Monday, Dec. 27, 1999

Marmes rock shelter key site

By Mike Lee Herald staff writer

Forty feet deep in the lower Palouse River rests one of the nation's most significant archaeological sites.

The Marmes rock shelter, a natural cave covered in water after Lower Monumental Dam was completed in 1969, yielded 5,000 cubic yards of dirt along with hundreds of artifacts, animal bones, weapons and even a tiny sewing needle.

At the time, Marmes also produced some of the oldest well-documented human remains found in North America, roughly 9,000 to 11,000 years old, according to Washington State University scientists.

By mid-2001 - more than 30 years after the site was flooded - the Confederated Tribes of the Colville Reservation plan to finish the site exploration archaeology report that was never completed.

"This is one of those projects that for several months a year, you sleep with it, you live with it," said Brent Hicks, archaeologist and project leader for the Colvilles. "We really want to find out more about how these people lived and when."

Also different are the reporting standards followed by modern archaeologists, especially compared with the "rush job" done in 1969 when the project was running out of money and lead scientists were turning to other ventures.

"It doesn't come anywhere near what a true final report would be for most sites, certainly by today's standards," said Hicks. "There is so much information that never came out."

For example, archaeological teams at Marmes in the mid-60s screened the dirt for artifacts using a quarter-inch screen. Today, they would likely use eighth-inch screens - and probably find more fish bones and other tiny clues about the Marmes culture.

Should the Lower Snake River dams ever be torn down to recover salmon, archaeologists could have another look at Marmes - just north of Lyons Ferry State Park at the impounded mouth of the Palouse River - and several other ancient sites along the Snake. Hicks said the greatest advantage of that would be to look at the sedimentary record of the site, which researchers currently can "observe" only through field notes.

And those field notes aren't that good.

"One of the great tragedies of this site ... is the lack of record-keeping," Hicks said. "That robs us of more information than having the missing items that were not found."

"Each week, they found more precious evidence of early life in the Americas," said Keith Petersen and Mary Reed in their history of the Lower Snake River, Controversy, Conflict and Compromise, prepared for the Army Corps of Engineers.

"Overlying the ancient remains rested layer upon layer of evidence of human use of the rock shelter, extending into the 20th century when white residents occasionally explored the cave."

It was, as the authors said, "10,000 years of human history, condensed into one small area, waiting to be uncovered."

Among other artifacts in and around the 25-foot-deep cave, archaeologists found the jawbone of an arctic fox and forest animals such as the pine marten, indicating a far different landscape in the ancient past. What they found is being kept at WSU, and the human remains have not been reburied.

The story of the discovery and its untimely end, as culled from Petersen's and Reed's riveting account, goes like this:

In 1953, residents along the Lower Snake River showed prominent WSU archaeologist Richard Daugherty a rock shelter once inhabited by ancient people on property owned by Roland Marmes. A decade later, he started excavating the site as the Corps was finishing Lower Monumental Dam.

They spent 18-hour days digging and sifting with hand tools. A tent city and a laboratory was built at Marmes. More crews were added.

"The odds of finding such a complete package of evidence within one site again are so great it is almost impossible," said the renowned H. Marie Wormington at the time.

But the archaeologists weren't working fast enough.

So Daugherty convinced his friend and influential U.S. Sen. Warren Magnuson of Washington to scratch out $1.5 million from the federal budget for a levee to keep the site dry when the river rose behind the dam. The proposal died in the House, but Magnuson went to President Lyndon Johnson.

Over the objections of the Corps, in 1968 the president ordered a levee built at Marmes. Engineers set to work expecting problems trying to keep the site dry and nervously eyeing the coming fish passage season, when the new dam had to be working.

The Corps planned to pump out water that seeped into the shelter - but the engineers underestimated the incoming flow.

In February 1969, water spilled into Marmes at 45,000 gallons a minute. The pumps couldn't keep up.

"There wasn't a damn thing we could do," said Harry Drake, the Corps chief engineer, as quoted by Petersen and Reed.

Archaeologists scampered in and covered the site with plastic and fill dirt to keep the floodwaters from washing it all away.

And then the water flowed over it.

The leaky levee still stands, and the site is visible from an overlook that is about a quarter-mile hike above Lyons Ferry State Park. The overlook includes basic interpretive signs about the area's archaeology. It's about a three-mile hike to the actual site, though the road is blocked and there are no signs. The state doesn't encourage visitors to go there because of its archaeological importance.

That's still a matter of speculation, but evidence suggests that as early as about 12,000 years ago the area would have been inhabitable because of the flooding and pooling caused by the melt of the last Ice Age, Hicks said.

In general terms, however, it's clear Marmes was a dwelling place well before 7,000 years ago, when the Mount Mazama eruption in Southern Oregon covered the region in ash thick enough to make lowlands like the cave unlivable. When people returned, their stays were intermittent.

The Colvilles aren't doing new studies about the human remains found at Marmes. Hicks described them as "fairly extensive" but downplayed their significance in recreating Palouse history, much of which can be interpreted from tools, garbage and animal bones.

Hicks expects that ancient Palouse residents from Marmes would have met and mixed with the people who buried the cache of stone Clovis-style tools found in 1987 near Wenatchee.

And Kennewick Man, should he really be 9,200 years old, probably would at least have known about the rock shelter roughly 40 miles east of where he was discovered in Columbia Park three years ago.

"The odds that they had not encountered each other in their lifetimes would be very, very low," Hicks said. "They may have been related."