Little Girl Disappeared While Bringing Lunch to Sisters — 32 Years Later, He Threw Away One Napkin 

There is a bridge in the north end of Tacoma, Washington called the Proctor Bridge, and underneath it the ground falls away into a gulch. In late March, the trees have not filled in yet. You can still see through them. From the picnic tables at the top, Puget Park looks small and ordinary, and about the size of a city block.

 From the bottom of the ravine, it does not look small at all. On a Wednesday morning in the spring of 1986, three sisters walked into that park. The oldest of them was 12 years old. She wore glasses. She had long blonde hair. She was small for her age, small enough that people who did not know her guessed lower. That morning, the three of them realized they had left their lunch at home.

 She was the one who noticed. She was the one who went. So, she got on her bicycle, and she rode out of the park, back through the neighborhood, into an empty house, and then all the way back again. And here is the detail that this entire case turns on. The one thing you need to carry with you for the next 30 two years. She came back.

 She chained her bicycle to the rack in the same place she always chained it. She set a brown paper bag down on a picnic table, sandwiches inside, made at home, carried back for her sisters. Then, she walked off to find them. When the two younger girls came back to the picnic area in the early afternoon, the bag was still sitting on the table.

 The bicycle was still chained to the rack. Their sister was not there. So, they called for her. The three of them had a call they used between themselves, a sound they had invented, the kind of signal that carries through bare trees and down across a ravine, and always, always comes back. They called it into the gulch. Nothing came back.

 You are watching Cold Case Model. Her name was Michella Welch, and this case stayed open for 32 years. It did not close because a witness finally came forward. It did not close because somebody confessed. It did not close because a suspect slipped up. It closed because a family tree built out of the genetic material of distant cousins who had never met him, walked detectives to the front door of a man who had been living quietly a short drive from that park the entire time.

 He had never been arrested for anything. He had a job people trust you with. He had been invisible to every tool law enforcement had and he stayed invisible right up until the moment he was not. If you are new here, subscribe. Everything on this channel is built the same way out of public records, court filings, and published reporting.

Nothing is invented, not one line of narration, not one image on your screen. No artificial intelligence generated pictures of real people, no imagined conversations, no made up thoughts put into the head of a child who cannot correct me. That standard is why these take weeks instead of hours. Subscribing is the entire reason it stays possible.

Now, back to Tacoma. Richella Welch was 12 years old in the spring of 1986. She was a sixth grader. She was the oldest of three girls and her sisters were both younger than she was. Her father was no longer in the home. Her mother had spent several hard years putting a life back together with three children in it.

 By that March, that work was starting to show. There was a house. There was a new job and there were piano lessons for all three girls, which is the sort of detail that tells you almost everything you need to know about what a mother was trying to do with what she had. Piano lessons are not a necessity. Piano lessons are a statement.

 They mean somebody has decided that the hard part is behind them and it is time to start building forward again. They were about six weeks into that new chapter. Family members later described Richella as fiercely independent, as responsible far beyond her years, as kind. A member of her family wrote publicly decades later that what happened to Michella was the most devastating thing that ever happened to them and that it left a hole in their lives that nobody was ever able to fill.

Think about how long it takes to be able to write a sentence like that. Not to feel it, to write it, to sit down and put it into words for other people to read. That took more than 30 years. What all of it adds up to, in practical terms, is a 12-year-old who watched her mother work and quietly picked up whatever she could reach.

 Second set of hands. Second set of eyes. Which is why when three girls in a park realized they had nothing to eat, nobody had to ask her to go. The park was known ground. It was a couple of miles from the house. It sat right by where the piano lessons were held, which is why the whole morning had been planned around it.

 It was the kind of place a family uses so often that it stops being a destination and just becomes part of the day. Nobody thinks about it. Nobody makes a decision about it. You just go. That is worth sitting with for a second because it is the part of this story that quietly terrifies every parent who hears it.

Nothing unusual happened that morning. There was no bad decision. There was no warning that got ignored. There was no argument. No stranger phoning the house. No reason for anybody to have done anything differently. It was a Wednesday in a public park 2 miles from home in the middle of the morning in a neighborhood full of families in a city where children walked places by themselves as a matter of course.

