"Yardsticks" in Neighbor Galaxy Double Universe's Size
Before we discuss the problem with the Cepheid variable distance scale that was discovered in the 1940s, researchers had already seen evidence of a problem with the distance scale. Astronomers are always looking for different types of "standard candles" – objects of known brightness for which a distance can then be calculated using the observed brightness. One such object that astronomers tried to use in the early days of extra-galactic observations was the globular clusters of a galaxy. Globular clusters are a type of star cluster that tend to lie in the halo of a galaxy. In 1931, Hubble compared the brightest globular cluster in our galaxy to the brightest one in Andromeda, and found that the one in Andromeda was intrinsically much dimmer than the one in ours.
Later, he compared the globular clusters in our galaxy to those in Messier 33 (M 33), another spiral galaxy that is further than Andromeda. He found that the globular clusters in M 33 were still fainter than those in Andromeda, based on the distances he had earlier found. Hubble recognized that it was possible for globular clusters to vary from galaxy to galaxy, but it seemed unlikely that they would preferentially get dimmer the further they were from our galaxy. This pointed to a problem with the distance scale, but Hubble did not have an explanation for the origin of the problem.
When Baade discovered that there were, in fact, two different populations of stars, it did not take researchers long to connect the dots between the discrepancy that Hubble was seeing in the globular clusters of distant galaxies and the Cepheid distance scale. Baade's images of Andromeda and Andromeda's two companion galaxies, Messier 32 and NGC 205, showed that there were two populations of stars, which he called "Type I" and "Type II" – Type I stars were bluer and brighter whereas the Type II were redder and fainter. He recognized that the globular clusters were rife with Type II stars, whereas the disk of the galaxies tended to have both Type I and II stars.
As it happens, the Large Magellanic Cloud, the location of the Cepheid variables that Henrietta Leavitt observed, is populated with Type I stars. However, when Shapley calibrated the Cepheid period-luminosity relationship, he was using observations of Cepheid variables in globular clusters – where Type II stars reign. The Cepheid variables that Hubble observed in the Andromeda nebula were Type I, but he unknowingly used Shapley's calibration that were calibrated for Type IIs. The problem lies in the fact that there are two different calibrations of the period-luminosity for the two different populations of Cepheid variables.
The re-calibration of the Cepheid scale for the Type I Cepheids showed that the distances to the Type I Cepheids was off by about a factor of 2. In other words, astronomers were finding distances that were half as far as they should have been. Andromeda suddenly went from being a "mere" 800,000 light-years away to being about 1,800,000 light-years away. With this new distance, the brightest globular clusters in Andromeda were now about the same intrinsic brightness as those found in our own galaxy.
To bring this discussion into a more modern note, astronomers have studied the differences between Type I and Type II stars in more detail. Typically, these are now referred to as Population I and Population II instead of Type I and Type II. Astronomers have found that one difference between Population I and II stars it the amount of heavy elements in the stars. Population II stars show very little heavy elements where Population I have more. It is thought that Population II stars are, therefore, older stars, formed earlier in our Universe's lifetime. This is because the Universe started as mainly hydrogen and helium. This is still true, but the abundances of the elements heaver that hydrogen and helium have slowly increased as stars fuse hydrogen into heavier and heavier elements. Population I stars were likely formed out of the remains of earlier, Population II stars.
Other resources
The following web page has additional information:
