of Edinburgh, Session 1884 - 85 . 
115 
met with much success, for though, when the sky was clear, the 
light was rather brighter than it is here at present, yet I have seen 
it quite as bright here. Still the spectrum was fairly bright, and I 
don’t think that I could have helped seeing bright lines if there had 
been any to see. I got no photograph. I made, however, one 
observation, which seems to me to be of importance, and recalls a 
former observation made in the West Indies. We had sunset about 
6 o’clock ( m . m . t .), and by 7.15 or 7.30 all trace of the sunset 
colours (reglow and all) had disappeared, and the Z. L. showed 
very clear and a little brighter than the milky way. It was broad 
below, but much narrower above — being about 15° at the horizon 
and only G° at an altitude of 30°, and was distinctly visible up to 
about 60°. How the point that I noticed was this, that it did not 
seem to set at all, but only to fade out. Even at 10 p . m ., four hours 
after sunset, you could trace it up as high as before, and it had still 
the same shape — broad below and narrow above. This was not a 
solitary observation, but was repeated every night on which the 
moon and the mists — of which I had unfortunately too many — 
would permit of it. I suppose that if this is really the case, it must 
belong to the earth, and not to the sun. 
I did not confine myself to the Z. L., but did what I could at 
atmospheric lines and at atmospheric electricity. I made hourly 
observations of the latter from 6 or 7 a . m . to 10 or 12 p . m ., and have 
got a very fair approximation to the curve of variations for the day 
between these hours showing a maximum, very well marked, about 
2 p . m ,, and probably a minima near 7 a . m . and 7 p . m .; but the 
evening observations were so interfered with by mists that they are 
not of much use for giving points in the curve. But the mist 
observations seem to me by far the most interesting from another 
point of view, viz., that of the formation of thunderstorms. 
Dodabettah, being the highest point of the Hiiigres, was usually 
above the clouds in the afternoon, but at times mist was blown over 
from the comparatively cool easterly side, and was dissipated as it 
reached the crest of the hill, now in such a dissipating mist, the air 
‘potential ivas below the normal for that time of day. On the 
other hand, with a heavy condensing mist, the air potential ivas far 
above the average. It seems to me that this is direct evidence in 
favour of the condensation theory of thunderstorms. I have now 
an immense mass of electrical observations on hand, some of them of 
