Harvard College Observatory: Stellar classification systems, 1
Williamina Fleming, at right, uses a special frame backlit by a mirror, which allowed her to view the images on the observatory’s glass plates. At left is Mable Stevens, working as a recorder.
Harvard University Archives, HUV 1210 (9-6).
In 1882, Edward Pickering laid out a plan in a short pamphlet suggesting that interested parties might search for five types of stars with varying magnitudes:1
I. temporary stars, or those that shine out suddenly, sometimes with great brilliancy, and gradually fade away, otherwise known as novae; II. long-period variables, or those undergoing great variations of light; III. stars undergoing slight changes according to laws as yet unknown; IV. short period variables, or stars whose light is continually varying, but the changes are repeated with great regularity in a period not exceeding a few days; V. Algol stars, or stars which … every few days suffer a remarkable diminution in light for a few hours.
Toward the conclusion, he provided motivation for undertaking the study of variable stars in case any reader should need convincing:2
Apart from the value of the results attained it is believed that many amateurs will find it a benefit to accustom themselves to work in a systematic manner, and that they will thus receive a training in their work not otherwise easily obtained outside of a large observatory. The lesson should be taught that time spent at a telescope is nearly wasted unless results are secured worthy of publication and having a permanent value.
Heeding Pickering’s clarion call, many amateurs followed through with the plan of observing variable stars and reported the resulting data to the observatory.
Meanwhile, under the banner of the new Henry Draper Memorial, Williamina Fleming began the task of cataloging stars by their spectra, as recorded on photographic glass plates made in Cambridge, Massachusetts, and in Arequipa, Peru. In her investigations, she discovered that a preponderance of stars categorized as variables by brightness were Type III stars by their spectra, using the classification system of the Italian astronomer Angelo Secchi.
Initially, Secchi’s system had contained two classes, one of which consisted of red and yellow stars like the Sun, while the other contained blue stars such as Sirius. As he was exposed to more and more spectra, Secchi changed his classifications to allow for five different types: Class I: blue stars such as Sirius, which display a small number of hydrogen absorption lines; Class II: stars similar to the Sun, with spectra rich in narrow dark lines; Class III: reddish stars such as Betelgeuse, which display broad dark bands; and Class IV: stars with strong carbon lines.3 In 1877, Secchi added a fifth class of emission-line stars.
A natural part of Fleming’s job was to devise a new classification system.
References
- Edward C. Pickering, A plan for securing observations of the variable stars (John Wilson and Son, 1882), p. 5.
- Ibid., pp. 14–15.
- Jay Holberg, Sirius: The brightest diamond in the night sky (Springer, 2007), p. 87, doi:10.1007/978-0-387-48942-1.
Cite this resource
Bretislav Friedrich and Maria McEachern, “Stellar classification systems, 1,” Harvard College Observatory history guide, American Institute of Physics, 2026, https://www.aip.org/history/harvard-observatory/stellar-classification-1.
Note that this material was originally developed in concert with the Williamina Fleming history guide as part of the Women in the History of Quantum Physics collection.