128 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1919. 
NEW STARS. 
Within historical times some 27 new stars have suddenly flashed 
out in the heavens. Some have been of interest only to the astrono- 
mer ; others, like that of last June, have rivaled Sirius in brilliancy. 
All have shown the same general history, suddenly increasing in 
light ten thousandfold or more, and then gradually, but still rela- 
tively rapidly, sinking into obscurity again. They are a very inter- 
esting class, nor has astronomy as yet been able to give any universally 
accepted explanation of these anomalous objects. Two of these novae 
had appeared in spiral nebulae, but this fact had not been weighed 
at its true value. Within the past two years over a dozen novae 
have been found in spiral nebulae, all of them very faint, ranging 
from about the fourteenth to the nineteenth magnitudes at maxi- 
mum. Their life history, so far as we can tell from such faint 
objects, appears to be identical with that of the brighter novae. 
Now, the brighter novae of the past — that is, those which have not 
appeared in spirals — have almost invariably been a galactic phe- 
nomenon, located in or close to our Milky Way, and they have very 
evidently been a part of our own stellar system. The cogency of 
the argument will, I think, be apparent to all, although the strong 
analogy is by no means a rigid proof. If 27 novae have appeared in 
our own galaxy within the past 300 years, and if about half that 
number are found within a few years in spiral nebulae far removed 
from the galactic plane, the presumption that these spirals are them- 
selves galaxies composed of hundreds of millions of stars is a very 
probable one. 
If, moreover, we make the reasonable assumption that the new 
stars in the spirals and the new stars in our own galaxy average 
about the same in size, mass, and absolute brightness, we can form a 
very good estimate of the probable distance of the spiral nebulae, 
regarded as island universes. Our galactic novae have averaged 
about the fifth magnitude. The new stars which have appeared in 
the spiral nebulae have averaged about the fifteenth magnitude, but it 
would appear probable that we must inevitably miss the fainter 
novae in such distant galaxies, and it is perhaps reasonable to assume 
that the average magnitude of the novae in spirals may be about 
the eighteenth, or 13 magnitudes fainter than those in our own galaxy. 
They would thus be about one hundred and sixty thousand times 
fainter than our galactic novae, and on the assumption that both 
types of novae average the same in mass, absolute luminosity, etc., the 
novae in spirals should be four hundred times farther away. We do 
not know the average distance of the new stars which have appeared 
in our own galaxy, but 10,000 light-years is perhaps a reasonable esti- 
