DIFFUSE MATTER IN SPACE 
385 
1000 parsecs would contain 1*5.10 24 electrons—the number contained in 
5 gm. of matter. Light traversing this column would be reduced in in 
tensity in the ratio e -1 or roughly 1 magnitude. An absorption of l m per 
1000 parsecs is just large enough to be of serious importance in stellar 
investigations in the galactic system. There is no reason to suppose that 
the diffuse cloud extends much beyond the galactic system—its motion 
determined by Plaskett shows that it is associated with our local system 
in particular—so that we must not assume that the absorption extends 
equally to the globular clusters; but anything seen outside our system 
would be dimmed at least one magnitude, and this would give important 
corrections to the deduced distances of globular clusters and spiral 
nebulae. 
But a density of 500 free electrons per cu. cm. is much greater than 
we can admit. We have seen that 1 atom per cubic centimetre is about 
the maximum possible, and since it would not be more than doubly or 
triply ionised, there cannot be more than 2 or 3 electrons per cu. cm. We 
can probably conclude safely that electron-scattering in interstellar space 
is too small to cause appreciable dimming of even the most distant objects. 
There is perhaps a small risk in neglecting it in the determination of the 
distance of the Andromeda nebula (300,000 parsecs) ; but even if the matter 
were supposed to continue with undiminished density through inter- 
galactic space, the required correction would not alter the order of mag 
nitude of the distance of the nebula. 
If dimming by electron-scattering is of cosmical importance at all, it 
must be in local regions of the sky, where the interstellar matter is more 
condensed. 
When a gas is ionised the electron scattering (which does not redden) 
is very much larger than the Rayleigh scattering (which reddens), so that 
the combined result is absorption without appreciable reddening. We 
have above us a column of air containing about 1000 gm. per sq. cm. If 
this were fully ionised, the electron scattering would reduce the light of 
a star overhead by 1000/5 = 200 magnitudes; if singly ionised, the re 
duction would be 28 magnitudes. The Rayleigh scattering which is not 
appreciably different for ionised air and the actual atmosphere is, of 
course, trifling in comparison. 
We might tentatively use observations of reddening as a test whether 
nebulous material is ionised or not. If there is absorption with reddening 
the number of free electrons must be small; if there is no detectable 
reddening the material must be considerably ionised. The only difficulty 
is that we cannot be sure that the absorption is not produced in some 
quite different way, e.g. by meteoric matter. Measurements of reddening 
by obscuring patches and gaseous nebulae have been attempted, but it is 
too early yet to state any firm conclusions.