When Lisa Graaf took the stage at ISHI 36, she had a story to tell about a mandible that had waited 19 years for a name.
It was broad and heavy—chunky, as an NFI anthropologist would later describe it—when fishermen hauled it from the North Sea in 2003. Golden inlays glinted from several teeth, evidence of modern dentistry and someone who’d once been well off. But there was no name. No family. No story to tell investigators who the person had been or how they came to rest on the seafloor.
For nearly two decades, that mandible sat in the Netherlands Forensic Institute’s collection of unidentified remains. DNA was extracted. Profiles were generated and uploaded to the Dutch Missing Persons Database. But no matches emerged. The case went quiet, joining the thousands of others where the North Sea keeps its secrets.
Until 2020, when an author named Martin Hendriksma started asking questions about a hiding place between the walls of an old house.
A Story Between the Walls
As Graaf recounted, Martin Hendriksma wasn’t looking for a missing person when he first heard about the discovery in Heemstede. He was an author chasing a good story—and the homeowners who’d found a concealed space behind their walls during renovations had exactly that. Inside the hiding place: a mattress, two broken bottles, and a torn newspaper dated November 28, 1941.
Hendriksma’s research through municipal archives revealed the home had belonged to Ernst Octavius Moltzer during World War II. The Moltzer family name carried weight in the Netherlands—they’d acquired the Lucas Bols distillery in 1868, claimed at the time to be the largest in the world. Ernst himself had been born into privilege in 1910, joined the board of directors at 20, and was a decorated athlete: national rowing champion in 1935, Olympic sailing team member in 1936.
“He had all things going for him, it would seem,” Graaf said during her presentation. But in 1941, Ernst made a choice that would cost him everything. He left the family company rather than profit from business dealings with German occupiers, and the trail went cold.
Through his research, Hendriksma discovered the truth: Ernst had become an *Engelandvaarder*—literally, “sailor to England.” These were the men and women who risked everything to cross the North Sea in tiny boats, often with four or five people crammed aboard vessels never meant for open water, to join the Dutch resistance operating from England. Around 1,700 made it. Many more didn’t.
On November 14, 1941, Ernst and two close friends set sail on a modified boat called the *Oso*. They were never heard from again.
The Question That Changes Everything
With this information, Hendriksma had his book subject. But he had two questions for Dutch police: What happens to remains lost at sea? And if Ernst’s remains were ever found, could they be identified?
Through her presentation, Graaf walked the ISHI audience through what Hendriksma learned. The answers led him to two specialized units: the National Expertise Center for Missing Persons and the Maritime Police’s Bureau for Missing Persons on the North Sea. The system they described was methodical. Fishermen who catch human remains in their nets hand them over to police. Carbon dating determines if the remains are from someone who lived after 1920—the cutoff established because anyone who went missing before then likely has no living relatives close enough genetically to enable identification with standard testing.
If carbon dating places remains after 1920, DNA profiles are generated using 23 autosomal STR loci and stored in the Dutch Missing Persons Database, where they’re compared against profiles from family members of missing persons.
When asked about the significance of the 1920 date, Graaf explained, “This date was chosen based on the data where we don’t expect any close relatives to be living today. If you go further back, then perhaps no one alive even remembers that person going missing, or they are too distantly related to them to be able to search effectively.”
Hendriksma was advised by police to find relatives—ideally first-degree, or anyone in the paternal line. Through exhaustive genealogical searches across archives in the Netherlands and beyond, he found a distant male relative: a second cousin once removed, paternally related to Ernst. The man agreed to provide a DNA sample.
For Graaf, the genetic distance was daunting. “I have to be honest with you, given the large genetic distance, my only hope was that the Y-STR would yield a possible match.”
The Power of Kinship Databases
When the reference sample reached Graaf’s lab at the NFI, it was processed using the Promega PowerPlex ® Fusion 6C and PowerPlex ® Y23 chemistries. The profiles were submitted to the Dutch Missing Persons Database, which contains DNA profiles from unidentified human remains, missing persons, and relatives of missing persons.
The database uses two comparison approaches. CODIS software performs autosomal comparisons between unidentified remains and missing persons profiles, looking for complete matches. But kinship analysis—connecting remains to family pedigrees—requires more sophisticated tools. For this, the NFI uses Bonaparte, software developed in-house in 2007 with the University of Nijmegen.
“It stores all pedigrees that we create along with the profiles, and performs all calculations that we could possibly need. No matter how complex the pedigree, with a click of the button, we have the results,” Graaf explained.
The autosomal profile from Ernst’s second cousin once removed didn’t generate a match. But the Y-STR profile did: a 17/17 match to an old profile labeled only as “mandible.” The same mandible fished from the North Sea in 2003.
Because the NFI retains all extracts from unidentified remains, Graaf was able to upgrade the profile to PowerPlex Y23 within a week. The match held: all loci matched between Ernst’s distant relative and the unknown mandible.
Building the Case
The Y-STR match was promising but not conclusive. Police and Hendriksma continued searching for closer relatives. Hendriksma eventually located a nephew (Ernst’s sister’s son) and a granddaughter—both living outside mainland Europe, one in the United States.
