This article concerns the murders of eight people in Austin, Texas, in 1884 and 1885. Six of the eight victims were Black women, killed during a period of intense racial violence and inequality in post-Reconstruction Texas. The investigation was shaped by the racial politics of the era in ways that determined which deaths were investigated urgently and which were not. Both the murders and the investigative failures are treated here with the seriousness they deserve. This is also a case in which most of the surviving documentation comes from newspapers and legal records of a period in which the deaths of Black Americans received systematically different treatment than those of white Americans. That asymmetry is part of the story.
Part One: The City on the Colorado
In the autumn of 1884, Austin, Texas, was a city in the process of becoming something. It had been the state capital since 1839, had survived the Civil War and the confused, violent years of Reconstruction that followed, and was, by the mid-1880s, experiencing the first phases of what its boosters called transformation. The new state capitol building — an enormous pink granite structure that would become one of the most recognisable buildings in Texas — was under construction. The University of Texas had opened its doors in 1883. A new railway spur had connected the city to the main national network. Population was growing. Hotels were being built. Austin wanted, urgently, to be taken seriously as a modern American city.
Underneath this ambition was a social structure that the aspirations of Austin’s civic boosters did not discuss in their promotional literature. The city’s Black residents — approximately a quarter of the total population by the early 1880s — lived in a condition of legal freedom that was, in practice, heavily circumscribed by the violence, economic exclusion, and systematic discrimination of post-Reconstruction Texas. The Reconstruction amendments to the Constitution had formally abolished slavery and granted citizenship and voting rights to Black Americans. The political and paramilitary campaigns of the 1870s had systematically dismantled the enforcement mechanisms that might have made those rights real. In 1884 Austin, Black residents could not vote in any election that mattered, could not serve on juries, could not access the same schools or most of the same public spaces as their white neighbours, and could not expect the police force — which was entirely white — to respond to their concerns with the same urgency it brought to the concerns of white citizens.
Many of the city’s Black residents worked as domestic servants in the households of white Austin families — cooks, laundresses, nursemaids, general household workers who lived in small outbuildings and servants’ quarters behind their employers’ main homes, occupying a position of physical proximity to wealth combined with near-total social invisibility. They were the people whose labour made the aspirational households of 1880s Austin function. They were also, in December 1884, the people whose deaths were about to be systematically underinvestigated by a police force that found it difficult to see them as fully equal victims.
The first victim died on the night of 30 December 1884.
Part Two: Mollie Smith and the Winter of 1884
Before the first canonical murder, there had been a pattern of attacks that, in retrospect, the Austin police should have connected. Between July and November 1884, at least three women — all Black, all domestic servants — had been attacked in their beds at night. Two had been stabbed in the face. One had been struck on the head with a smoothing iron. All had survived. The attacks had been treated as separate incidents. They had not generated significant newspaper coverage. They had not produced a suspect.
On the night of 30 December 1884, twenty-five-year-old Mollie Smith was asleep in the servants’ quarters behind the home of her employer, W.K. Hall, on Hickory Street. She worked as a cook. Her boyfriend, Walter Spencer, was sleeping in the same building. Sometime in the small hours of 31 December, an intruder entered the outbuilding. Mollie Smith was struck on the head with an axe. She was dragged outside into the yard, where she was raped and mutilated. Walter Spencer was struck on the head during the attack and left for dead. He survived, though severely injured.
Mollie Smith’s body was found the following morning in the snow near the family’s outhouse, in a pool of blood so extensive that the ground around her had been turned dark red. The Austin Daily Statesman covered the murder in a few paragraphs on an inside page. The police investigated briefly. Their most immediate action was to arrest Walter Spencer, the surviving boyfriend — the obvious suspect in the conventional investigative logic of the period, which assumed that violence against Black women was most likely the work of Black men they knew. Spencer was tried and acquitted within two days. The investigation then largely stalled.
Five months passed. Then, on 6 May 1885, Eliza Shelley, a servant who worked for a physician named Dr Lucian Johnson on West Pecan Street, was found dead in her room in the servants’ quarters. She had been struck so many times with an axe that, in the account of a reporter who attended the scene, her skull had been almost split in two. The Austin papers covered the story briefly. It was the second dead Black servant girl in five months. It was not yet a pattern that the city’s press or police appeared ready to treat as requiring urgent public attention.
