The sandbar on Mill Creek the author believes to be the site of the Three Knolls massacre. The Indians would have been camped on the sandbar in the foreground; the settlers would have been firing from the bank to the right. In this picture dating to the 1990s, California Department of Forestry archaeologist Richard Jenkins works a metal detector in an unsuccessful search to find confirmation of the theory.
There had been little conflict with the Indians in Butte County in 1864, as might have been expected in light of the expulsion of the native people that followed the killing of the Lewis children the previous year.
But in 1865, the Mill Creek War returned to south of Yahi country, with a bloody raid by Indians, followed by an attack of retribution that has become known as the Three Knolls Massacre.
It’s the best documented massacre of the war, with four detailed accounts published — one just days after the attack — and a brief newspaper item.
And yet, there is much about the incident that remains unsettled, including where it happened, how many people were killed, and the tribe of those who were slain.
As the story is told now, in revenge for an raid in the Concow Valley, southeast of present-day Paradise, a group of whites tracked the raiders to a village on Mill Creek where the entire Yahi nation was gathered. Some 30 of the Indians were killed, with only five or six escaping.
The implication was that the survivors were Ishi’s companions at the end of the Yahi’s time. Theodore Kroeber wrote that Ishi’s father was among killed in the attack, even though she also said Ishi refused to talk about the incident.
But it’s a story that breaks down almost immediately when you look back to the primary sources, and think about what they actually wrote.
To begin with, the Concow Valley is dozens of miles from Yahi country, much farther than the severely pressed Yahi could have mounted an offensive.
Patching together the story from the various accounts, we can determine the trouble began on Monday, Aug. 7, 1865. At about 6 p.m. a party of three to five Indians appeared at the Robert Workman house in the Concow Valley. Mrs. Workman and her sister, Miss Mary Rosanna Smith, fled into the house, but the Indians followed, forcing their way in.
Mrs. Workman ran out the back, apparently to call for the aid of the hired hand, John Banks. A pursuing Indian caught up to her as she left the house, and knocked her down with a blow from a rifle butt. The Indian aimed the rifle at her, but turned the muzzle towards Banks as he came running up, and shot him dead. The Indian then picked up a large rock, crashed it down on Mrs. Workman’s chest, and left her for dead.
The raiders then apparently turned their attentions to Miss Smith, who was “ravished.” Her throat was cut and her body was mutilated. Banks also had his throat cut, but Mrs. Workman escaped that fate, escaping to a place of safety while the Indians were otherwise distracted.
The house was ransacked, and the raiders making off with gold coin valued at about $2,500, a silk dress, thread and other items.
Mrs. Workman made her way about 200 yards to the home of G.G. Marquis, and was taken to the Porter Quartz Mill, where the residents of the valley were gathering in fear.
The Butte Union Record reports there were only four men in the valley at the time of the raid, the rest being at a political convention in Oroville.
The Indians don’t appear to have felt much threatened, as they lingered in the area until Wednesday morning, Aug. 9, when they shot and wounded Joseph Miller near Frenchtown, a now-gone town that then lay in the eastern end of the valley. Two other houses were reportedly robbed, and one was burned down.
But by Wednesday afternoon, the white settlers had a posse of six or eight men together, and started in pursuit. The trail led to the northwest, toward Chico. By Thursday night, the group was camped at Butte Mills, very near the current Butte Creek Country Club.
As they continued north the following day, they ran into Sim Moak, who decided to join the raiders after hearing their tale. Moak rode ahead toward Solomon Gore’s house on Rock Creek, near the foot of the Cohasset Ridge, intending to raise a party and block the Indian crossing of that stream.
He was too late; the raiders had come and gone.
The Concow men arrived at Gore’s about noon on Friday, Aug. 11, and were joined by a number of young men from the area, including Gore’s son-in-law, Robert Anderson. The reinforced group moved on to Hi Good’s place on Deer Creek, and camped for the night.
Good killed a sheep for the raiders to eat, and John Bidwell sent up 40 pounds of bacon and 40 pounds of hardtack for the campaign ahead.
On Saturday morning, Aug. 12, the party headed into the hills, crossing Deer Creek Flats — a bluff overlooking that stream — and descending into Deer Creek Canyon. They moved upstream a few miles and camped for the night near the foot of Iron Mountain.