 The ordinariness is not a detail of this case. It is the case. Let me walk you through the timeline because the timeline is the only witness that never changed its story. The three girls arrived at Puget Park at around 10:00 in the morning. At some point in the next hour, they realized the lunch had been left behind at the house.

 At around 11:00, Michella got on her bicycle and rode home for it. She was gone for a while. Making sandwiches for three people takes a few minutes. Riding 2 miles each way takes longer. And then she came back into that park with a paper bag in her hand, and she locked her bicycle to the rack, and she set the bag down on the picnic table, and everything in the world was still completely fine.

 Her sisters were not at the table. They had walked over to a nearby business to use the restroom. So, she went to look for them. There was a boy in that park that morning. He was 13. He was a classmate of Michella’s, and he told detectives afterward that he had seen a man standing underneath the Proctor Bridge earlier in the day.

 He said the man was watching the girls. He described a white male in his middle 20s, thin, wearing a torn denim jacket, and dirty white tennis shoes. That description became a composite sketch, and that sketch was put in front of an entire city. Nobody ever came forward. Later, at around half past 1:00 in the afternoon, a second person reported seeing Michella near the edge of the gulch, talking to a man.

 The man appeared to be pointing down a trail. The description of that second man did not match the description of the first one. I want you to hold both of those men in your head, because I am going to come back for them, and what the record eventually says about them is not what you are expecting.

 Her sisters came back to the picnic area at around a quarter past 1:00. The bag was on the table. They waited. They played for a while near the ground under the bridge. They came back to the table again. The bag was still on the table. The bike was still on the rack. That is when they went to the edge and used the family call, and that is when nothing came back. They found their babysitter.

 The babysitter reached their mother. Officers arrived at around 10 minutes past 3:00 that afternoon, and began to search the park. The search went into the night, and it grew, and it moved down off the flat ground, and into the gulch, which is not a manicured city park. It is steep. It is thick. It is the kind of terrain where a person can be 20 ft away from you and completely out of sight.

 At around half past 11:30 that night, a tracking dog led searchers to a spot more than a quarter of a mile from the picnic table. Michella was found there. The medical examiner ruled her death a homicide. She had been harmed. I am not going to take you any further into that and this channel never will.

 There is nothing in the specific that honors her and there is nothing in them that you need in order to understand what happened here or why it took so long to answer. She was a 12-year-old child. The details of the worst hour of her life are not entertainment and they are not going to be turned into any on this channel. What you do need is one detail from that night and it is a boring one.

 Evidence was collected at that scene in 1986 and it was kept. Nobody standing in that ravine in the dark had any idea what that evidence eventually be able to do because the science that would eventually do it had not been invented yet. It would not exist in a usable form for another 20 years. The people who bagged it and labeled it and filed it away were not saving it for a technology they were waiting on.

 They were following procedure. Everything that happens in the rest of this story, every answer, every arrest, every name rests on that one unglamorous decision made by people who could not possibly have known why it mattered. Remember that, too. It comes back. Tacoma changed after that spring.

 It changed in the way cities do when a child is taken in daylight in a public park. Parents started driving children to places they had always been allowed to walk. Kids stopped taking the shortcut through the trees. The route through the park stopped being a route. The composite sketch went everywhere. Into newspapers, onto walls, into to hands of anybody who would take one.

Thousands of tips came in. In the first year alone, investigators logged more than 10,000 hours on this case. That is not a symbolic figure. That is a genuine, enormous, sustained commitment of a city’s police resources aimed at one crime. And none of it moved. Not because the detectives were lazy, and not because they did not care.

 You are about to find out exactly why it did not move. And the reason is going to be worse than incompetence. Because it was not anybody’s fault, and it cost 30 years anyway. Before I take you into it, do one thing for me. Go down to the comments and type the words I’m still here. Not for the algorithm. Type it because of what those three girls had.

