Proctor’s Ledge: Perley Put Salem’s Hangings There in 1921

Where did the old story put the gallows?

The 1692 accounts read here point only at “the Hill.” Calef, a Boston merchant who wrote against the trials, says the cart carrying eight people to execution on September 22 was “going to the Hill” when it stuck (Exhibit B). In his 1866 edition, Samuel G. Drake added a note that “The Hill” “has ever since been pointed out as Witch Hill, or more generally Gallows Hill.”

Charles Upham's history of 1867 chose a spot on that hill. He described a bare cliff on the western side of Salem and wrote, “At the top of this cliff, on the southern brow of the eminence, the executions are supposed to have taken place” (Exhibit E). He suggested the town chose the highest point to defy the Devil in his own realm, the air.

Upham was candid about his footing. “There is no contemporaneous nor immediately subsequent record,” he wrote, “that the executions took place on the spot assigned by tradition.” His support was a letter of 1791 reporting that a man born in 1692 had heard from his mother's nurse that she saw the condemned “hanging on Gallows' Hill” from a chamber window.

Perley later took that letter apart (Exhibit F, p. 8). The house where the man died was not built until 1730, so it could not be where he was born. He was born on May 22, 1692, Perley added, so the only hanging the nurse could have seen in those weeks was Bridget Bishop's, on June 10.

Gallows Hill, Salem, where the witches were hung: the title this photograph carries in the Boston Public Library's catalogue, dated between 1895 and 1905. It is a title from around 1900; this page sets out where the evidence now places the hangings, and the photograph does not show which part of the hill is in view.

The hilltop story sits at the edge of what this site calls courtroom folklore, popular versions of a famous trial that contradict the surviving record and still outlive it. The 1692 papers read here do not name the spot, so they never contradicted the summit directly. What undercut it came later, from land records.

What did Perley find in the land records?

Perley spent years tracing Salem's land titles, roads and early houses. A letter he printed, dated October 16, 1911, already mentions his articles on the site (Exhibit F, pp. 17–18). His full case appeared in the Essex Institute Historical Collections in January 1921.

His method shows in the first clue he tested. A newspaper item of about 1845 passed on a family memory of a woman who had watched the hangings from her doorway. Her descendants later lived on Boston Street, and a reader would assume she had too. The land records, Perley found, showed that section was undivided common land with no houses until 1718, so the clue failed (Exhibit F, p. 1).

The records showed that in 1692 the whole rough, ledgy tract south of the North River was common land, with, Perley argued, no path to the top. The only way out of town in that direction crossed the river at Boston Street, and the old road skirted the foot of the ledges. Perley argued that the young sheriff, George Corwin, simply took the condemned to the nearest common land beyond the town.

Then the deeds narrowed the spot. Around 1746 a lot on the lower ledge passed from Moses Steward to Thorndike Proctor, a grandson of John Proctor, who had been hanged in 1692. Proctor planted locust trees there, and in 1775 his widow Abigail was assigned the acre on which, the probate record said, the locust trees stood (Exhibit F, pp. 10–11).

Those trees connect to a diary. On August 14, 1766, John Adams walked to “Witchcraft hill” and wrote that “Somebody within a few years has planted a number of locust trees over the graves,” as Perley printed the entry. Adams did not say which part of the hill he climbed.

The summit had already been searched. Perley reported that Abner C. Goodell, a Salem student of the trials, said he had searched the top and the sides of the hill, on and off for twenty years after Upham's book, for a crevice like the one Calef described, and found none (Exhibit F, pp. 13–14).

Perley also reread Calef's stuck cart. Winfield Scott Nevins, another writer on the trials, had taken it as proof that the cart was climbing a steep summit. Perley answered that Calef has the cart stuck while “going to the Hill,” so before it got there, and that “the afflicted and others” blamed the Devil, which they would hardly do for a plain steep slope.

Perley's distances, as he gave them in 1921, converted at 16.5 feet to the rod and 3 feet to the yard. Drawn from the table published with this page (Exhibit F, pp. 15–16).

Perley's distances, with conversions (CSV) (where the data comes from) Data: Sidney Perley, Essex Institute Historical Collections 57 (1921)

Family memory filled in the rest. Perley heard that Dr. Andrew Nichols, in 1849, had pointed at large trees on that lot and told his son it was where the witches were hung. The landowner's family said they felled two big trees there about 1860, one growing from a crevice in the rock. They gave Perley a small piece of one trunk; the great fire of 1914 later destroyed the house, the shed and the remaining sections (Exhibit F, pp. 12–13).

