230 1. C. Mendenhall— Coefficient of Expansion 
from imperfect calibration. Both standards had been kept for 
three months previously hanging in a room whose temperature 
never exceeded 80° C. e comparison therefore represents 
the thermometers had been exposed to 1: 0° C. I shall recur 
to this point in the continuation of the paper. 
[To be continued. } 
Art. XXX.—On the Determination of the Ovefficient of Ex- 
pansion of a Diffraction grating by means of the Spectrum; 
by T. C. MENDENHALL, of Tokio, Japan. 
In all precise measurements of wave-lengths, a correction 
for variations in the temperature of the grating is very essen — 
tial. If the ruling be upon glass the coefficient of expansion 
be ma 
concerning a ruling upon metal. Many of the beautiful grat — 
ings produced within a few years upon Mr. L. M. Rutherfurd’s 
rom the well-known equation 
A=s sin } 
in which 4 represents the wave-length, s, the distance Lepore: 
the centers of two adjacent lines upon the grating and 6 the 
angle of deviation, we easily obtain the following: 
se 
‘Tellier aces tien be Cake: one ga 
8 
from which it appears that, in order to determine the ¢0 
=-—cot bd5b 
