×

Architectural conservator William Flynt uses timber patterns to determine when structure was built

  • William Flynt, a historian at Historic Deerfield, holds a drill bit he used to remove core samples from timbers in the Shepherd Barn in Northampton. He uses the practice of studying tree rings to determine the age of homes known as dendrochronology. FOR THE RECORDER/ANDY CASTILLO

  • William Flynt, a historian at Historic Deerfield, measures tree rings as part of a process to determine the date of a historic house at the Memorial Libraries in Deerfield. FOR THE RECORDER/ANDY CASTILLO

  • A core sample William Flynt took from a house shows the age rings of a piece of timber. He has found this process to be a more accurate way of dating the true age of a home since historically construction techniques were sometimes used at atypical times. FOR THE RECORDER/ANDY CASTILLO

  • William Flynt drills out core samples from a historic home in Sheffield in 2004. Flynt has become an expert in determining when historical houses were built with dendrochronology revolutionizing his work. SUBMITTED PHOTO/Will Garrison



For the Recorder
Friday, May 18, 2018

DEERFIELD — On the second floor of Historic Deerfield’s Memorial Libraries, in a small room filled with massive tomes and the aroma of decaying pages, architectural conservator William Flynt studies timber patterns to determine the age of historic New England houses.

One recent day, cross section core samples drilled from the timbers of a historic house on Cape Cod, Flynt’s most recent project, rest on a table. At a desk in the middle of the room, looking through a microscope, Flynt, 64, of Dummerston, Vt., painstakingly counts the age rings on one of the cross sections to figure out when the tree was cut down. Knowing when the trees were felled, in turn, will help him to estimate when the house was built.

“There are drawers and drawers of these things. I try to keep it by states,” Flynt says, pausing from his work, and gesturing to a row of cabinets filled with thousands of samples. They were taken over a period of nearly two decades from timbers in 200 houses.

The practice of studying tree rings to determine calendar dates is called dendrochronology. It’s typically pursued by scientists looking into what causes forest fires, or climate change. Applying dendrochronology to dating historic structures isn’t as common, according to Flynt.

More traditional means of dating houses usually rely on the identification of construction features typical to certain time periods.

However, Flynt says, that can be unreliable because construction techniques were sometimes used at atypical times, and older houses were often built in sections over a period of many years. Thus, for a more accurate historical assessment, Flynt counts tree rings.

Flynt, who has a master’s degree in historic preservation from the University of Vermont, has worked at Historic Deerfield for 40 years.

His grandparents, Henry and Helen Flynt, started the museum in 1952 as a way to restore historic houses in Old Deerfield. Growing up, Flynt went to Deerfield Academy and later attended Williams College. Now, he oversees research and preservation of the Deerfield museum’s historic homes, gives public presentations, and works as a consultant for organizations throughout New England.

One of those organizations is Historic Northampton, which hired Flynt through a Community Preservation Act grant a few years ago to figure out when its Shepherd Barn was built.

Solving a mystery

The barn appeared on Northampton’s official map in 1853, but historians couldn’t determine when exactly it was built because the barn has construction features typically used during a few different centuries.

“There were a lot of dates that were floating around, but no one really knew,” said Laurie Sanders, co-executive director of Historic Northampton.

On another day, inside the barn, dark except for pinpricks of sunlight streaming through cracks in the old timbers, Flynt points out distinguishing construction features like English tie joints — a common way of joining wood used during the 17th and 18th centuries — and a beam that runs from one end of the roof to the other called a purlin — which splits the roof’s load — first used in the early 19th century.

“You see bits and pieces that could be 18th century or could be 19th century. So, how old is the barn?” he asks.

To find out, Flynt drills out about 10 samples from timbers throughout the structure. The round samples are sanded flat to reveal the wood’s growth rings, from the date of the tree’s seeding to when it was felled.

Each tree ring represents a single year’s growth. The amount of space between each ring is influenced by the climate of that particular year, Flynt says. During a wet year, trees grows more, and rings are therefore wider apart. Conversely, rings are closer together following a dry year because the tree hasn’t grown as much. Ring patterns of different trees, which grew in the same region, can be matched together, he says.

“You’re looking at the pattern — dry year, wet year, dry year, wet year, dry, dry, dry,” Flynt says. “You need 50 to 60 rings at a minimum to get things to correlate with the reliability that you need. I don’t care when it started growing, I care when it was cut down.”

Each ring pattern is converted into a computer file by Flynt, who does so by looking at each sample through the microscope, which is connected to a computer, and manually pushing a button at the beginning of every ring to measure it to the nearest micron. The ring measurements are converted to a series of numbers by Cofecha, a free dendrochronology software program developed by the Tree Ring Laboratory at Cornell University in New York.

Flynt then matches the core samples based on their digitized measurements in order to understand how they relate to each other. The data is organized in a single chart of numbers for each house.

This information, in turn, is matched against an expansive database of measurements collected from trees of the same species and in the same region, which have been dated. He’s hoping to find at least one that matches the sample he’s evaluating.

Over the years, Flynt has created an extensive chronology of tree ring patterns that spans decades.

By using this method, he matched the samples taken from the Shepherd Barn to the bigger database and nailed down an exact date for the structure. “Most of the timbers were cut down in the winter of 1802, which was standard. They cut in the winter when the leaves were off, and the tree didn’t have a lot of sap, and they would do their framing and joinery in the winter and spring,” he says. “Between 1800 and 1802 most of the stuff was cut down. So the earliest the framing could go up would be the spring of 1803.”

Flynt notes that dendrochronology isn’t the only kind of science that can be used to date historic houses. During one dendrochronology project in Martha’s Vineyard, which he was hired to date by a historical society, none of the wood samples could be matched into a larger database.

Taking a different approach, Flynt sent a few of the samples to the University of Arizona where scientists determined its age through carbon dating, a method of determining the age of organic matter by measuring the amount of carbon it contains. Flynt was then able to determine that the house was built a few decades after the historic society previously estimated.

Serendipitous meeting

Dendrochronology stems from American astronomer Andrew Douglass’ research during the early 1920s into the way sunspots impact climate. He began studying tree rings in order to understand climate change and its impact in Arizona. Originally, scientists matched tree rings visually, and it wasn’t until more recently that dendrochronology was applied to houses.

Flynt fell into the practice unintentionally. In 2002, while trying to figure out the date of Historic Deerfield’s Wells-Thorn House on Old Main Street, which has structural elements from different time periods, a friend referred him to Paul Krusic, a scientist forester trained in dendrochronology. At the time, Krusic was living in New Hampshire and coincidentally had a sister attending Deerfield Academy.

During a visit to Deerfield Academy, Krusic agreed to help Flynt learn the procedure as a skill to date historic homes, and took him on as a sort of apprentice. Together, they took samples from the Wells-Thorn House and compared them to the Stebbins House next door, which historians knew was built in 1773 based on dated journal records that referenced the house’s construction and purchase agreements.

Based on how the core samples lined up, Flynt estimated the Wells-Thorn House was built in 1747.

Since that first project, Flynt has become an expert in determining when historical houses were built; dendrochronology has revolutionized his work. He has completed projects for private clients, restoration contractors, historical societies, and places like Historic New England, the National Park Service, and Old Sturbridge Village. In the Pioneer Valley, he’s worked extensively documenting homes, especially in Old Deerfield.

Last summer, Flynt cut his hours back at Historic Deerfield to three days a week in preparation for retirement, during which he intends to continue his work in dendrochronology.