72 
IOWA ACADEMY OF SCIENCE Voi.. XXVIII, 1921 
by the filament. .Such diffraction-bands make it almost impossible 
to get an accurate brightness-match. A telescope having an ob- 
jective lens of 25 cm. focal length and an eyepiece of 18 mm. focal 
length will be found satisfactory. 
The limiting diaphragms C and B must be of such a size that dis- 
appearances of the pyrometer-lamp filament against the back- 
ground-image are possible. Angle z> must be smaller than 
angle u. Both C and B should be as large as possible but their 
size must be consistent with ability to obtain disappearances — in 
order to minimize the diffraction effects at the edges of the aper- 
tures, and they should be as far from the lamp D as possible. A 
stop B, which may be just small enough to be satisfactorily used 
at the lower temperatures, may prove to be too large when making 
brightness-matches at the higher temperatures. The larger B is, 
the farther down the temperature-scale can be extended. There- 
fore it will be well to use at least two different stops when making 
the complete calibration of a pyrometer. 
A pyrometer lamp having a tungsten filament in a spherical bulb 
5 cm. in diameter will be found very satisfactory. The filament of 
such a lamp is usually in the form of a hairpin loop with a small 
kink at the top of the loop. This kink is useful in that it enables the 
operator to be certain that he is using the same part of the filament 
at all times. A lamp having a filament of 0.033 mm. in diameter is 
useful if the temperature of small objects, such as wires, is to be 
measured ; for larger objects a larger filament will be found more 
satisfactory. The diameter of the lamp filament should aways be 
considerably smaller than that of the background-image. When 
measuring the temperatures of wires it should be remembered that 
Lambert’s Cosine Law of Emission does not hold in general.^® 
For example, a tungsten wire appears brighter at the edge than 
near the center, while the reverse is true for a carbon lamp fila- 
ment. Consequently, for most accurate work, the pyrometer-lamp 
filament should be placed parallel to the background-image. If the 
background-image is large and the brightness-match is made at a 
point near its center, the filament may be placed so as to cross the 
image. For most purposes sufficient accuracy will be obtained 
even though the filament and image are crossed and the image is 
not large. 
The objective lens of the pyrometer, the kink of the lamp fila- 
ment, and the telescope must be very approximately axially 
aligned. Therefore provision should be made for a vertical ad- 
justment and for a rotational adjustment about a vertical axis for 
