|Discovery of Galaxies
Key points: How it was established that other galaxies are island universes of stars; standard candles and distances; the distance ladder
While the Milky Way was considered to be a thin, distorted disk of stars with the sun near the center in the early 20th century, other galaxies were confused with gaseous nebulae, and were assumed to be part of Milky Way.
The best pictures showed other galaxies to have spiral arms, but they also appeared smooth and were thought to be a special kind of gaseous nebulae, the "spiral nebulae".
By the 1920s, a debate was ranging among astronomers about whether the spiral nebulae were gaseous objects (HII regions) or separate "island universes" like the Milky Way. Vesto Slipher had obtained spectra that showed them to have continuous spectra, not emission lines, consistent with their being made of stars.
Harlow Shapley thought that the Milky Way was so big that it had to dominate the Universe, and thus that the spiral nebulae were gaseous and part of the Milky Way. Part of his evidence was they avoided the plane of the Milky Way and they would have been found over the whole sky if they were independent systems. Both of these arguments were wrong, because Shapley did not appreciate the effects of interstellar dust and extinction. Extinction toward the globular clusters had made their stars too dim and consequently Shapley had placed them too far away, and extinction in the plane of the Milky Way hid the spiral nebulae behind it.
On the other hand, many astronomers did not yet accept Shapley's arguments that the solar system was to the side of the Milky Way -- again because extinction hid the true shape of the system from most types of observation.
Shapley and Heber Curtis "debated" the two views on April 26, 1920. Shapley was under consideration for director of Harvard College Observatory, and did not want to do anything that would be unnecessarily provocative and undermine his position. He gave a lecture that was very watered down, with many popular ideas that were uncontroversial. Curtis gave a more advanced talk, and achieved extra points for technical sophistication for bringing his outline on projection slides that provided the audience a written version while he talked! Both debaters were right (Shapley -- shape of the Milky Way; Curtis -- size of the Milky Way, spirals are island universes) and wrong (Shapley -- spirals are gaseous; Curtis -- sun is near center of Milky Way (Kapteyn model)). Shapley's main item of evidence -- measurements by van Maanen that indicated that the spiral arms moved perceptibly with time -- was soon shown to be wrong. More on the debate at http://antwrp.gsfc.nasa.gov/diamond_jubilee/debate_1920.html
Why was this debate so important
The final "proof" came with the use of the new 100-inch telescope on Mount Wilson by Hubble (in 1923-26). Although the excellent images of this telescope were also important, the critical advance was provided by the then widely used photographic plates that allowed him to accumulate deep exposures on large fields to resolve stars in Local Group galaxies, demonstrating that they are island universes.
Picture of Hubble at the 100-inch telescope. From S. Mais, http://www.soteoria.hpg.ig.com.br/Hubble/page1.htm
|Here is one of Hubble's photographic plates of the center of M33, with Cepheid variables marked. If you look closely (and if the original resolution has been preserved from plate to journal to scanner to web page to your computer), you will see that the gray "nebulosity" is broken up into little specks that are individual stars. (The plate is a negative, as was the convention for publishing results -- the stars look black and the sky white) (From ApJ, 63, 236, 1926)|
|Hubble marked the Cepheid variable stars on the plate above. Here are
his measurements of some of the Cepheid variable light curves (some of these stars are in
other regions than the one in the picture above).
Cepheids have a Period-Luminosity relation that makes them good standard candles and allowed Hubble to prove that M33 is so far away that it cannot be part of the Milky Way.
This proved the "island universe" model, that the "spiral nebulae" are galaxies like the Milky Way.
Why do we need to determine the distances to the galaxies
|The group of galaxies called "The
Local Group" to which the Milky Way belongs has played a crucial role
in our determination of distances to galaxies.
Here is a sketch of the arrangement of the local group -- in three dimensional projection. Each of the two large galaxies (M31 = Andromeda and the Milky Way) has a retinue of small galaxies.
And here is a close up of the space around the Andromeda Galaxy, showing its set of satellite galaxies in more detail. Galaxies on solid stalks are above the grid plane and those on dashed stalks are below.
Although Hubble proved the "Island Universe" hypothesis by measuring distances in the Local Group with Cepheid variable stars, Cepheids cannot be detected in very distant galaxies. Therefore, astronomers have reached out by calibrating distance indicators useful over ever greater distances. All of these tests are built on "standard candles", objects where we think we know the luminosity and so can assume that the apparent brightness is determined by the distance through the inverse r2 law. The local group galaxies were to a large degree where the distance indicators were explored, tested, and established. With improvement in instruments, we were then able to use them on galaxies much farther away.
The series of steps we use to measure distances is described as a ladder, with each rung dependent on the ones below supporting it. These rungs we list below in the order of the distance range they work over:
Radar: Radar is the most accurate way to measure distances in the solar system, giving an accurate foundation for parallax.
Parallax: Modern space observatories let us measure parallax to 0.001 arcsec, directly determining distances to hundreds of parsecs.
Main sequence fitting: Since we know that similar stars should appear at the same place on the H-R diagram, we adjust the distances of cluster H-R diagrams until their main sequences overlap. From the placement of stars on the H-R diagrams, we can estimate their absolute magnitudes and thus determine both the true and relative distances to the clusters.
Cepheids: directly calibrated in Milky Way, for example from clusters with distances from main sequence fitting. Can be observed to ~20 Mpc
Brightest stars: relatively constant and useful over about 4x the distance that Cepheids are
Novae: reach about the same peak brightness as the brightest stars and serve as a check
Globular clusters: the brightest ones have about the same brightness and are visible over slightly greater distances than the brightest stars. Also useful for ellipticals.
21-cm line width: the width of the HI 21-cm line reflects the rotation rate of the galaxy which in turn is a measure of the mass. Most galaxies have nearly the same ratio of mass to luminosity, so if we know the mass we can compute the luminosity (technically, this correspondence is called the Tully-Fisher relation).
Supernovae: can be seen over huge distances (to ~ 50% of the age of the Universe). White dwarf based supernovae are expected to all have similar peak luminosities.
|As in this illustration, the ladder is a bit rickety, and the errors
build up in the sequence of steps, making some of our distance estimates rather uncertain.
(From Wojtek Kozak, http://www.wkozak.com/ColourEditorialDrawings7.htm.)
Test your understanding before going on
|An illustration from Chinese mythology of the Milky Way, from Postel & Guerrero http://www.artistexpo.com/Collprints.html||
Pillars in M16, NASA, HST, J. Hester, P. Scowen, APOD http://antwrp.gsfc.nasa.gov/apod/ap951106.html
Click to return to syllabus
|Click to return to the Milky Way||
hypertext G. H. Rieke
Click to go to the Interstellar Medium