They had a sound that belonged only to them. And the entire understanding between them, the whole silent contract of it, was that when one of them sent it out, one of the others sent it back. That was the deal. That was what it was for. You call, and somebody answers. And that is how you know you are not alone in there.

 On the 26th of March, 1986, two little girls stood at the edge of a ravine and sent that sound out into the trees. And for the first time in their lives, nothing came back. So, type it. I’m still here. Fill the comments with the answer that Gulch never gave them. And subscribe, because cases like this one are built out of records, filings, and archives.

 And that work only keeps happening if people stay. Now, back to Tacoma, and to the summer of 1986. On the 4th of August, a little more than 4 months after Michella Welch died, a 13-year-old girl named Jennifer Bastian left her house on a bicycle. She was riding to Point Defiance Park, which sits roughly 3 miles from Puget Park, in the same north end of the same city.

 She was training for a bike tour. She left a note behind saying what time she would be home. She never came home. Her body was found weeks later in a wooded section of that park. She had been killed. Now, line the two of them up the way investigators had to line them up. Jennifer was 13. Michella was 12. Both girls were blonde.

 Both girls were alone. Both girls were on or near a bicycle. Both girls were in a public park in the north end of Tacoma. Both were taken in daylight. Both were found in wooded ground. Both cases went silent afterward with no name, no arrest, no confession, nothing. Four months apart, three miles apart.

 Put yourself in that room. You are a detective in 1986. You have no DNA science worth anything. You have no national databases you can actually work with. You have witness statements, tips, sketches, and pattern. Pattern is all you have, and the pattern is screaming at you. So, the cases were consolidated. The theory became official.

 One man, two victims, one offender hunting children in the parks of North Tacoma in the summer of 1986. That theory held for almost 30 years. It was wrong. I want you to really understand what a wrong assumption does to an investigation because this is the part of the story that most retellings skip, and it is the part that actually explains the missing decades.

 Once the two cases are joined, every suspect has to fit both of them. That sounds reasonable. It is catastrophic because now, if a man was in Tacoma in March but demonstrably somewhere else in August, he is cleared. Not just of Jennifer’s case, of both of them. He falls out of the file. He is done.

 And if a man could have done it in August but has an alibi for March, exact same thing. Gone. Every man who could only have committed one of these two crimes was systematically eliminated by the requirement that he must have committed both. Think about the shape of that. The filter that detectives were using to narrow their suspect pool was, unknowingly, a filter designed to remove the right man.

 They were not failing to look hard enough. They were looking extremely hard for 30 years through a lens that was quietly deleting the answer every time it came near. They were searching for the wrong shape of a crime. The ’80s turned into the ’90s. Tips kept arriving and kept going nowhere.

 Suspects were interviewed and released. Detectives retired. New detectives inherited the boxes. The composite sketch of the thin man in the denim jacket kept getting circulated, and nobody ever matched it to a face. And the whole time, in an evidence room, the material collected in that ravine in 1986 sat in storage doing nothing, waiting for a science that had not been invented. The wait ended slowly.

 By the middle of the 2000s, forensic DNA analysis had finally matured to the point where it could do something useful with old, degraded biological evidence. Techniques that had been science fiction in 1986 were becoming routine. In 2006, 20 years after Michella’s death, investigators were finally able to develop a usable male DNA profile from the evidence preserved in her case.

 20 years, and now they had him. Not his name, but him. The actual, specific, unrepeatable genetic signature of the man who did this, sitting in a file in Tacoma. There was only one thing left to do with it. They submitted it to CODIS. CODIS is the national DNA database. It holds the genetic profiles of convicted offenders and arrestees, millions of them, gathered over decades.

 It is the single most powerful identification tool in American law enforcement. You put an unknown profile in, and the system compares it against every known profile it holds. And if your man is in there, the machine tells you his name. They ran it, no match. Sit with that, because it is not a small thing.

 It is the whole reason this case took another 12 years. CODIS can only find a man who has already been caught. That is what it is. It is a database of people the system has already put its hands on. If a man has been convicted of a qualifying offense, he is in there. If he has never been convicted of anything, he is not in there, and the machine has nothing to compare him to.