Perley rejected the idea of a built gallows, too: “The executions must have been upon the limbs of trees.” His reason was that no town or county record showed lumber or labor paid for one. The 1692 record is not so clear. Thomas Brattle, writing on October 8, 1692, described two of the condemned behaving well “from the Goal to the Gallows, and whilst at the Gallows” (Exhibit C).

What did a prisoner see on August 19, 1692?

Perley's case rested on deeds and memories. A line in the court papers, printed in 1864, gives it a witness from the day itself (Exhibit A).

Rebecca Eames of Boxford was charged with afflicting Timothy Swan by witchcraft. She was examined before the Salem magistrates on August 19, 1692, the same day five people were hanged (Exhibit D). Near the end of the record, she was asked if she was “at ye execution.” She said she was “at ye hous below ye hill” and saw a few people there.

Look at what the answer leaves out. It names no house and no hill, and the record does not say which execution was meant. It is also a clerk's summary of a prisoner's answers, most of them admissions, not a survey.

A schematic, not drawn to scale: Perley wrote that from the Symonds family house on North Street one looked over the river to the lower ledge, while the top of Gallows Hill could not be seen from it. Drawn for this page after Perley (Exhibit F, pp. 8 and 16).

Even so, the phrase fits Perley's ledge better than Upham's summit. The houses along Boston Street sat below the lower slope. According to Benjamin Ray, a member of the later research team, the local historian Marilynne Roach found the line in the court records and connected it to Perley's case (Exhibit H).

Perley singled out one of them, built about 1685 by John Maccarter, a dyer, about two hundred yards from the crevice (Exhibit F, p. 16). Maccarter sold it on November 12, 1692, for what Perley judged a price far below its worth, and Perley wondered, without proof, whether the hangings in view of its windows had something to do with it.

A site that rests on later testimony rather than a record made at the spot has a parallel in the Kensington Runestone, whose find story also comes down to what people said afterward.

What did the 2016 team add?

On January 12, 2016, Salem State University announced that two of its professors, the historian Emerson (Tad) Baker and the geologist Peter Sablock, had worked with the Gallows Hill Project to confirm Proctor's Ledge as the execution site (Exhibit G). The team's full report was not available to read for this page.

Ray's account in the University of Virginia's Salem archive fills in the method (Exhibit H). He writes that the team formed in 2010. By his account, a mapping specialist modeled what could be seen from the Boston Street houses, and with today's trees left out, the front of the ledge was in view while the summit was not. He also writes that the name Proctor's Ledge comes from its eighteenth-century owner, Thorndike Proctor, and that the City of Salem bought two small plots there in 1936 for a memorial that was never built.

The Proctor's Ledge memorial below the slope, photographed in May 2018. According to Ray, the City of Salem dedicated it on July 19, 2017, 325 years after the hangings of July 19, 1692.

The new work tested Perley rather than replacing him. His case had been land records and family memory; by Ray's account, the team brought in the Eames line, a line-of-sight model and a ground survey. None of it is a document from 1692 that names the spot.

Why are there no graves at the ledge?

Calef's account of August 19 explains why a grave would be hard to find. After George Burroughs was hanged, Calef wrote, “he was dragged by the Halter to a Hole, or Grave, between the Rocks, about two foot deep,” and put in with John Willard and Martha Carrier (Exhibit B). Calef adds that a hand, a chin and a foot were left uncovered.

Illustration

An illustration, not a record: a yard across the river from a low ledge, the kind of view Perley described from houses on North Street, with the ledge in sight and no hilltop beyond. No drawing or record from 1692 shows the place of the hangings.

Perley recorded a Buffum family tradition that Joshua Buffum, who lived on Boston Street, helped relatives carry bodies down to the river and home by boat (Exhibit F, p. 14). The same tradition held that Buffum saw the uncovered hand and foot from his house and went out after dark to cover them.

The hilltop, Perley pointed out, lay about 120 rods from that house, more than a third of a mile, while the crevice was about 53 rods away in open view (Exhibit F, p. 15). That is a family story printed more than two centuries later, not a record. The crevice the landowner's family pointed out had been emptied of its soil for their garden about 1860.

Ray writes that the team's survey used ground-penetrating radar and found crevices under the surface, but that the team chose not to excavate them. So the ledge has the same limit as any reading of ground without digging, which the page on reading layers of ground treats in general: a survey shows where soil and rock lie, not what happened there.