A, W. Wright—Spectrum of the Zodiacal Light. 41 
he room was not lighted, ex- 
cept by the diffuse illumination from the sky, and care was 
line, which is sometimes visible, is constantly present and be- 
longs to the zodiacal spectrum; third, to discover whether 
polar aurora. A record was kept of the circumstances of the 
different observations, and of the state of the sky at the time 
they were made. 
it cannot be regarded as belonging to the zodiacal light. In 
its general appearance the spectrum is not different from that 
of faint daylight or. of twilight. It extends from somewhat 
elow D to near G. In order to fix its dimensions more defi- 
nitely, an extended series of observations was ma e, and the 
limits in both directions were determined with great care, 
bility were afterward read off on the scale. These varied 
Somewhat with the state of the atmosphere, but the determina- 
tions made on the best nights agreed very well with each other. 
For the purpose of comparison, those were selected which were 
found on a few of the clearest nights. The width of the slit, 
in all these cases, was ten divisions of the scale, or 1°22™™- 
Adding ten to the readings at the lower or red end, as a correc- 
tion for the breadth of the slit, the numbers are as follows: 
Lower limit, 54, 52°5, 57°7, 57°5, 50; mean, 54°35. 
Upper limit, 123°5, 119°5, 120, 124, 120; mean, 121-4. 
Since the intensity near the limits declines by almost insen- 
sible gradations, and the circumstances of the observations at 
different times, as atmospheric conditions, sensibility of the 
ari a 
