ag 
ves 
206 EH. A. Hazen—The Sun Glows. « 
into an ocean of gigantic coral formations. The frost work was 
from ‘560 to ‘75 inch thick upon the branches and trunks of 
the trees. The frost work after a time formed on the coat of 
the observer as if accumulated by degrees. These frost parti- 
cles were too minute to be seen directly, but could easily be 
recognized either by looking through a great depth or by the 
effect of the sun’s light upon them. ; 
It is not definitely known what causes the ordinary sunrise 
and sunset colorings, but it is certain that they may occur in all 
their magnificence in clear skies. he opinion generally held, 
that they are caused by the absorption of the rays at the blue 
end of the spectrum by moisture or dust particles, will not ex- 
‘plain the present glows. It has been thought, however, that 
they may be explained if we consider them due to reflection 
from definite particles in the air. One difficulty in this is that 
we must assume that reflection and absorption would produce 
an effect precisely similar to the ordinary phenomenon. It may 
be said that the ordinary coloring may be caused by reflection, — 
but it is a little singular that such an explanation has not be- 
fore been suggested by meteorological authorities. 
nder the third condition suggested above, some form of 
electrical action may be mentioned as amply sufficient to fulfill 
the requirements. We have no means of deciding the height — 
at which frost particles are ordinarily sustained. Siemens has 
suggested that some form of vapor must be found at and be-— 
yond the limits of our atmosphere; however this may be, there 
must be some cause for so large an increase in the amount of 
frost particles in the upper regions of the atmosphere as to pt — 
duce the present remarkable after-glows. That electrical actioh — 
has been increased of late is evidenced by the behavior of elee 
trical instruments in India immediatelv after the tremendous % 
convulsion at Krakatoa. Also we note the recent increase 2 
tion. The following table of mean monthly sunspot numbers 
tions have been made by Prof D. P. Todd, of Ambers 
Mass. I have added the observations since May, 1883, 10F 
comparison. : 
