sufficient sensitiveness. More recently I have substituted for the graded galvano¬ 
meter one of a different pattern, called by Sir W. Thomson a ‘lamp-counter,’ and 
designed for finding the number of lamps in use on an electric light circuit. To keep 
* The alternative method referred to in the text consists in measuring the current passing’ through 
the radiation wire, and, simultaneously, the difference of potentials between two points at known distance 
apart. (See figs. 5a, . 5b, Plate 24.) It possesses one great advantage, viz., the avoidance of the 
disturbance by conduction of heat from the ends, but it wo aid be difficult or impossible to use this 
method with a vacuum, except in an enveloioe of glass. 
