1.1.8. The Discovery of Universal Expansion

The last major step in observations which have led to the development of our modern cosmological model was provided by Hubble. Earlier observations by Slipher (1914) had shown that galaxies had spectral features (emission or absorption lines) and that most of them exhibited a Doppler shift towards longer wavelengths, commonly called a redshift. Hubble and Humason systematically measured more galaxy spectra and again found that most galaxies exhibited a redshift, indicating radial motion away from the observer. Hubble noticed that not all galaxies exhibited the same redshift, so the amplitude of the redshift must depend upon some other property. But before that property could be elucidated, it had to be established if these nebulae (galaxies) were objects located in our own Galaxy or well beyond its boundaries. Resolution of this issue would require measuring the distance to the Andromeda Nebulae (M31).

The first credible effort was done in 1922 by E. Opik who derived a distance to M31 of about 450 kpc or approximately 1.5 million light years - substantially larger than the estimated size of the Milky Way Galaxy. Opik reasoned that, since galaxies exhibited approximately equal flux ratios in different filters (that is the colors of galaxies are not greatly dissimilar) then, to first order, all galaxies have similar stellar populations and hence similar values of mass-to-luminosity ratio (M / L). If M / L does not vary greatly from one galaxy to another, then a galaxies rotational velocity, which from Newtonian mechanics must be determined by M, is also an indicator of its intrinsic luminosity since M / L is assumed to be constant. Under this assumption, the distance to M31 follows trivially from measuring its apparent flux.

Further support for the extragalactic nature of M31 came in 1925 when Hubble discovered variable stars in M31 that had similar properties to those detected earlier in the Large Magellanic Cloud. This provided a relative distance scale between the LMC and M31 and indicated that the M31 was located at a distance of 300 kpc. Hubble discovered Cepheid variables in most of the galaxies in the Local Group but, in this environment, redshift is not well-correlated with distance as the Local Group is loosely gravitationally bound. Observations of Local Group galaxies therefore would not reveal universal expansion.

Hubble took spectra of fainter and smaller galaxies as well and noticed that their observed redshifts were considerably larger than anything in the Local Group. Hubble also noticed that galaxies which were faint and had small apparent angular sizes, tended to have larger redshifts than galaxies which appeared bigger and brighter. Lacking a suitable means for determining distances to these smaller and fainter galaxies, Hubble made some assumptions on the nature of these galaxies. By assuming that galaxies were either of constant brightness or constant physical diameter Hubble could deduce that the smaller and fainter galaxies with the higher redshifts were farther away than the brighter bigger galaxies with smaller redshifts. In this way, Hubble could make a plot of galaxy distance versus observed galaxy redshift.

This plot, using Hubble's original data, is shown in Figure 1-4. Although the data are quite noisy, there is a general trend for more distance objects to exhibit a larger redshift. These data were sufficient to empirically demonstrate that recessional velocity was proportional to distance and hence that the Universe was in a state of uniform expansion. Einstein's General Relativity and the static Universe could now be resolved - the Universe itself was expanding. This determination of uniform, linear expansion of the Universe has a clear prediction. If galaxies are moving apart from one another today, then in the past they must have been closer together. Indeed, there must have been a time when all the galaxies in the Universe were together in the same space. At this time, the Universe was very dense and in a physical state well-removed from how it is observed to be today.

Figure 1-4: Hubble's original data plot which shows the correlation between redshift and distance for his sample of galaxies. Although the data is quite noisy, the overall trend is adequately represented by a linear law which redshift is directly proportional to distance. That law is depicted by the solid line.