16,2 Fisher: Low-Sun Phenomena in Luzon 155 
mosphere to the eye of the observer, who sees it on its second 
tangency at the surface—three passages through the lower atmos- 
phere, with a total absorption many times that in one such 
passage, even though there did not intervene the weak reflection 
by the upper air. In fact, the light actually observed passes 
along paths considerably higher and more transparent than those 
of the elementary theory, and the computed lower limit is based 
on an angular depression of the sun considerably too small. 
Horizon observations of dawn are impossible at Manila, on 
account of all sorts of local conditions described above. But, 
on watching the passage of the various stages of dawn across 
the sky, I thought that the zenith passage of each stage might 
be observed with reasonable accuracy, a value deduced for the 
corresponding altitude (or depression) of the sun, and hence a 
better lower limit for the atmospheric extent. 
Besides the fact that the zenith is more generally visible than 
the horizon, zenith observations have the evident superiority 
over horizon observations that they involve one less absorption 
during tangential passage of light through the lower air. The 
light deviated vertically downward passes the absorbing layer 
by the shortest possible path, though probably a deviation or 
a of 90° produces less intensity than one which is very 
small, 
On the morning of April 18, 1919, looking in all directions 
from the summit of Mount Santo Tomas, Benguet, the low haze 
layer was fairly well defined. Arayat, an isolated volcanic cone 
in Pampanga Province, 1,024 meters high, did not project 
through it. The two eastmost peaks in Zambales Province, 
between the central plain and the sea, thrust dark blue summits 
above the haze, one only a little, the other quite a good deal. 
They are N egron, 1,590 meters, and Pinagtabo, 1,781 meters. 
So that, on this morning, the haze layer ceased at a height of 
about 1,500 meters. Assuming it uniform in height and turbid- 
ity—it really seemed somewhat more turbid in its upper parts— 
calling the earth’s radius 6,371 kilometers (this is the mean of 
vO equatorial radii and one polar), and neglecting refraction, 
ht reaching a point on the surface horizontally would have 
traversed the have a distance of 
V2xX1.5 6,371=138.3 kilometers; 
and if it is reduced to the fraction n of its intensity by absorption 
when passing through vertically, it would be reduced after hori- 
Zontal passage to n'*.*/1.5 of its original intensity. For the whole 
atmosphere it has been estimated that vertically incident light 
