Rays emitted by Glowing Platinum. 4651 
to use Matthiesen’s formula for the change of electric resistance 
with the temperature, only showed the impracticability of this 
method. 
Only the middle portion of a glowing wire can be said to be 
of equal temperature throughout. If we measure the resistance 
of the wire when hot and cold (in itself no easy task), the 
change corresponds to a difference of temperature which gives, 
so to speak, the mean temperature of the whole wire; a quantity 
which must then be used, together with A and & (inner and 
outer conductivity of the metal), and with the dimensions of 
the wire, in the calculation of the distribution of temperature 
throughout its various parts. Aside from the difficulty of find- 
Ing an applicable expression for this distribution, our imper- 
fect knowledge of the quantities A and & for platinum, as fune- 
tions “3 the temperature, would render the calcuiation of doubt- 
ul value. 
The method finaliy adopted was to measure directly the 
expansion of the wire. By observing it from end to end with 
the leucoscope, while glowing, it was found that for a portion 
in the middle, about 60™ long, the light radiated was, for the 
Whole distance, of like character. This then was the greatest 
admissible length of the piece to be measured. In reality the 
Section chosen was much shorter (45™), so that certainly within 
its limits, only imperceptible differences of temperature occurred. 
degree of the platinum thermometer may be de ned as 
that change of temperature which causes in a platinum wire a 
linear variation of 1:1-00000866. Then for a wire 45™" long, 
one degree corresponds to an expansion of about 0004", and 
it was desirable in determining the temperature to be able to 
Measure its length to within a few ten-thousandths of a milli- 
meter. For this purpose I used a finely constructed Helm- 
holtz’s Opthalmometer; the following description of which is 
taken from Helmholtz’s “Handbuch der physiologischen 
Optik,” (p. 8). “The opthalmometer is essentially a teles- 
Cope arranged for short distances, before the objective lens of 
which two glass plates stand side b ide, sO that one-half of 
the lens looks through the one, the other half through the other 
Plate. When both plates are in a plane perpendicular to the 
axis of the telescope, there appears a single image of the object 
In view. Let them be turned a little, however, toward opposite 
sides, and the single image divides into two halves of a double 
Image; the distance between which increases with the angle 
between the plates. This distance can also be calculated from 
the angle which the plates make with the axis of the telescope.” 
f a ray pass obliquely through a glass plate its displace- 
Ment S will be, (fig. 4), : 
Am. Jour. Sct.—Tarrp p Sanne, Vou. XVIII.—No. 108, Dzc., 1879, 