 The most powerful tool in American forensics returned a result that, translated into plain English, meant this: The man who killed Michella Welch has no record. He has never been convicted of anything that would put him in this system. He has been living an ordinary life, holding down ordinary jobs, and as far as this database is concerned, he does not exist.

 He was not hiding. He did not have to hide. He was just a man with a clean record, and a clean record made him invisible. So, the profile went into the system, and the system said nothing, and the case sat. Seven more years went by. In 2013, investigators finally managed to develop a DNA profile in the Jennifer Bastian case as well, from evidence preserved from 1986.

 They ran that one through CODIS, too. No match there, either. Same story, same silence. Two profiles, two unknown men, neither of them in the system. And then somebody did the obvious thing, the thing that nobody had been able to do for 30 years, because for 30 years there had been nothing to do it with. They took the profile from Michella’s case, and they took the profile from Jennifer’s case, and they compared them to each other.

 They did not match. Two different men, not one killer working two parks in a single summer. Two entirely separate individuals, unconnected to each other, who had each taken a girl’s life within a few miles of the same city in the same year, and who had then each walked back into an ordinary life and stayed there undisturbed for decades.

 A single comparison run in an afternoon destroyed the foundation that 30 years of investigation had been built on. Every consolidated suspect list, every man cleared of both because he could be placed away from one, every theory, every profile of a single offender’s behavior, every hour of the more than 10,000 logged in that first year and the countless thousands after it, all of it filtered through an assumption that the evidence had now proven false.

 It was not wasted work, but it was work aimed at a target that was never there. And while we are here, let me close the loop I promised you. Those two men in the park, the thin man in the torn denim jacket under the bridge, the man near the edge of the gulch who appeared to be pointing down a trail, neither was ever identified, neither was ever charged, neither, in the end, had anything to do with the man the DNA eventually named.

Two composite sketches, 30 years of a city studying two faces on a poster, and when the answer finally arrived, it came from a completely different direction, and it did not look like either of them. That is what a cold case actually is. It is not a puzzle with one missing piece. It is a room full of pieces, most of which belong to a different puzzle entirely, and no way to tell which is which until something changes.

 Something was about to change. Around that same time, a detective in Tacoma’s cold case unit was working these files. Her name was Lindsey Wade. Wade had grown up in that city. She was a child in 1986. She was the exact right age to remember what happened to the Parks and to remember what the adults around her started doing differently afterward and to remember thinking twice about the shortcut through the trees.

 She did not come to these cases as a stranger. She came to them as somebody who had been inside the fear they caused. She carried them for years, and the first genuinely new tool she got her hands on was not the one that solved it. It was called forensic phenotyping. Investigators worked with a private laboratory that could take crime scene DNA and use it to predict physical characteristics, not identity.

 Characteristics, likely ancestry, likely eye color, likely hair color, likely skin tone. From that, the lab could generate a predictive image of what the unknown man might have looked like. Tacoma released those images to the public, and I want to be precise about what that is and what it absolutely is not, because this technology gets misunderstood constantly.

 A phenotype composite is not a photograph. It is not a likeness. It cannot name anybody. It is a statistical estimate of a range of appearance rendered as a face so that the public has something to look at. It narrows a field. It generates calls. It puts a case back on television. It does not solve a case. It has never solved a case by itself. It did not name him.

 The thing that named him showed up in 2018, and it changed cold case investigation in the United States permanently. Here is how it works, and once you hear it, you will never be able to unhear how obvious it is. If a man has never been arrested, his DNA is not in any police database. That is the wall.

 That is the wall this case had been sitting behind since 2006. But, his second cousins are not in a police database, either. His great aunts are not. His distant relatives, scattered across the country, people he has never met and could not name, are not in a police database. Some of them are somewhere else entirely.

 They are on public genealogy websites because they got curious, because they wanted to know where their family came from, because somebody bought them a kit for Christmas. They spat into a tube. They mailed it off, and they uploaded their genetic profile to a public database that lets people find their relatives. They are not suspects.

