NIGHT-SKY RADIATIONS 
small, rarely as much as a factor of 2. During the full 
nighttime hours there appear to be many small irregular 
and complex intensity fluctuations which vary with the 
INTENSITY 
3000 
4000 
5000 
WAVE LENGTH (A) 
Fig. 2.—Spectral energy distributions of night-sky light. 
(After Babcock and Johnson [3].) 
6000 
place in the sky and which are different for the various 
wave lengths. There are strong intensity changes during 
twilight. 
Diurnal Variations. In France, Dufay and Tcheng 
Mao-lin [10] obtamed 588 spectrograms during 189 
nights from October 1940 to January 1944. From these 
spectrograms systematic measurements were made of 
the intensity variations of the oxygen lines 5577 and 
6300 A and of the sodium pair 5893. The oxygen line 
5577 A usually showed a weak maximum around mid- 
night (not more than 40 per cent) often obscured by 
fluctuations; 6300 and 5893 A merely weakened slowly 
during the night, the enfeeblement for 6300 being 
greater than that for 5893 A, which amounted only to 
10 or 15 per cent. In Russia [23] photocell measurements 
of radiation in the infrared region 850 to 11000 A indi- 
cated that the intensity was at a maximum around 
midnight, being about twice as great then as at 10 
P.M. and 2 A.M. 
Twilight Effects. The green line 5577 was observed to 
exhibit no twilight enhancement. On the other hand, 
the sodium yellow emission 5893 dropped suddenly in 
intensity to one per cent of itsformer value at about ten 
minutes after sunset [7] with a corresponding recrudes- 
cence near dawn. Likewise the red oxygen lines 6300 
showed a similar but slower change during twilight 
[13]. The strong NZ flash at twilight discovered by 
Slipher is intense in aurorae but absent from the 
night sky. 
Annual Variations. In temperate latitudes 5577 has 
slight maxima in October and February, and a 27-day 
variation in intensity, hence a faint correlation with 
solar phenomena; 6300 and 5893 have maxima near 
midwinter and minima in summer of amplitude factors 
343 
of about 2.5 and 5, respectively; they seem indifferent 
to solar phenomena [2, 10, 12]. 
Latitude Variations. A survey [17] of the visual bright- 
ness of the night sky in various latitudes was carried 
out with standardized visual photometers which had a 
field of view about 11° in diameter. Visual measurements 
of the night-sky luminosity refer mainly to the strong 
atomic oxygen line 5577. This is because, by multiplying 
the night-sky spectral energy curve of Fig. 2 by the 
sensitivity curve of the dark-adapted eye, one finds 
that 5577 has the main effect (8/10) with only minor 
influence from the oxygen red lines 6300, 6363, the 
sodium yellow lines 5896, 5890, and the molecular 
bands in the blue and green. 
Observations were made at stations in Bocaiuva 
(Brazil), Bikini, Maryland, Whiteface Mountain (New 
York State), and Greenland, at latitudes —17°, +12°, 
+39°, +45°, and +63°, respectively. The results are 
plotted in Fig. 3, in which the ordinates are the night- 
S80 ms TAR LLL ALA A 
myLE 
BE (oe) 
ee ® ® m 
E ¢ le) 
< @ 
ia) 
@ 8 x 
2 x 
x x 
100%— x x 
A WHITEFACE MT 
O BOCAIUVA O BIKINI 
@ MARYLAND, GLEAR @® GREENLAND FEB, MAR,1949 
E X MARYLAND, HAZY 
o° [OSM 2 OSES OI OS OG OS Nm Onmc Omen Or 
ZENITH Z HORIZON 
Fic. 3.—Average night-sky brightness at several 
localities [17]. 
sky brightness B in millimicrolamberts (mul; 1 
myuL = 10-9 lambert = 2.96 X 1077 candle ft~*) and 
the abscissas are the zenith angle Z of the place in the 
sky. Each point of Fig. 3 was the average of values 
observed in several directions for several nights when 
no polar aurorae were visible [17].? From the data of 
Fig. 3 it was concluded that there were no changes 
with latitude in the visual night-sky brightness which 
could not be attributed to variations in haze. 
In connection with the geographical distribution of 
night-sky intensity, Farnsworth and Elvey [13] con- 
cluded that their photographic observations with a 
glass-prism spectrograph indicated no major differences 
at Bosque Alegre, Argentina (32°S), Portrerillos, Chile 
(27°S), and in the southern New England states; their 
data were too meagre for much generalization. Rayleigh 
[21] concluded from his visual observations that the 
average intensities of the blue, green, and red spectral 
regions at Terling, England (51°N), Capetown, Africa 
(34°S), and Canberra, Australia (35°S) were of com- 
parable magnitude, but tended to be highest at Terling 
2. Data for latitude +-63° were observed in February and 
March, 1949, by personnel at a United States Air Force weather 
station in Greenland using a Naval Research Laboratory visual 
low brightness photometer. 
