454 Observations on Twilight. [August, 
Now the arc in the western sky which bounds the illu- 
minated segment begins to be very distinctly perceptible. 
When the sun is 6° below the horizon it is about at the alti- 
tude of 75 0 . For 170 degrees along the western horizon 
extends the orange-yellow stratum, of 2 5- 0 in height ; above 
it lies a green band of double the width, whilst the rest of 
the illuminated segment is a whitish blue. The boundary 
arc sinks rapidly downwards, and when the sun is io° below 
the horizon it is scarcely as many degrees above. Sometimes 
a second, but fainter, rosy light can be seen in the western 
sky, but of less extent and shorter duration than the first. 
When the sun has sunk to ixf every trace of red colour has 
vanished from the western sky, and the sharply- defined 
luminous segment hastens down to the horizon. 
Still with the descent of this segment the phenomenon 
has not always reached its end. There sometimes occur, 
later whitish colourations in the western sky, and whitish 
segments, which, however, never display any trace of colour. 
On one occasion a luminous appearance of 6° altitude was 
observed when the sun was already 23 0 below the horizon. 
The phenomena observed during the dry season in Spain 
are inferior to those described, both in intensity of colour 
and in extent. 
The phenomena of the dawn observed in Germany differ 
widely from those recorded in the south of Spain. In the 
latter country the phases are more distinct than in Germany. 
The succession in time is more definite, and the strata of 
different colour are more distinctly marked off from each 
other; the boundary between the light segment and the 
dark sky is very decided. 
On the other hand, with the exception of the counter- 
dawn, the duration and the extent of the colouration are 
greater in Germany. As regards the colours, there prevails 
the difference that in Spain green is almost invariably seen 
in all shades, which is rarely the case in Germany. On the 
other hand, in Germany the red colourations of the Spanish 
heavens take a more flesh-coloured and purple tone, and the 
violet colours seem deeper. Further, in Spain there exists 
a greater difference between the twilight phenomena of 
summer and winter than in Germany. 
In England we need scarcely say that, as a rule, the phe- 
nomena are too much complicated by smoke and cloud to 
admit of useful study. 
At an earlier date — in 1864 — Prof, von Bezold made a 
careful study of the phenomena of ordinary twilight, and 
published his results in " Poggendorff’s Annalen” (vol. cxxiiii., 