It should be noted that at time, Anderson’s account departs from the version of Sim Moak and Daniel Klauberg, the other two authors of lengthy accounts of the campaign.
Anderson has the raiders camping on Deer Creek Flat and electing him captain at that time. But Klauberg had Good taking command the night before. Anderson is most certainly mistaken on these points.
Anderson also says the raiders crossed Deer Creek Canyon the next morning near the mouth of Sulfur Creek, climbed up onto Digger Pine Flat, and headed north across the “two Dry Creeks.” This would seem to refer to Dry Creek and Little Dry Creek. Crossing those two creeks would require a course a mile or two to the west of the path Moak describes. Klauberg agrees with Moak on a starting point further east, but says little about the path between the creeks.
Moak’s account seems the most reasonable at this time, as Anderson’s route would lead to an area the Indians had abandoned years before. His confusion is understandable, given the number of times he headed into the hills to battle the Indians, and the 44 years between this particular attack and the publication of his memoirs.
Thus, at dawn on Sunday, Aug. 13, we can assume the raiders rose and began the climb out of Deer Creek canyon. They then turned down into Dry Creek, the next canyon north.
And with that, the first problem with the of the story of the Three Knolls emerges.
Everyone who has attempted to map the Three Knolls puts them upstream from Black Rock, the primary landmark on Mill Creek. The path that Moak, Anderson and Klauberg all describe leads to an area well downstream from the Black Rock.
Some have seized on a comment Moak attributed to Anderson, that he was sure the Indians were at Black Rock, to label one or another of the sites around that monolith as the Three Knolls. But that doesn’t make sense. If the village was there, why would
Anderson refer to a trio of landmarks that are so obscure that he and Good disagreed on which side of the creek they were on, instead of identifying the village by the obvious and dominating reference point Black Rock provides?
Had the raiders been heading to a village at Black Rock or upstream from there, they would not have climbed down into Dry Creek canyon. The gulch is about a thousand feet deep, not the kind of terrain you cross if you don’t need to.
And to get to Black Rock, it isn’t necessary or desirable to cross that canyon. A better and more direct route would turn east at the ridge top between Deer and Dry creeks, and follow the high ground as it climbs relatively gently to the bluffs above Black Rock.
Moak not only has the raiders crossing Dry Creek Canyon, but he says they then headed down and across Boat Gunwale Creek Canyon, which is about 800 feet deep in that vicinity. Again, it’s not a route you’d take to get to Black Rock or above.
And at the point where they headed down toward Boat Gunwale Creek, it’s not likely that Anderson and Good would have made a false turn. That ridge line carries the Lassen Trail, an immigrant road used heavily in 1849 and 1850. The two leaders of the group were familiar with the route, and would have known that following the road eastward would have carried them to the vicinity of Black Rock much more quickly.
It appears instead that the whites were following the Indians along the obvious path between Deer and Mill Creeks. There’s a clear notch in the side of the ridge carrying the Lassen Trail, and a low saddle with a convenient spring, which is an obvious place to cross the ridge. The Indians probably went through there, and Good and Anderson led their party through as well.
Today, the only north-south trail in the Ishi Wilderness roughly follows the course the raiders probably took.
At the Lassen Trail, one man — Joseph Dikeman — gave out and turned toward back the valley. The 16 who remained pressed on, crossing Boat Gunwale Creek Canyon and coming to the ridge overlooking Mill Creek. They were about a mile from the creek at this time, several miles downstream from Black Rock.
As they moved down an open ridge toward the creek, Anderson noted an Indian picket across the stream, at a lush area low on a bluff. The description matches an area on Long Point where a series of springs would provide a nice little oasis even today, if range cattle hadn’t torn up the area badly, leaving it a muck hole.
The raiders turned back, wound behind a high point in the ridge and followed the cover of a gully down to Mill Creek. The route gets a little vague here, as there are at least three watercourses that fit the general description. The paths all lead to the south side of the creek between Long Point and Table Mountain, the area I believe the Yahi called Tuliyani. (See Chapter 4.)