When contacted about a possible long-lost relative, both agreed immediately to provide DNA samples. The nephew would establish a maternal line through mitochondrial DNA, while the granddaughter would contribute second-degree autosomal data.
Meanwhile, Graaf consulted with an NFI anthropologist about what the mandible itself could reveal. “He immediately saw that there were multiple golden inlays present,” she recounted. “There are also some techniques that he saw that were used for restorations. And these were signs of modern dentistry. So it must point to someone being well off, but also someone who lived quite, quite recently. He further noticed that the mandible was quite chunky. The picture doesn’t do it justice, really. It’s a very broad, large mandible. So it must possibly belong to someone who was on the taller or broader side and both fit our missing man, Ernst Moltzer.”
When the new DNA results came back, the case came together. The entire mitochondrial DNA control region between Ernst’s nephew and the mandible matched. Combining the nephew’s and granddaughter’s profiles yielded an autosomal likelihood ratio of approximately 460:1.
In the Netherlands, different DNA profile types are combined and reported on a verbal scale rather than numerical. The verbal scale ranges from “indistinguishable” (1-2) through increasingly strong support levels up to “extremely more likely” (over 100,000).
“In this case, we reported that the DNA profiles that we obtained were extremely more likely if the other is Ernst Moltzer, given the shown pedigree, than if it had belonged to anyone else unrelated to these three individuals,” Graaf said.
For autosomal weight of evidence, Bonaparte handled the calculations. Y-STR rarity was assessed through database searches using Familias software, while mitochondrial DNA prevalence was checked using the EMPOP database.
With this convergence of evidence—Y-STR match, mitochondrial match, autosomal likelihood ratio, anthropological observations—Dutch police formally identified the mandible as belonging to Ernst Moltzer.
Bringing Him Home
With the formal identification complete, Ernst’s family chose to bury him at one of the Netherlands’ World War II cemeteries. After more than 80 years missing, he would finally have a marked grave. The identification also provided the concluding chapter to Hendriksma’s book, which had started this entire investigation four years earlier.
“It served as a final fitting chapter,” Graaf said. “Providing closure to his family, who had been in the dark for over 80 years.”
The case demonstrates what’s possible when multiple disciplines converge. Journalists and authors have time, resources, and archive access that can unlock cases years or decades old. The public—like the North Sea fishermen who now routinely hand over remains they catch—can be crucial partners when properly engaged. Dedicated police units maintain detailed records that create connections across cases. And DNA kinship analysis, especially when combining autosomal, Y-STR, and mitochondrial profiles in a dedicated missing persons database, becomes extraordinarily powerful.
“Kinship analysis is a very powerful tool, especially if you’re able to combine autosomal, Y-STR and mitochondrial DNA in cases like this,” Graaf emphasized. “This power is further enhanced if you can apply it to a dedicated DNA database for missing persons and unidentified human remains.”
Even forensic anthropology played a role—the golden inlays and dental work weren’t strong evidence alone, but they pointed investigators toward the right time period and socioeconomic background.
Lessons for Labs
Through her presentation, Graaf offered practical takeaways from Ernst Moltzer’s case for forensic DNA analysts working missing persons investigations:
- Preserve everything. The NFI’s policy of retaining all extracts from unidentified remains allowed Graaf to upgrade the Y-STR profile years after initial testing. Without that extract, the first Y-STR match might never have been strengthened.
- Think kinship from the start. Even distant relatives can yield investigative leads, especially through Y-STR profiles. Don’t dismiss potential matches because of genetic distance—they can be the thread that unravels the entire case.
- Combine evidence types strategically. The power here came from layering Y-STR (paternal line), mtDNA (maternal line), and autosomal profiles (multiple relationships) with anthropological observations. No single piece was conclusive alone.
- Software matters. Purpose-built tools like Bonaparte that can handle complex pedigrees and perform multiple calculation types streamline what would otherwise be prohibitively time-consuming analysis. Know what your software can and can’t do.
- Engage beyond the lab. As this case demonstrated, success requires more than technical expertise. Hendriksma’s curiosity, police dedication, cooperative fishermen, and family members willing to provide samples from across continents all proved essential. Missing persons work requires building and maintaining those networks.
As February 3 marks National Missing Person’s Day, Ernst Moltzer’s story reminds us why this work matters. Every unidentified remain represents someone’s son or daughter, brother or sister, parent or friend. Someone left behind an absence that never fully closed.
For 80 years, Ernst’s family lived with unanswered questions. The hiding place in the walls of his house spoke to resistance and sacrifice, but they never knew what became of him after that November night in 1941. Now they do. Now they can visit a grave. Now there’s an answer.
Lisa Graaf’s presentation “How a Multidisciplinary Missing Persons Investigation Brought a Dutch War Hero Home” was delivered at ISHI 36. Graaf is a forensic DNA kinship analyst at the Netherlands Forensic Institute, specializing in missing persons, unidentified human remains, familial searching, and immigration kinship cases.