1380 
This indicates that our chances of detecting 
the lines in the spectrum of the corona will 
be greatest in the photographie part of the 
spectrum. Moreover, it appears to explain 
the absence of radiant heat in the light 
emitted from the corona, the particles being 
too small to scatter these longer waves to 
any appreciable extent. Abbott, of the 
Smithsonian party at Wadesboro, found 
the corona cold in comparison with his bo- 
lometer, and infers from this that the 
corona neither reflects sunlight nor emits 
light in virtue of incandescence, expressing 
the opinion that the luminosity is analogous 
to that of vacuum tubes transmitting elec- 
tric discharges. It seems to me that the 
polarization of the coronal light makes this 
theory untenable, and that the absence of 
heat rays can be explained fully by the 
small size of the particles. I am aware 
that the absence of radiant heat in the 
emitted light has yet to be accounted for. 
My own notion, based on experiments 
which are now in progress, is that the re- 
flected or scattered light is vastly in excess 
of the emitted, and that the absence of the 
Fraunhofer lines is more probably due to. 
line-of-sight motion of the particles, than to 
simple drowning out by emitted light. 
My experiments on the ratio of emitted 
to reflected light of a body brought to in- 
eandescence by powerful solar radiation are 
not yet completed, consequently I do not 
yet feel prepared to make any very positive 
statement in regard to this matter. A full 
account of this work will appear shortly in 
the Astrophysical Journal. 
Any observers planning to use a Nicol 
prism in connection with a spectroscope in 
the manner described will find a gas or 
eandle flame illuminated with a beam of 
sunlight concentrated by means of a large 
mirror or lens, extremely useful in making 
preliminary experiments. 
For work on the polarization of the co- 
rona I believe that the artificial corona 
SCIENCE. 
[N.S. Vou. XIII. No. 318. 
described elsewhere will be found most use- 
ful for preparatory work. Not only is it 
polarized, and polarized in the same way as 
the real corona, but it resembles it in a 
most striking manner, and can be easily 
made of the same brilliancy. It would be 
well to work with particles of different 
size, giving different percentages of polari- 
zation, and the picturesque refinements for 
producing the polar streamers could of 
course be omitted. For work of this sort it 
would be well to use a lamp with a ground 
glass bulb, as the conditions of illumination 
would then more nearly approach those in 
the real corona. 
Data regarding the plane of polarization 
in the streamers would be useful in formu- 
lating a theory of the streamers. These, it 
seems to me, can be conceived as. formed in 
two ways: they may be streams of coronal 
particles moving in curved paths, in which 
ease the plane of polarization should be 
everywhere strictly radial, or what is ex- 
tremely improbable, they may be caused by 
divergent beams of light coming from the 
polar regions of the sun and moving in 
curved paths, owing to the rapid decrease in 
the refractive index of the sun’s atmos- 
phere in an outward direction. If this were 
the case, the plane of polarization would 
turn with the streamer. 
This latter hypothesisis extremely vision- 
ary, and I do not present it seriously, for it 
is almost impossible to conceive of any way 
in which the isolated beams of light could 
be formed, unless perhaps by vortex funnels 
more highly luminous than the surround- 
ing surface of the sun. Such fanciful spec- 
ulations are hardly worth indulging in, 
though they have interested me for the 
moment in connection with the matter of 
the possible curvature of light rays in the 
sun’s atmosphere alluded to in a recent 
paper by Julius in the Astrophysical Journal. 
R. W. Woop. 
UNIVERSITY OF WISCONSIN. 
