The Great Nebulae 
19 
15-18] 
“ island-universes ” distinct from the universe which contains our sun. 
Knowing their distances, it is easy to calculate their sizes; the diameter of 
the great Andromeda nebula, which subtends nearly three degrees in the sky, 
must be about 15,000 parsecs; the diameter of M 33, subtending about one 
degree, must be about 5000 parsecs. We may call these nebulae “extra- 
galactic ” on account of their distance, or may speak of them as the “ great 
nebulae ” on account of their size. 
17. As seen in a telescope, the great nebulae differ enormously in apparent 
size, shape and brightness. But Hubble has recently shown* that differences 
in size and brightness in nebulae of the same shape are almost entirely due 
to a distance effect. Nebulae of the same shape may thus be thought of as 
similar manufactured articles, or as astronomical plants belonging to the 
same species. Hubble further finds that even nebulae of different shape shew 
only slight differences in intrinsic luminosity, and no great differences in size. 
These circumstances make it possible to estimate the distances of all 
nebulae, down to the very faintest visible, with fair accuracy. The faintest 
nebulae which can be seen photographically in the Mount Wilson 100 -inch 
telescope give only about a hundred-thousandth part of the light of the 
brightest. Assuming the difference in light to be due to a distance effect, 
it is found that the 18th magnitude nebulae must be at a distance of about 
140 million light-years. This represents the range of vision of the 100-inch 
telescope for objects having the luminosity of the great nebulae; it is the 
greatest distance with which practical astronomy has so far had to deal. 
Within this distance Hubble estimates that about two million nebulae must 
lie, these being fairly uniformly spaced at distances of about 570,000 parsecs 
apart. 
We can construct an imaginary model of the system of the great nebulae 
by taking about 50 tons of biscuits and spreading them so as to fill a sphere 
of a mile radius, thus spacing them at about 25 yards apart. The sphere 
represents the range of vision of the 100 -inch telescope; each biscuit repre 
sents a great nebula of some 4000 parsecs diameter. A few nebulae of 
exceptional size must be represented by articles rather larger than biscuits, 
while our system of stars, up to Kapteyn s tenth spheroid, would be repre 
sented by a flat cake 13 inches in diameter and 2 £ inches in thickness. On 
this scale the earth is far below the limits either of vision or even of 
imagination. It is little more than an electron in one of the atoms in our 
model; and we should have to multiply its dimensions many millions of 
times to bring it up to the size of even the smallest particles which are 
visible in the most powerful of microscopes. 
18. This completes our brief astronomical survey of the dimensions of the 
universe. We have dwelt mainly on the lay-out, as regards distance, of the 
* Astrophys. Journal, lxiv. (1926), p. 321. See Chapter xm below.