Work With Variable Stars and Nebulae

The Horsehead Nebula

Glass Plate photonegative b10250

Williamina Fleming was prolific in her science. You might recognize the Horsehead Nebula, one of her early discoveries from 1888 (Fig. 4). While the newer image is colorized and flipped for visual understanding, Fleming first found the object as a small, faint grey smudge on a glass plate negative (more accurately, she found a curious blank spot, like an eraser smudge -- the Horsehead Nebula is a 'dark nebula', meaning it is visible to us only because it is illuminated from behind, and we see its silhouette) (Fig. 2 & 3). This was not the only faint object her detailed eye found -- Fleming also discovered 10 novae, 59 nebulae, over 300 variable stars, and was part of the first team to recognize the existence of white dwarf stars.

Discovery of Horsehead Nebula

Using Williamina Fleming’s Notebooks 

Williamina Fleming and the Harvard Computers all kept detailed notes on their work at the Observatory. The following notes made by Fleming correspond with Plate i10491, from which she was identifying variable stars (Fig 5). The colored boxes diagram her work process. 

Fleming first drew a diagram (red box) of the star field in question, circling the new variable star and labeling the major stars nearby. In a table beside this diagram, she documented the positions and known magnitudes of the labeled stars (green box). Fleming then determined the magnitude of the new star by using the “Pickering-Fleming Method” of comparison (cyan box). This method goes as follows: she compared the target variable star “v” with known magnitudes, slightly dimmer and slightly brighter star, noting how many tenths of a magnitude difference there is between the measurements. In the cyan box her notes read that d is 0.1 dimmer, and “e” is 0.5 brighter than “v” (Fig. 6).

Comparing the Notebooks with Glass Plates 

To determine a new variable star’s cycle, Fleming spent upwards of two days inspecting past HCO photos of this area of sky. Using this method of time-domain astronomy, she measured the star at each moment in time to see how its brightness varied. In this entry she compares the variable, v, on plate b11191, with three other stars (shown in Fig. 8). According to her notes (cyan box), c is 0.5 dimmer, d is the same brightness, and e is 0.7 brighter (Fig. 7).

Since most stars on the observatory's glass plates appeared too small for the unaided eye to examine, the computers used a magnifying glass. With larger images, the computers could better position the stars in their sketch (Fig. 9, compare with Fleming's notes from plate i10491 in Fig. 5 above). On top of the glass plate is a tool astronomers call a “fly spanker” (Fig. 10). The flyspanker allows the computers to compare the brightness of a star in a photo to a calibrated scale. For brighter stars and longer exposures, more light reaches the plate, imprinting a larger dot. 

Newly discovered variable star in second plate

Plate referenced in Figs 5 & 7

Star chart from Fig 5 on referenced plate with fly spanker

Preparing the Data 

Publication of discovery of new variable star

The observatory could not publish Fleming's findings as they appear in the notebooks -- it is indecipherable! To reduce this data into publishable form, Fleming would clean and organizer her calculations, but first she would check to make sure her data points were strong enough to hold against critics. While many HCO plates contained her target star, she excluded certain plates whose brightness was too dim, or whose focus was blurred. Demonstrating the skill and pride with which the old observatory staff performed their work, she only excluded three out of fifty-four plates that images the Virgo region (Figure 11a). Of course, for transparency and reproducibility, she also noted the specific dates for these plates (Fig 11b).

Read on to the next section, to see how Williamina Fleming transformed this raw data into public knowledge!

Image Sources

2. (1893). Plate b10250.  Center for Astrophysics | Harvard & Smithsonian, Photographic Glass Plate Collection, Cambridge, MA.

3. (1888). Plate b2312.  Center for Astrophysics | Harvard & Smithsonian, Photographic Glass Plate Collection, Cambridge, MA.

4. Crawford, K. (2011). Horsehead Nebula [Image]. Retrieved from https://en.wikipedia.org/wiki/Horsehead_Nebula#/media/File:Barnard_33.jpg

5. Fleming, W. P. (1896). Variable stars. Project PHaEDRA (Sequence Volume 8), Center for Astrophysics | Harvard & Smithsonian, Cambridge, MA.

6. Fleming, W. P. (1896). Variable stars. Project PHaEDRA (Sequence Volume 8), Center for Astrophysics | Harvard & Smithsonian, Cambridge, MA.

7. Fleming, W. P. (1896). Variable stars. Project PHaEDRA (Sequence Volume 8), Center for Astrophysics | Harvard & Smithsonian, Cambridge, MA.

8. (1894). Plate b11191. Center for Astrophysics | Harvard & Smithsonian, Photographic Glass Plate Collection, Cambridge, MA.

9. (1894). Plate b11191. Center for Astrophysics | Harvard & Smithsonian, Photographic Glass Plate Collection, Cambridge, MA.

10. Wargelin, M. (2018). Photograph of HCO human computer “fly spanker” [Image]. Cambridge, MA: John J. Wolbach Library.

11a & 11b. Pickering, E. C. & Fleming, W. P. (1896). Harvard College Observatory Circular No. 6 New variable stars. Astrophysical Journal 3, 296-7.