8 
SIR WILLIAM CROOKES ON THE PREPARATION OF 
cut off some of the heat rays which the glass transmitted. If, on the other hand, I 
select a piece that allows the red rays near the lines B and C to pass (as shown on the 
photo-spectrograph), some of the action on the radiometer balance would be due to the 
luminous red rays, and it is not advisable to entirely cut these off as the residual 
colour thereby might be affected. The plan ultimately adopted was gradually to 
increase the thickness of the mica until the photo-spectrograph showed that the 
visible red had just ceased to affect the plate. After experimenting on some hundred 
kinds and thicknesses of brown and black mica I succeeded in getting a piece which 
appeared to satisfy all requirements ; the examination of the different kinds of glass 
made in my laboratory during the last four years has been carried on by help of this 
specimen. 
Measurement of Infra-red Radiation. 
Before 1 had tried many experiments with the thermopile and black mica 
another difficulty cropped up. The heat radiated from the Nernst glower gradually 
warmed up the apparatus, and unless a long time was allowed between each observa¬ 
tion I never coidd get the index ray of light to return to zero. This and the little 
sensitiveness of the instrument used induced me to try a mercurial thermometer in 
place of the thermopile, all other arrangements being as before. The thermometer 
was specially constructed with a concave bulb coated with carbon from burning 
camphor. It was divided into tenths of a degree, the scale being about 9 mm. to a 
degree. 
To test, a selected plate of black mica was permanently fixed in the apparatus 
between the bulb of the thermometer and the slide containing the glass. ' First of all, 
upon lighting the Nernst lamp, the rise in temperature was taken, having only the 
mica plate between the lamp and the thermometer. The thermometer was observed 
from a distance through a cathetometer. It was allowed to rise about 1° C., and then 
with a stop-watch the further rise in sixty seconds was observed. A mean of four 
observations in this way gave a rise in sixty seconds of— 
175 
1'60 
1725 
F675 
mean 1‘683. 
In making a determination of diathermancy, glass plates 2 mm. thick were prepared 
and placed close to the thermometer bulb in a holder having a -g-inch hole; ten 
observations were made and the mean taken as the value, which was given in percent¬ 
ages of F683. 
Thus with glass No. 129 ; mean rise in sixty seconds = 0'26° C., 
0'26 x 100 
15-4° C. 
1'683 
