In a cramped Harvard University sub-basement, a team of women is working to document the rich history of women astronomers.
More than 40 years before women gained the right to vote, female "computers" at Harvard College Observatory were making major astronomical discoveries.
Between 1885 and 1927, the observatory employed about 80 women who studied glass plate photographs of the stars. They found galaxies and nebulas and created methods to measure distance in space.
They were famous - newspapers wrote about them, they published scientific papers under their own names. But they were virtually forgotten during the next century.
But a recent discovery of thousands of pages of their calculations by a modern group of women has spurred new interest in their legacy.
Surrounded by steel cabinets stuffed with hundreds of thousands of plate glass photographs of the sky, curator Lindsay Smith Zrull shows off the best of Harvard's Plate Stacks collection.
Each glass plate is stored in a paper jacket and initialled to show who worked on it.
But for decades no-one kept track of the women computers' full names. So Smith Zrull started a spreadsheet about 18 months ago and adds initials when she discovers new ones and then tries to locate the full names in Harvard's historical records.
"I'm slowly starting to piece together who was who, who was here when, what they were studying," Smith Zrull says.
She has about 130 female names. About 40 are still unidentified.
She points at a glass plate crowded with notes taken in four different colours. "One of these days, I'm going to figure out who M.E.M. is."
Not all of the initials belong to the computers. Her list has grown to include assistants and, in some cases, astronomers' wives who helped with their husbands' work.
Dozens of women worked on the glass plate photographs at Harvard. "Which is a pretty amazing number considering women were still trying to get social approval to go to college, let alone work in the sciences," Smith Zrull says.
She now oversees a digitisation project at Harvard-Smithsonian Center for Astrophysics to make the glass plates available to the world.
Since 2005, a custom-built scanner has been making its way through the collection of more than half a million plates from 1885 to 1993. The team scans 400 plates per day - they're at about the halfway point now - and Smith Zrull estimates about three years remains.
'People forgot they were there'
As the scanning continued last year, Smith Zrull turned her attention to the notebooks belonging to the women computers, and realised many of the books were missing.
"I started doing a little bit of digging and eventually came across some proof that we might have boxes in storage off-site, which is very common for libraries around Harvard."
Smith Zrull found 118 boxes, each containing between 20 and 30 books.
Inside were more notebooks from the women computers, as well as notebooks from astronomers who predated photography and made hand-drawn sketches of planets and the moon.
"People didn't know they existed when they were in storage," Smith Zrull says. "As different curators came and went here, I suppose people forgot they were there."
To resurrect their legacy, she enlisted the help of librarians at the centre, who planned to go through the boxes and begin the labour-intensive process of cataloguing them. Project PHAEDRA (Preserving Harvard's Early Data and Research in Astronomy) was born.
'OK, we've hit pay dirt'
But then there were two quick discoveries in the plate stacks - Smith Zull found a handwritten catalogue of the books from 1973.
"At some point in 1973, someone who we assume is named 'Joe Timko' went through all of these boxes at an item level and recorded as much information as he could find," says head librarian Daina Bouquin. They had no sense of why it was done, "but we thought, 'OK, we've hit pay dirt.'"
Then someone found a typewritten version of the 1973 catalogue, adorned with a Post-it saying "Finally done! Rachel." On the very last page was a handwritten path to a computer file, a spreadsheet on a Harvard server that hadn't been accessed since 2001.
The discovery sped up the digitisation project by months, if not years. The librarians went from having only 30 characters on each box, to machine-readable data they could quickly turn into real records.
"Thank you Joe Timko and possibly Rachel, wherever they may be," says Bouquin.
The library has completed transcription of about 200 volumes. There are many more to come - nearly 2,300 - but the work has begun. Right now, notebooks from two women are listed on the Smithsonian Transcription Center website.
Bouquin hopes the public will help transcribe the books, but anticipates it will still be years before everything is readable.
"You'll be able to do a full-text search of this research," Bouquin says. "If you search for Williamina Fleming, you're not going to just find a mention of her in a publication where she wasn't the author of her work. You're going to find her work."
'She's the one who really found it'
Fleming is the first famous woman computer from Harvard. Fleming emigrated to the United States from Scotland in the late 1870s.
While pregnant, she was abandoned by her husband and found work as a maid in the home of Edward Pickering, the observatory director. In 1881, Pickering hired Fleming to work in the observatory.
She would go on to discover the Horsehead Nebula, develop a system for classifying stars based on hydrogen observed in their spectra and lead more female computers.
Wolbach Library unveiled a new display in early July showcasing Fleming's work, including the log book containing the nebula discovery.
"When the [Horsehead Nebula] was discovered, it was just a little 'area of nebulosity in a semi-circular indentation,'" says librarian Maria McEachern, who has helped the team sort through the notebooks.
"Years later that it became known as the Horsehead Nebula," McEachern says. A male scientists at another institution who named it was the one who got credit.
"It wasn't even until recently that people have been doing more scholarship and finding out that, yes, she's the one who really found it."
But Fleming was just the first computer to make her mark on astronomy.
Pickering hired Henrietta Swan Leavitt in 1895. She was tasked with measuring and cataloguing the brightness of the stars. Her major discovery - a way to allow astronomers to measure distance in space, now known as "Leavitt's Law".
Annie Jump Cannon joined the observatory in 1896 and worked there until 1940. Cannon created the Harvard Classification System for classifying stars, which is the basis of the system still in use today.
Cecilia Payne-Gaposchkin came to the Observatory in 1923 and earned a doctorate from Radcliffe College (FC) in 1925, but she struggled to get recognition from Harvard.
For years she had no official position, serving as a technical assistant to then-director Harlow Shapley from 1927 to 1938. It wasn't until the mid-1950s that she became a full professor and later, the first woman to head a department at Harvard.
And because of Smith Zrull's discovery, transcription of each of these women's notebooks are underway.
'They've always been there'
"I like to think resilience goes a long way, but I think some of these women go a little above and beyond what we think of when we think of overcoming things," Bouquin says.
Both Bouquin and Smith Zrull said they want to give young girls more role models like the Harvard computers - role models who weren't well-known when they were young.
"Yes, look at Sally Ride, look at modern women who people associate with the space-based sciences, but go back further," Bouquin says. "They've always been there. As long as they could be, they were there."
Smith Zrull - who hated history as a teenager - said she struggled to find women who encouraged her.
"It really took me a long time to start to find women who I felt were like me, who did important things," Smith Zrull said.
"I think more women need to know, you're not alone, you can do it."
The World is a co-production of the BBC World Service, PRI and WGBH. You can listen to more here.