 They have done absolutely nothing. They do not know he exists. They just happened to share a fraction of their DNA with him. So, investigators took the profile from Michella’s case, the one that CODIS had been silent about for 12 years, and instead of asking the police database, is this man in here? They asked a public genealogy database a different question entirely.

 Is anybody related to this man in here? And the answer came back yes. Not him, cousins, distant ones, people with no idea why they mattered. And from those cousins, genealogists began to build. This part is not glamorous, and it is not fast. It is not a computer spitting out a name. It is human beings doing archival research.

 Census records, birth certificates, marriage licenses, obituaries, newspaper archives, church records, building a family tree backward and outward from a handful of distant relatives. Hundreds of names, generation by generation, until the branches start to converge. And then you start cutting. Was he alive in 1986? Was he the right age? Was he male? And above all, could he have been in Tacoma, Washington in the spring of 1986? Branch after branch falls away.

 Hundreds become dozens. Dozens become a handful. And then it comes down to one. His name was Gary Charles Hartman. He had lived in the Tacoma area for essentially his entire adult life. He was a nurse. He had held a job that requires a license, a background check, and the trust of vulnerable people. He had no conviction that would ever have placed him into a criminal DNA database, which is the precise, mechanical, unromantic reason that CODIS had been silent for 12 years.

 Not a cover-up, not a lucky escape. He simply had never been convicted of anything that would put him in the system, and so the system had nothing to say about him. He was not a ghost. He was a neighbor. Now, a family tree is not proof. Genetic genealogy does not convict anybody. What it produces is a name and a hypothesis, and that hypothesis has to be tested against the actual crime scene DNA, which means detectives needed his DNA directly, and they could not simply walk up and ask for it.

 So, they did what has now become standard practice in this kind of case. They followed him, and they waited for him to leave something behind in public, something he had discarded, something he had thrown away and had no further interest in, in a place where anybody could pick it up. He did. They recovered it.

 They sent it to the laboratory, and they compared it to the profile taken from that ravine in 1986. It matched. 32 years, 3 months, and a handful of days after a 12-year-old girl chained her bicycle to a rack and set a paper bag down on a picnic table, the man who took her had a name. Gary Hartman was arrested in June of 2018. He was at that point a man in his 60s who had spent his entire adult life within a short drive of Puget Park.

 He had gone to work. He had come home. He had aged. He had done this for 3 decades while a city put up sketches, and a family went to graves, and detectives burned tens of thousands of hours looking for a man who was, the entire time, right there. He had never left. The case went into the court system, and this is where a lot of people watching are going to get angry, and I want to explain it properly instead of just letting you be angry.

Hartman ultimately pleaded guilty to the murder of Michella Welch. He was then sentenced, and the sentence he received was governed not by the law as it exists today, but by the law as it existed in 1986. That is not a loophole. That is a basic principle of American law. You are sentenced under the sentencing scheme that was in force at the time of the offense.

 A court in 2019 cannot reach back and apply 30 years of subsequent legal change to a crime committed before those changes existed. The sentencing framework of 1986 was, in practical terms, far less severe than what a similar crime would draw today. So, the sentence handed down was in the region of 26 years. Not for lack of will. Not because anybody in that courtroom thought it was enough, because the man committed the crime in 1986 and hid inside the passage of time, and the passage of time turned out to have a legal value. Every year he stayed

un-caught was a year that worked in his favor. That is the part of this story that does not resolve. There is no version of the sentencing where the arithmetic feels correct, but there is one thing the arrest did do, and it is not nothing. It named him publicly, permanently, in open court, with her family in the room, with the record of what he did written down and attached to him for the rest of history.

For 32 years, Michella Welch’s case had an empty space where a person should be, and a family had to live next to that empty space every single day. The space has a name in it now, and there was one more thing left to answer. Because remember, there were always two cases. The same year, using the same method, investigators went after the other one.

The DNA profile from the Jennifer Bastian case, the one that had also gone into CODIS and come back with nothing, was put through the same genealogical process. Cousins, records, a family tree built backward, branches cut away until one man was left standing. His name was Robert Dwayne Washburn.