There’s a point about the terrain here worth noting. There’s not much flat ground on the north side of the creek in this area. Instead the bluffs rise up steeply for a few hundred feet, and then flatten out, climbing more gradually to the next divide. That higher table is divided by several creeklets into three ridges. Viewed from many points on the surrounding higher ground, the ends of these ridges look like three knolls rising out of the trees in the creek bottom.
View looking upstream in Mill Creek Canyon from the saddle between Table Mountain and Black Oak Grove shows what the author believes to be the Three Knolls in the foreground, on the north side of the creek.
I’m not certain this is what Anderson meant. The reference is so strange, as Mill Creek Canyon is so extreme and so dramatic that the idea a knoll could be a landmark is tough to believe. These three ridges are at least significant enough to serve as landmarks, and they really do provide one of the better ways to identify the area I believe is Tuliyani.
The reference is clearer than any other trio of humps or bumps in Yahi country.
The ridges/knolls also match the description of Klauberg. He doesn’t mention knolls, but says they camped on a ridge opposite the camp, and moved down both sides of an adjacent ridge to launch the attack.
And down along Mill Creek in this area is a locations that match the descriptions of the massacre site exactly. There’s a lens-shaped sandbar beneath a high bank, with a natural ford at the downstream end of the flat. It is just as Anderson, Moak and Klauberg described the site.
At any rate, the raiders came to the creek downstream from the site, crossed over and rested in some reeds while Anderson and Good scouted around. They saw some Indian women headed upstream, and returned to their group. It was about dark when the raiders followed the women upstream, moving uphill away from the creek to avoid detection.
They stopped on a slope, and Anderson and Good again went scouting to find the Indian camp. They wading up the creek to the point where an Indian’s dog challenged them. Anderson describes Indians sitting up from bedrolls around campfires on the sandbar, then settling back down to sleep.
And here’s the second problem of this story. The Three Knolls Massacre has become the account of an attack on an established village. It is a mistake I made for quite a few years.
But the slaughter, the accounts agree, happened on a gravel and sandbar, a feature born of submersion under high water. The Yahi weren’t dumb enough to build a village in a flood plain. They didn’t have the time or resources to build structures that would just be washed away with the next predictable flood.
This was not an attack on an established village, but rather on the temporary camp of a group of Indians on the move.
But still, Anderson quotes Sandy Young as saying he’d passed through the massacre scene later in 1865, and reported it looked like a cyclone had hit. It’s hard to imagine that description applying to an attack on an overnight camp where there would have been little more than bedrolls and campfires.
Young was John Bidwell’s representative in Big Valley, now submerged under Lake Almanor. He led an expedition down Mill Creek in 1865 in response to raids by the Mill Creeks on the Mountain Maidu who were partially his responsibility, as many of them were employed building Bidwell’s Humboldt Road from Chico to Susanville.
At any rate, the Three Knolls raiders moved a few hundred yards back to pass the night.
As the eastern sky began to lighten on the morning of Monday, Aug. 14, 1865, the white raiders began to move toward their sleeping targets. Sim Moak and Frank Curtis were sent off upstream to block the Indians escape that direction. William Merithew (his name also appears as Merathew and Matthews) moved to the creek downstream from the village.
The remainder spread into a loose half-circle and moved down the ridge toward the sleeping Indians. Good and Hardy Thomasson were in the center of the line on the west side of the ridge; Anderson and Henry Curtis held the middle of those on the eastern slope. Klauberg, judging from his observations, was with Good.
They were a few dozen yards from the village when an early-rising Indian came climbing up the slope toward them. He was Billy Sims, a valley Indian gone bad, Anderson reported. It’s easy to imagine him stopping, stunned, as he saw the whites creeping down the hill at him in the pale light before dawn.
He recognized Anderson and aimed his rifle at him. Henry Curtis whistled a warning and Anderson ducked behind a tree. Sims bolted, not to warn the village, but to escape across the ford to the other side of the creek. He was fighting across the current when Good shot him through the breast. Sims struggled on to the other bank, staggered along the shore for a few steps, and collapsed.
The white raiders then ran to the bank above the sandbar, and fired as fast as they could at the Indians who were scrambling for safety just below them. Many ran for the ford; others headed for the cottonwoods and alders along the creekbank and headed upstream.