Part Three: The Year of Terror
In the twelve months that followed Eliza Shelley’s murder, Austin was subjected to a campaign of nocturnal violence that killed six more people and left several others seriously injured. The attacks followed a consistent pattern: they occurred at night, while victims slept; the preferred weapon was an axe or a sharp instrument; the victims were attacked in or near their sleeping quarters; and the killer left almost no useful evidence beyond a distinctive set of bare footprints — missing one toe from the right foot — that appeared repeatedly at the scenes.
On 23 May 1885, Irene Cross, a servant living on East Linden Street, was attacked with an axe as she slept. A reporter who arrived at the scene described her head injuries in terms that suggested she had, in his words, been scalped. She died of her wounds.
A relative pause through June and July. Then, in August, three attacks in quick succession. Clara Dick, a domestic servant, was attacked and seriously wounded. Then Rebecca Ramey, another servant, was attacked — she survived, though with severe injuries. But her eleven-year-old daughter Mary Ramey, who had been sleeping beside her mother, was killed. Mary Ramey was a child who had been in her mother’s bed. She was eleven years old.
On 6 September 1885, two more people were killed in the same night. Gracie Vance, a servant, was attacked and killed. Orange Washington, a man who had been nearby, was also killed. Two other women who were present — Lucinda Boddy and Patsey Gibson — were attacked and left seriously wounded.
By the autumn of 1885, Austin was beginning, belatedly, to register what was happening. The Austin Daily Statesman had been increasing its coverage. Citizens were beginning to ask questions at public meetings. The mayor had been confronted about the inadequacy of the police response. Approximately 400 men had been arrested and questioned at various points through the year; all had been released. The police had no viable suspect.
It was at around this point — the autumn of 1885 — that a local writer named William Sydney Porter wrote a letter to a friend that gave the killer the name he would carry into history. Porter, who would later become famous under the pen name O. Henry, was in his early twenties and working at the General Land Office in Austin. His letter, referring to the killings with the darkly comedic register he had already developed as a writer, called the perpetrator “the Servant Girl Annihilator.” The name was sharp, memorable, and reductive — it described the victims entirely in terms of their employment — but it stuck. The newspapers began to use it. It has been used ever since.
Part Four: Christmas Eve
On the evening of 24 December 1885, the pattern that had defined the year shifted in a way that changed everything about how the case was subsequently treated.
Two women were killed that night, approximately a mile apart. The first was Susan Hancock, the wife of Moses Hancock, a white woman living in her own home on San Marcos Street — not a servant, not a resident of servants’ quarters, not a Black woman, but a white woman from a middle-class household, killed as she slept in her own bed. The second, killed later the same night, was Eula Phillips, seventeen years old, also white, killed in her home on Water Street. Eula’s husband James Phillips was also attacked during the assault on his wife; he survived with head injuries.
The reaction was immediate and unmistakable. For most of 1885, the deaths of Black servants in outbuildings had produced limited press coverage and a slow, resource-constrained investigation. The deaths of two white women on Christmas Eve produced something altogether different. The Austin Daily Statesman ran extensive front-page coverage. The governor was informed. Citizens held public meetings demanding action. The police force, which had responded to months of Black servant deaths with a relatively unhurried pace, found itself under intense pressure from white Austin to solve the case immediately.
The investigative response to the Christmas Eve murders also demonstrated, with brutal clarity, the racial assumptions at the heart of the entire investigation. The police arrested two men and charged both of them with the murders of the women they had been sharing a house with. Moses Hancock was charged with his wife Susan’s killing. James Phillips, Eula’s husband — who had himself been attacked and injured — was charged with Eula’s killing. Both men were detained. The theory that these were domestic killings — husband-murders, husband-murders disguised as the work of the Annihilator — was embraced by a police force that, faced with white female victims, appeared to find it easier to assume the killer had been a familiar man than to accept that the same force terrorising the city’s Black servant population had turned on white women too.
James Phillips was tried and convicted of killing Eula. He was sentenced to seven years in prison. The conviction was overturned within six months on the grounds of insufficient evidence. Moses Hancock’s trial ended in a hung jury. Neither man was ultimately convicted. The theory that the Christmas Eve killings were domestic murders committed by husbands died of its own inadequacy. But its brief life illuminated the racial logic of the investigation with exceptional clarity: the killer of Black servants had been imagined as a Black man acting from familiar proximity; the killer of white women had been imagined as a white man acting from familiar proximity. In neither case had the same sustained outward-looking investigative pressure been applied that might have found the actual perpetrator.
After Christmas Eve 1885, the killings stopped.