 He was not in Tacoma anymore. He had been living in Illinois for years, a couple of thousand miles from Point Defiance Park. He was arrested in 2018, and he too pleaded guilty. Two cases, two men, two arrests, same year, same method. 30 years of investigation built on the belief that these two crimes were one man’s work ended with both of them solved within months of each other by proving conclusively that they never were.

 There is a strange cold symmetry to that. The theory that held this investigation back for three decades was finally destroyed by the same technology that solved both halves of it. The mistake and the answer arrived in the same envelope. So, what actually broke this case? It was not a confession. It was not a tip.

 It was not a witness who finally found their courage after 30 years. And it was not a detective staring at a corkboard until the pattern jumped out. It was three things, and none of them are dramatic. The first was a procedural de- cision made in a dark ravine in 1986 by people who had no idea why it would ever matter. They collected evidence.

 They preserved it correctly. They filed it. And then it sat in a room for 20 years, useless, taking up space, surviving budget cycles and moves and reorganizations because somebody kept deciding not to get rid of it. Every single thing in this story happened because that box did not get thrown away.

 The second was patience, not a virtue in this context, a tactic. A cold case unit is an institution built on the belief that a case is not closed just because it is not moving, and that the correct thing to do with an unanswerable question is to keep it carefully until the world produces the tool that answers it.

 Lindsey Wade carried these files for years without a break in them. That is the job. Most of it is nothing happening. And the third was a scientific idea that nobody in 1986 could have imagined, which is that a man’s own privacy is not the only privacy that matters. He never gave his DNA to anybody. He did not have to. His relatives did for entirely innocent reasons, and that was enough.

 That idea is genuinely powerful, and it is genuinely uncomfortable, and both of those things can be true at once. It has, since 2018, been used to identify suspects and to name unidentified bodies in hundreds of cases across the country. It has closed cases that everybody involved had privately given up on. It has given families answers after 50 years.

It has also raised real questions about consent and about what a person is agreeing to when they upload their genetic profile to a website to find out where their great-grandparents were born. Those questions are still being worked out in courts and in legislatures right now. Michella Welch’s case sits at the beginning of that story.

 It is one of the cases that proved the method worked. I want to end where we started because I think the beginning of this case is the part that keeps getting lost underneath everything that came after it. Strip away the DNA. Strip away the databases and the family trees and the 32 years and the courtroom and the sentence that does not feel like enough.

Go back to the morning. Three girls in a park. Somebody realizes there is no lunch and a 12-year-old gets on a bicycle. Nobody made her. Nobody asked her. She noticed a problem that belonged to all three of them and she decided it was hers because that was apparently who she had already become by the age of 12.

She rode 2 miles home through her own neighborhood. She let herself into an empty house. She made sandwiches for her sisters. She put them in a brown paper bag and then she rode 2 miles back into that park and she She her bicycle where she always chained it and she set that bag down on the picnic table where they would find it.

 And then she went to go get her sisters so they could eat. That is the last free decision Michella Welch ever made. Everything after it was made by somebody else. Everything about how her life ended was taken out of her hands by a man who did not know her and who would spend the next three decades acting like it never happened.

 But that decision was hers. Nobody else gets it. The bag was still on the table when the police arrived that afternoon. It was still there when the search moved into the trees. It was still there when the light went and the flashlights came out and the dog went down into the gulch. It sat on that table through the whole first night of a search that would take 32 years to end.

 Her sisters called into that ravine on the 26th of March, 1986, and the sound did not come back. It came back in 2018. It came back late, and it came back smaller than it should have, and it came back in a courtroom instead of a park, but it came back. Her name was Michella Welch. She was 12 years old. She was the oldest of three sisters, and she was the one who noticed, and she was the one who went, and she was the one who came back.

 If you want more of this, records, filings, sourced from the first line to the last, no invented dialogue and no artificial imagery, and no pretending to know what a child was thinking, then subscribe because that is genuinely the only thing keeping it alive. The next case is already open on the desk.