Chasing Giants

For the past week I've been chasing down the history of the Cepheid distance scale as part of my "real job" (as opposed to my grad work). I've never been a big history buff, but when it comes to history of science, math or numbers, I dive right in. The target for my research was to figure out the state of the Cepheid scale as of 1955, the date of the issue of the "Cosmic Times" newspaper that we're working on.

Universe twice as large as originally thought

By 1955, the Universe was known to be far vaster than distance calculations had so far indicated, as the distances measured by Hubble and others are half of what they should have been. An error in the calibration of the Cepheid variable period-luminosity scale caused the calculated distances to be off.

The first inkling of trouble with the distance scale came in the mid-30s with observations of globular clusters in the Andromeda nebula. Edwin Hubble noticed that the peak luminosity of globular clusters in Andromeda were 1.5 magnitudes fainter than those in our own galaxy. Under the "equivalence principle", it would be expected that globular clusters in all galaxies would be somewhat similar, so it was a surprise to see Andromeda's globular clusters to be dimmer than expected. On the other hand, the equivalence principle is best taken at large scales, with large samples, so it didn't worry Hubble too much. However, measurements of M 33, another nearby galaxy, showed that its globular clusters are even dimmer than those of Andromeda. At this point, Hubble and other researchers realized that there was a problem with the distance scale, but did not know where to turn for the answer.

In the 40s, however, Walter Baade, using special red-sensitive glass plates developed by Kodak and taking advantage of the black-outs in California due to the war, imaged Andromeda using the 100-inch Hooker telescope. Baade achieved far better detail than had previously been possible. The images showed that there were two classes of stars. One class was concentrated in the disk of the galaxy while the other was distributed in the halo of the galaxy.

In fact, Jan Oort had originally identified two classes of stars in the teens, which were called high-velocity and low-velocity stars. Baade's image of Andromeda drove home that these were two classes, and that they inhabited different parts of the galaxy. In fact, this made it clear that the Cepheids in the globular clusters were also a different class of objects from those in the disk. The problem with the distance scale, then, came because Henrietta Leavitt had developed the period-luminosity relationship using disk Cepheids in the Large Magallenic Cloud, but Harlow Shapley had calibrated that scale using globular cluster Cepheids.

The re-calibration increased the measured distances to galaxies by a factor of 2, so almost overnight, the size of the known universe doubled.