Part Five: The Investigation’s Failures
The aftermath of the Christmas Eve murders produced the most concentrated investigative effort of the entire case — and, ultimately, the most visible failure. The Austin police conducted sweeps of the city’s population in numbers they had never attempted for the servant murders. Approximately 400 individuals were eventually questioned. Specific suspects were identified, investigated, and eliminated. The investigation went on for months after the last murder without producing an arrest.
The racial politics of the investigation had shaped it from the beginning in ways that limited its effectiveness. The assumption, held consistently by Austin’s white press and police through most of 1885, was that the killer was a Black man — partly because most of the victims were Black, and partly because the prevailing prejudice of post-Reconstruction Texas could not readily accommodate the idea that a white man would have reason to kill Black servants. When the Christmas Eve murders demonstrated that the killer was willing to kill white women, this assumption was disrupted, but it was replaced not with a more methodical search but with the husband-murder theory, which was itself a product of the same instinct to locate the killer within a familiar domestic frame rather than to look outward.
A specific investigative failure of particular importance was the handling of a piece of physical evidence that should have been central to the entire case. At multiple scenes, investigators had recovered bare footprints in soil or snow that appeared consistently — the prints of adult feet, missing one toe from the right foot. The prints had been noted. They had been documented to a degree. They had not been systematically cross-referenced with suspects in a way that might have been possible even with the investigative tools of 1885.
There was also, by some accounts, a more fundamental archival loss. Reports from the early twentieth century, referenced in some historical treatments of the case, suggest that Austin police files from the 1884-85 period were destroyed — whether by fire, deliberate disposal, or administrative neglect varies by account. The destruction of physical evidence meant that later investigators and historians working the case have had to reconstruct it primarily from newspaper reports, which were themselves shaped by the racial and class biases of their era.
The most thorough modern account of the case is Skip Hollandsworth’s 2016 book The Midnight Assassin: Panic, Scandal, and the Hunt for America’s First Serial Killer, which drew on newspaper archives, court records, and historical research to produce the most complete reconstruction of events currently available. Hollandsworth’s work established the case’s significance in American criminal history and revived public interest in a set of events that had been substantially forgotten for most of the twentieth century.
Part Six: The Suspects
The investigation of the Servant Girl Annihilator produced no conviction and no definitive identification of the killer. In the 140 years since the murders, the case has accumulated a small number of named suspects whose candidacy has been argued with varying degrees of seriousness.
The most discussed suspect in the contemporary investigation — and the one who has attracted the most sustained attention from modern researchers — is Nathan Elgin, a nineteen-year-old Black resident of Austin who worked as a domestic cook and who was known to have moved through the relevant neighbourhoods. In January 1886, approximately a month after the final murders, Elgin was shot by Austin police officers after he grabbed a woman from outside a saloon in a drunken state and attempted to drag her away. He was shot while resisting arrest and died the following day of his wounds. When his body was examined, it was found that he was missing a toe from his right foot. The killing stopped. The footprint evidence from several crime scenes had consistently shown a right foot with a missing toe. Elgin fit the physical evidence, he fit the geographic profile, and the violence after his death ended.
The case against Elgin is entirely circumstantial. He was never charged with any of the murders. He could not be questioned after his death. The physical evidence linking him to the scenes — the missing toe, the footprints — is suggestive but not conclusive, as missing toes are not entirely uncommon. His youth (nineteen in 1886, meaning he would have been seventeen at the start of the killings in December 1884) has been both used to argue for and against his candidacy, depending on whether one considers early-starting serial violence to be plausible. He remains the most likely single suspect in the mainstream historical assessment, but he is a probable rather than a proven one.
The second named suspect emerged from among the staff of the Pearl House hotel, a prominent Austin establishment in the 1880s. A Malaysian cook known in the historical record only as Maurice — his surname, if it was known at the time, did not survive into the available documentation — worked at the Pearl House during the relevant period. The hotel was located near several of the murder sites. Contemporary newspaper reports noted the geographic proximity and raised the possibility of Maurice’s involvement. The additional detail that gave the theory its subsequent legs was this: at approximately the time the Austin killings stopped, Maurice left Austin, spent some time in New Orleans, and eventually settled in London. The Austin-American Statesman, in November 1888, made the connection explicit — the killings in London that autumn bore certain similarities to the Austin murders, and Maurice’s trajectory from Austin to London made him, on the face of it, a plausible link between the two sets of crimes.
