LABORATORY OPTICAL PYROMETER 
73 
the lens, for the telescope, and for the lamp. It must also be pos- 
sible to rotate the telescope through a small angle in a vertical plane 
through the optical axis of the system ; the lamp should be capable 
of being adjusted in a horizontal direction perpendicular to the 
optical axis of the system. 
After the pyrometer has once been calibrated the relative posi- 
tions of the parts must remain fixed, otherwise the temperature 
determinations will be incorrect. It is, however, allowable to move 
the different parts relatively to one another if the angles u and v 
are kept the same. Usually it will be found more convenient to 
recalibrate the instrument than to make adjustments which will 
keep u and v constant. The focusing of the pyrometer on the 
background should be accomplished by changing the distance be- 
tween A and B. 
Absolutely-monochromatic screens for use at G cannot be ob- 
tained, and filters which are very nearly monochromatic transmit 
so little light that they can be used only in measuring the high 
temperatures. “Partially-monochromatic” glasses, which absorb 
so little of the incident energy that they are usable even at rela- 
tively low temperatures, are available. Such a filter can be em- 
ployed if the correct one of all the transmitted wavelengths is se- 
lected. This wavelength is called the “effective-wavelength^,^ 
of the filter. It is such that, for any definite temperature interval 
for a particular source, the ratio of the radiation intensities for 
this wavelength is equal to the ratio of the integral luminosities 
through the screen used. The effective wavelength of a filter 
varies with the temperature to be measured, but for most work, 
if the average value is used, the error will not be too large. For 
example, an error of 0.001 micron at 2400° K would cause an er- 
ror of about 1.2° and at 3000°K an error of about 3° if the filter 
transmitted red light. Filters transmitting red light are most con- 
venient for general use since red light of a sufficient intensity is 
emitted by a hot body at a lower temperature than that at which 
light of the shorter wavelengths is radiated. The least relative 
change of brightness which can just be detected is about the same 
for red as for blue light but the change in intensity for a given 
small temperature change of the radiator is greater for blue than 
for red. Also, diffraction effects are less troublesome when one is 
using light of the shorter wavelength. Therefore greater accuracy 
is obtainable when light of shorter wavelengths is used, but the 
lowest measurable temperature is higher. 
Perhaps the most convenient form of absorbing screen to be 
