Astronomers have been classifying stars since at least 120 B.C.E., when Greek astronomer Hipparchus classified stars according to how bright they were. In 1666, Isaac Newton found that the light of the Sun could be dispersed into a series of colors, called a spectrum. These measurements give the brightness of light at specific wavelengths. Ever since, astronomers have been analyzing the spectra of stars.

In the 1800s, a major change occurred, one that proved to reveal many unknown aspects of how stars work: astronomers began to find holes in these spectra, fine dark lines scattered in patterns throughout the spectra. Different stars, astronomers soon found, had different patterns of these dark lines in the spectra. Discovering why is one of the great stories of modern astrophysics.

Astronomers eventually realized that each dark line corresponds to specific elements in the stars. This made it possible for astronomers to categorize these spectra in order to figure out what stars are made of. In the 19th century, astronomers tried out many different forms of stellar classification, but they came to little consensus. By 1900, there were more than 20 classification systems in use based on spectra. However, one classification scheme came to dominate which was developed by women who worked at the Harvard Observatory.

Starting in 1885, Edward Pickering, the Director of Harvard College Observatory, hired a corps of women to use photographs of stellar spectra to catalogue and classify stars. First came Williamina Fleming, Pickering’s former maid, who categorized stars by how much hydrogen they had. Then, Antonia Maury created a complicated system of classifying stars based on their temperature. Finally, Annie Jump Cannon, one of the signatories to the guestbook, merged the two theories to create a simpler system of classification than Maury’s — the classification that astronomers use today.

In 1913, Henry Norris Russell, who visited the Observatory on November 11, 1921, plotted a chart comparing the magnitude of stars to their spectral type. Using research from Ejnar Herzsprung, who determined that temperature and brightness are related in stars, Russell showed that Annie Jump Cannon’s spectral classes correspond to temperature and brightness.

Stellar classification methods in the 19^th and early 20^th centuries relied on looking at and differentiating stellar spectra. Williamina Fleming’s classification system, in particular, purported to categorize spectra by how much hydrogen they had. In 1925, however, for her PhD dissertation, Cecilia Payne found a different result. Payne looked at a recent theory published by astrophysicist Meghnad Saha, who found that atoms at different temperatures in stars occupy different states of energy. Then, using experimental data, she took that theory and discovered, contrary to the perception at the time, that all stars are predominantly comprised of hydrogen, some helium, and trace amounts of the rest of the elements on the periodic table. By inference, this led to the realization that in contrast to living organisms and rocky planets, the Universe is essentially just hydrogen.

Subrahmanyan Chandrasekhar was five years out of having made a major breakthrough in our understanding of stars when he visited the Observatory in February 1936. Chandrasekhar was at the time visiting the Harvard College Observatory — he visited for three months — before heading off to do research at the Yerkes Observatory. During this period, he was looking into the reason why white dwarfs, an outlier region of the Hertzsprung-Russell diagram, form. Building off a theory of his graduate supervisor, Ralph Fowler, who used the emerging field of quantum mechanics to determine why the extremely dense stars do not collapse on themselves. At a certain mass of star, Chandrasekhar found, the quantum mechanics pressure ceases to prevent the stars from collapsing. That limit is now known as the Chandrasekhar limit.