This is the Ripper connection, and it deserves honest treatment. The theory that the Servant Girl Annihilator became Jack the Ripper has been a recurring feature of both case’s speculative literature since 1888. It was raised in contemporary newspapers. It has been restated by various researchers over the intervening century. The version involving Maurice the Malaysian cook has been the most frequently cited vehicle for the theory. The version involving James Maybrick — the Liverpool cotton merchant already discussed in the Ripper piece on this site — has been argued by the writer Shirley Harrison.
The honest assessment of the Ripper connection theory is that it is more interesting than it is evidenced. The Austin and London murders share some structural features — nocturnal attacks on women from marginalised communities, mutilation of bodies, the use of cutting instruments. They also differ in significant ways — the Austin killer used an axe rather than a knife, attacked victims primarily while they slept in their beds rather than in streets, and left a specific physical signature (the toe-missing footprints) that does not appear in the Whitechapel evidence. The time gap — three years between the Austin killings ending and the Whitechapel murders beginning — is plausible but proves nothing. The theory requires a mobile killer who paused for three years and then resumed with modified methods in a different country on a different continent. It is possible. It is not demonstrated by any physical evidence currently available.
What the Ripper connection theory has done, as the Austin killings researcher and journalist Skip Hollandsworth has observed, is provide the Austin case with a frame of reference that audiences in the present century can recognise. The Servant Girl Annihilator is frequently introduced in popular coverage as “the killer who may have been Jack the Ripper.” This framing is not entirely dishonest, but it does something problematic to the case’s own integrity. It suggests that the Austin murders matter primarily as a possible precursor to a more famous case, rather than as a case with its own victims, its own investigative history, and its own significance in the history of American crime.
Part Seven: Why This Case Was Forgotten
The question of why the Servant Girl Annihilator case was forgotten — why it does not appear in the standard American true crime canon with anything like the prominence it deserves — has a straightforward answer that the case’s recent historians have been willing to state directly.
Six of the eight victims were Black women. They were killed in 1884 and 1885, two decades after the abolition of slavery, in a Southern state where the political and social infrastructure of the post-Reconstruction period was actively working to reconstruct as much of the pre-war racial order as the law would permit. The deaths of Black servants in outbuildings behind white employers’ homes were not treated, by Austin’s white press or police, as the deaths of people whose loss demanded the same sustained public attention as the deaths of white women of equivalent status would have demanded. The newspapers covered the murders briefly. The investigation was slow. The victims were frequently identified in print by their occupations — servant girls, cooks, laundresses — rather than as individuals with names and histories and families who were grieving them.
The shift in press and police attention that occurred after the Christmas Eve murders of Susan Hancock and Eula Phillips was not simply a matter of the investigation finally getting serious. It was a matter of the investigation getting serious about the right kind of victims — the victims whose deaths the white readership of the Austin Daily Statesman considered appropriately alarming. The servant murders had been alarming to Austin’s Black community from December 1884. They had been less than fully alarming to official white Austin until white women died.
This dynamic — the differential treatment of victims by race and social status — is not unique to Austin in 1884 and 1885. It appears, in forms that vary with the specific social context, in almost every major unsolved case involving victims from marginalised communities: the Yorkshire Ripper’s two-tier framing of “prostitutes” and “innocents,” the Milwaukee Police Department’s handling of the communities from which Jeffrey Dahmer’s victims came, the West Mesa murders in New Mexico, and many others. The Servant Girl Annihilator case is, among other things, the first documented instance of this dynamic in what is now recognised as an American serial killing case.
The case was further eclipsed by Jack the Ripper. The 1888 London murders generated a volume of press coverage, cultural commentary, and subsequent historical investigation that had no American parallel. The Whitechapel killings became, almost immediately, the dominant reference point in English-language culture for the idea of a serial killer preying on vulnerable women. The Austin murders, which had preceded them by three years, receded into the background of a story in which they were at best a footnote and at worst a disputed prelude.
Skip Hollandsworth’s 2016 book The Midnight Assassin was the first serious attempt to recover the case from this obscurity. By drawing on the Austin newspaper archives, the Travis County court records, and the available historical research on 1880s Austin social conditions, Hollandsworth produced an account that treated the murders as a significant event in their own right rather than as a possible precursor to a more famous case. The book has been the primary vehicle through which general audiences have encountered the Servant Girl Annihilator case in the present century.
Part Eight: The Names That Matter
Eight people were killed by the Servant Girl Annihilator between December 1884 and December 1885. Their names are, in some cases, all that survives of them in the historical record. The documentary asymmetry that gave more coverage to white Austin than to Black Austin in the 1880s means that the biographical details available for the victims are sparse compared to what would be recoverable for people of greater social visibility in the same period.
