SPIRAL NEBULAE—CURTIS. 129 
mate.* This would indicate a distance of the order of 4,000,000 light- 
years for the spiral nebulae. This is an enormous distance, but if 
these objects are galaxies like our own stellar system, such a distance 
accords well with their apparent dimensions. Our own galaxy at a 
distance of 10,000,000 light-years would be about 10 minutes of arc 
in diameter, or the size of the larger spiral nebulae. 
On such a theory a spiral structure for our own galaxy would be 
probable. Its proportions accord well with the degree of flattening 
observed in the majority of the spirals. We have very little actual 
evidence as to a spiral structure for our galaxy; the position of our 
sun relatively close to the center of figure of the galaxy and our 
ignorance of the distances of the remoter stars renders such evidence 
very difficult to obtain. A careful study of the configurations and 
star densities in the Milky Way has led Professor Easton, of Amster- 
dam, to postulate a spiral structure for our galaxy. 
4 Note added in June, 1920.—The estimate given above of 10,000 light-years as the ayver- 
age distance of the galactic novae is probably too large. Data obtained since the 
lecture was delivered now make possible another method of estimating the distance of 
the spirals, leading, however, to the same general result. 
Seventeen noyae haye appeared in the great nebula of Andromeda, the largest and pre- 
sumably the ciosest of the spirals. Sixteen of these were faint, averaging about magnitude 
17 at maximum and probably about magnitude 27 at minimum, on the assumption that 
they vary in this respect as do the galactic-novae. The seventeenth, 8S. Andromedae, was 
seventh magnitude at maximum, or 10 magnitudes brighter than the average of the 
fainter novae. 
The absolute magnitude, or absolute luminosity, of a star is that apparent magnitude 
which it would have if seen from the standard distance of 32.6 light-years, and may 
easily be found from the equation, 
Abs. magn.—=apparent magn.+7.6—5 Xlogarithm of distance, 
where the distance is expressed in light-years. Converting the apparent magnitudes of the 
16 fainter noyae into absolute magnitudes on the assumption that this spiral is 500,000 
light-years distant, the following comparison may be made with four galactic novae, whose 
distances, and hence their absolute magnitudes, are known, 
16 novae in F 
Andromeda,| 4,Stlactic 
ifat 500,000 
+ known 
light-years . 
distance. | distance. 
DSOMMUG HAC TILNOG AL IAAI sane s Sele cece em ene ene e ieee —3.9 —3.4 
Absolute magnitude at'minimum’. .-). . 99.2202. 02982% Le SME Ee 6.1 +7.2 
Though it must be admitted that the data for such a correlation are still very limited, 
a distance of 500,000 light-years is strongly indicated for the nebula of Andromeda, and, 
if the spirals are structures of roughly the same order of actual size, a distance of 
10,000,000 or more light-years would be expected for the apparently smaller spirals. 
At this distance for the nebula in Andromeda, S. Andromedae, evidently an exceptional 
nova, would have the very great absolute magnitude of —13.9. The ‘“ dispersion” of the 
novae in absolute magnitude is evidently very large, as indicated by the difference of 10 
magnitudes between S. Andromedae and the fainter novae in this spiral. An absolute 
magnitude of —14 does not seem impossible for certain exceptional novae in our own 
system. Tycho’s nova was brighter than Venus at its maximum, and if this nova was as 
close to us as 1,000 light-years, its absolute magnitude must have been about —13. 
