Mxzsox.— On the recent Sun-glows, 869 
it was gradually drawing towards the pole, and that it was connected 
with the Sunda eruption. But that the exceptional matter in the upper 
air is vapourous at all seems to me very doubtful. I would remark here 
parenthetically that it is very easy for us to be wise when a great many 
facts and opinions have been given and collected; but very difficult when 
the phenomena are quite new and few accurate observations have been 
made, and we have had no opportunity of comparing notes with other 
observers. To say this much is due to those who, from their position, are 
expected at all times and at a moment's notice to give a correct opinion as 
to the cause of whatever baffles ordinary comprehension in natural pheno- 
mena. That they should occasionally make a mistake is to be expected, 
but that they have done so as far as this aqueous theory is concerned, I con- 
ceive, as yet, is certainly ‘not proven." That they have, in some cases, 
changed their opinion, shows the difficulty and uncertainty with which the 
whole subject is surrounded. 
However, there is another test of the “aqueous vapour" theory—the 
spectroscope. When the glow is examined by means of this tell-tale instru- 
ment, if there were aqueous vapour in any quantity in the upper air, 
you would have the solar spectrum with its Fraunhofer lines and 
certain black bands—wet weather or rain bands as they are called 
—intervening, even as when a ray of light is seen through straum 
(M. Jansen). Now, as a matter of fact, during the period of the sun- 
glows, whenever the spectroscope was employed, it was found that the 
rain bands instead of being at a maximum were at a minimum. On the 
other hand, the dry-air band was well defined—in fact, to use the words of 
Lockyer, it was at a tremendous maximum. Notwithstanding some evi- 
dence to the contrary furnished by Mr. Michie Smith, of Madras, and Dr. 
Assman, of Maddebürg, this seems at first sight pretty conclusive. But it 
is not so, for there arises this question, Would the spectroscopic effect be 
altered according to the condition of aqueous vapour in the upper air? If the 
vapour were turned by the intense cold of the upper currents into icy 
particles, Would the rain band then appear? I do not think it would, and 
we shall see subsequently direct testimony that it would not. So that, after 
all, the aqueous vapour theory does not completely break down—it only 
becomes modified. We have ice crystals instead of watery vapour. I read 
that green and blue suns are frequently seen in polar regions; and icy par- 
ticles in the air refracting the sun's light into its prismatic components are 
very favourable for the display of sunset colours (“ Hard. Sei. Goss.”) The 
question is, what other independent proof have we that the upper air has 
been full of ice crystals? The general view of the world's weather 
recently does not seem to provide any evidence in this direction. Some 
24 j 