Mollie Smith, twenty-five, killed 30 December 1884. She worked as a cook for the Hall family on Hickory Street. She was found in the snow beside the outhouse of the property where she worked, in a pool of blood that had spread several feet across the frozen ground. The boyfriend who had been sleeping in the same building survived his injuries and was tried and acquitted for her murder within two days of being arrested.
Eliza Shelley, killed 6 May 1885. She worked for a physician on West Pecan Street. Her skull had been struck so many times with an axe that reporters at the scene described what they saw as the effect of being almost split in two.
Irene Cross, killed 23 May 1885. She lived on East Linden Street. The axe injuries to her head were described by a reporter as appearing to have scalped her.
Mary Ramey, eleven years old, killed in August 1885. She had been sleeping beside her mother Rebecca Ramey when the killer entered the room. Rebecca survived, severely injured. Mary did not. She was eleven years old. She is the youngest of the confirmed victims, and the one whose presence in the case most directly refutes any theory of selective targeting that the documentary record might otherwise support. She was a child who had been in the wrong bed.
Gracie Vance, killed 6 September 1885. A domestic servant.
Orange Washington, killed 6 September 1885. The only male victim in the confirmed list. He was present at the scene of the attack on Gracie Vance and was killed in the same incident. His name appears less frequently in the historical accounts than those of the female victims.
Susan Hancock, killed 24 December 1885. A white woman, the wife of Moses Hancock, killed in her bed in her own home on San Marcos Street. Her death on Christmas Eve changed the texture of the investigation’s public attention in ways that the preceding deaths of Black servants had not.
Eula Phillips, seventeen, killed 24 December 1885. The youngest of the Christmas Eve victims and the last confirmed victim of the Annihilator. She had recently married James Phillips, who was also attacked that night and survived. James Phillips was subsequently tried and convicted of her murder, then acquitted on appeal. He maintained for the rest of his life that he had not killed her. The overwhelming historical assessment is that he was telling the truth, and that Eula Phillips was killed by the same person who had been killing through the preceding year.
Three other women — Clara Strand, Christine Martenson, and Clara Dick — were seriously injured in attacks attributed to the Annihilator but survived. Rebecca Ramey, Mary’s mother, survived her attack and lived with the knowledge of her daughter’s death. Their names belong alongside those of the dead in any honest accounting of what happened in Austin in 1884 and 1885.
A Final Note
The Servant Girl Annihilator case has never been officially solved. No suspect was ever charged in connection with the murders. Nathan Elgin, the most plausible candidate in the mainstream historical assessment, died in a police shooting in January 1886 without ever being questioned about the murders, and the circumstantial case against him — persuasive as the missing-toe detail is — cannot be converted into anything approaching a proven conclusion from the surviving record.
The case is formally 140 years old. The victims’ direct descendants, if any survive, are several generations removed from the events. The physical evidence, to the extent it was preserved at all, has long since been lost to institutional neglect and the destruction of records. The Austin police department holds no active investigation. The case will, in all probability, remain unsolved.
What should not remain unsettled is the question of why the case was forgotten. The answer is not mysterious or complex. It was forgotten because most of the victims were Black women whose deaths were treated, by the institutions of 1880s Austin, as less fully requiring of justice than the deaths of people higher up the social hierarchy. It was forgotten because Jack the Ripper happened three years later and became the culturally dominant reference point for everything the Austin murders prefigured. It was forgotten because the mechanisms by which we collectively remember historical violence tend to amplify the cases that most directly resemble what we already find culturally legible — the Ripper’s letters, and to allow the cases that don’t fit the familiar templates to fade.
The Servant Girl Annihilator case predates the Ripper by three years. It may or may not have involved the same killer. It certainly involved the same structural conditions — vulnerable women, marginalised communities, an investigating institution that responded to different victims with different urgency, a press that covered different deaths with different degrees of alarm — that appear in case after case in the history of violence against women from the 1880s to the present. Understanding those conditions is, in the end, more important than naming the specific man who swung the axe in Austin in 1884 and 1885, though both matter. The name matters for Mollie Smith. The conditions matter for everyone who came after her.
Mollie Smith. Eliza Shelley. Irene Cross. Mary Ramey, eleven years old. Gracie Vance. Orange Washington. Susan Hancock. Eula Phillips, seventeen. Eight people. One year. One city that was trying, at the time, to present itself to the world as a modern American metropolis. The modernity it achieved did not include justice for the people killed in its servants’ quarters during the year it was being built.