The slaughter was limited by one or more Indian riflemen who appeared on a bluff on the opposite side of the creek, and began firing at the white attackers. The settlers scrambled for cover, and most of the Indians escaped in the interim. Anderson said he put a shot into the rocks between one of the Indian’s legs, and when he staggered back under cover, the village belonged to the whites.
The Indian snipers reinforce the idea that the targets of the raid were not Yahi. In addition to bullets, the white raiders were serenaded with eloquent curses in English. When Ishi came out of the wild, he knew no English words, and one could presume that would be the case for the entire tribe.
Good and Thomasson splashed across the ford to secure the opposite bank, cutting off an Indian woman’s escape in the process. She was taken prisoner.
Good scalped Sims and recrossed the creek to take the same trophies from the four or five Indians laying dead there. That was his right as the captain of the company. The men from Concow, wrought up by the rape and mutilation of Miss Smith, were mutilating the bodies. One Indian’s neck was cut to the bone, with the head being twisted around afterward in an unnatural way.
But here’s another problem: The number of dead didn’t approach the 30 cited in the common version of the Three Knolls attack.
Good put the number at nine in an account published in the Red Bluff Independent three days after the account. Klauberg said there were five bodies on the field and six or seven who escaped but “will surely die.” Moak agrees with the body count of five, and further identifies them as a man, two women and two children.
The toll of 30 appears to stem from a comment W.J. Seagraves made to Thomas Waterman when he was collecting information for his 1918 paper “The Yana Indians.”
Seagraves said he’d visited the scene years after and there were about 30 skeletons still scattered on the surface.
Waterman singled out Seagraves as one of his best informants, which leaves us another mystery to resolve. How could a trusted informant count far more dead than those who had participated in the attack?
Back at the Three Knolls, the attackers rifled through the Indians’ effects. They found a native infant alive, who Good promptly claimed. The boy, Anderson reported, had six toes on his feet.
One wonders if he may have been the son of a notorious Indian raider, a man the whites called Big Foot. He had large feet, with six toes on each. His distinctive tracks were often found at the scene of Indian “depredations,” but after the Three Knolls Massacre, were seen no more. He may have been shot while crossing the stream and floated away. Or he may have succumbed to wounds which allowed him to flee the carnage before dying.
The raiders gathered up such plunder as they could carry, and burned the rest. The captive woman was forced to carry the infant, and the group headed off toward the Sacramento Valley.
Anderson reported they “climbed the long hill and over its crest into Twenty-Mile Hollow.” This is another indication the three knolls weren’t upstream from Black Rock, as Twenty-Mile Hollow is about a dozen miles downhill from there. There’s no way you could go over the crest of a long hill into the hollow from Black Rock or upstream from there. You’d cross the crest several miles walk away from the hollow.
At the top of the hollow, the woman refused to go further, and was killed. The raiders continued down into the valley, with the Concow men moving on toward Chico. John Bidwell sent a wagon to carry them to town, and bought them dinner at the fanciest spot in town — Weatherbee’s hotel. He then provided them a ride back to Concow, where they arrived in glory, we can presume.
Anderson and Moak say the Indians were “chastised” to the point they presented only minor troubles to the settlers from then on. This ignores the murder of Mrs. Dersch the following year, but that was up in Shasta County, a long way from Chico in those days.
But it may not have just been the Three Knolls attack that brought peace to the vicinity. The discrepancies about the location, the death toll, whether a village or a temporary overnight camp was struck, and whether the victims were Yahi or someone else, beg an answer.
After years of thinking about it, I’ve come to the conclusion that accounts of two separate but similar attacks were blended together by Waterman. That error was widely distributed by Kroeber, and has become the “common knowledge” today.
It’s my belief that there were two attacks in Mill Creek Canyon in 1865. One struck a group of renegades camped temporarily on a gravel bar downstream from Black Rock and killed perhaps 10. The other, a few weeks later upstream from Black Rock, hit the last village of the Yahi, and left 30 or more dead, the majority of the tribe.
And I further believe that second attack was the one long misplaced at Kingsley Cave. The following chapter will explain why.
Sorry, comments are closed.