Reading Time: 67 minutes, [12335 words]
Negro Witnesses Under Pressure
Witness Tampering and Then Some: Manipulation Through False Pretexts, Bait-and-Switch Affidavits, Financial and Employment Inducements, Promised Opportunities, Disputed Testimony, Manufactured-Confession Allegations, Witness Removal, and the Campaign to Break Jim Conley
By Arthur Daniels
September 21, 2026
THE HIDEOUS AND UNTOLD HISTORY AFTER LEO FRANK'S VERDICT
There is a seldom-examined part of the Leo Frank case deserving far more attention than it receives. The controversy was never only about whether Frank was guilty or innocent, or whether Jim Conley was believable. It was about what happened to Negro witnesses after Frank’s defense investigators began searching for new affidavits, new confession stories, recantations, and other evidence capable of overturning his conviction.
Leo M. Frank, president of Atlanta’s B’nai B’rith lodge, was convicted on Monday, August 25, 1913, of murdering thirteen-year-old Mary Phagan. After the verdict, his attorneys and investigators launched an extensive campaign to obtain him another trial and ultimately vindicate him. They searched for new witnesses, gathered new affidavits, challenged State witnesses, pursued retractions and confession stories, and sought evidence pointing away from Frank and toward Jim Conley.
Again and again, this effort brought Frank’s investigators into contact with Negro witnesses whose testimony, affidavits, retractions, or accusations could help weaken the State’s case and strengthen the defense theory placing responsibility for Mary Phagan’s murder on Conley. What happened during those encounters forms an important and too often overlooked chapter in the history of the case.
Jim Conley was a central target of this effort.
Conley’s trial testimony supplied the prosecution with its detailed account of what happened inside the National Pencil Company after Mary entered the factory at noon on Saturday, April 26, 1913. Conley admitted helping move Mary’s body and writing the murder notes, but he said he did those things under Frank’s direction. If Frank’s investigators could destroy Conley’s credibility, place him somewhere inconsistent with his trial testimony, produce people claiming he confessed, develop degrading character evidence against him, or persuade Conley himself to accept responsibility for Mary’s murder, an essential part of the State’s case could be attacked and weakened.
A substantial portion of the surviving post-trial record concerns Negro witnesses and prospective Negro witnesses whose statements might damage Conley or assist Frank.
The principal Negro witnesses and prospective witnesses examined here are Annie Maud Carter, Albert McKnight, Ivey Jones, Mary Rich, Aaron Allen, William Calhoun, Mark Wilson, Frank Reese, and Fred Perkerson. A man named Whatley appears in the wider search. Jim Conley himself became the ultimate Negro target of several alleged confession schemes. Two additional Negro people appear without surviving names. One was reportedly sought to corroborate Reverend C. B. Ragsdale’s alleged Conley-confession story. Another was a Negro woman John Shields said Jimmie Wrenn wanted him to locate after Shields refused to give damaging testimony against Conley.
Minola McKnight occupies a different position. Her experience arose from State questioning in June 1913, months prior to the Burns-era post-trial campaign. She belongs here as a counterpoint because her detention, questioning, disputed affidavit, courtroom repudiation, and later history show why allegations of pressure upon Negro witnesses cannot be examined as though coercive pressure could flow from only one direction.
The sources themselves ordinarily used Negro, Negress, or colored. This article preserves Negro in the title and narration because it was the ordinary racial terminology found throughout the 1913 and 1914 newspapers, sworn affidavits, legal records, and court proceedings examined here. Historical quotations retain the wording of the sources.
The post-trial investigation ran through William J. Burns and men working under, beside, or around his operation, including Dan S. Lehon, C. W. Burke, C. A. Isom, J. W. “Jimmie” Wrenn, Dr. George Wrenn, Stiles Hopkins, and others. Their employment relationships were not identical. George Wrenn, for example, should not simply be called a Burns detective. The record places him more safely within the wider circle surrounding Frank’s defense investigation.
This article does not require believing every accusation made against Frank’s investigators.
Some witnesses changed stories. Some were prisoners. Some affidavits were met with counter-affidavits. Some claims were denied under oath. Some expenditures plainly represented ordinary investigative wages or travel expenses. A railroad ticket is not a bribe. A meal is not subornation of perjury. Paying an investigator a daily wage is not proof of corruption. An indictment is not a conviction. A grand jury accusation cannot be transformed into a jury verdict. The Ragsdale-related prosecution of Lehon, Thurman, and Tedder ended in acquittal, and any serious history has to say so.
Those qualifications do not weaken the strongest evidence.
They clear away the weak claims and leave the strongest ones standing.
One investigator admitted using a false railroad-accident story to question a Negro State witness about the Frank case. The lead investigator acknowledged directing the approach and preparing the resulting affidavit from the subordinate investigator’s notes. Albert McKnight said a much better-paying job was placed in front of him while a changed affidavit was being pursued and separately alleged words were added after his signature. Mary Rich said the strongest part of an affidavit attributed to her was untrue. Aaron Allen said he was pursued for a Conley-confession story even though he did not know Conley, while an investigator working inside the operation preserved Allen’s statement saying the desired affidavit would be false. Two Negro jail trusties said they were asked to manufacture a Conley confession. Carter was moved outside Georgia after giving evidence useful against Conley. A judge later described her removal as technical contempt until she was returned. The State formally accused people within Frank’s investigative orbit of manufacturing Carter-Conley correspondence, while the defense denied it and produced handwriting evidence.
The central historical question is therefore not whether every allegation became a criminal conviction.
The question is how often investigators crossed, or were accused under oath of crossing, the line separating investigation from witness manipulation.
That is where the concept of witness tampering becomes useful.
What witness tampering means
Witness tampering is a broad modern term for improper attempts to influence, pressure, intimidate, deceive, reward, punish, move, silence, or otherwise interfere with a person because of testimony or evidence connected with a legal proceeding.
The phrase should be used carefully in a 1914 history.
Georgia’s present statute is called Influencing Witnesses. It covers, among other conduct, offering a benefit, reward, or consideration with intent to deter truthful testimony, and it also addresses corrupt persuasion, attempted corrupt persuasion, intimidation, threats, and misleading conduct undertaken with intent to influence testimony. (Georgia Code, 1)
The modern Georgia statute was not the law governing the Frank investigation in 1914, so it cannot be applied retroactively as a criminal charge against historical actors. It is useful for explaining why conduct such as paying for false testimony, corruptly persuading someone to change evidence, or deliberately misleading a witness in order to affect testimony is now recognized as a serious interference with judicial proceedings.
Georgia did possess an older offense directly relevant to the Frank-era disputes: subornation of perjury.
Georgia’s present subornation statute defines the offense as procuring or inducing another person to commit perjury or false swearing. Its statutory history traces through the Penal Code of 1895 and Penal Code of 1910, placing the legal concept squarely inside the period of the Frank litigation. (Georgia Code, 2)
Georgia case law also draws an important line. Completed subornation requires the underlying false swearing actually to occur. A simple unsuccessful effort to get somebody to swear falsely does not, by itself, establish completed subornation.
This becomes important with witnesses such as Aaron Allen, Frank Reese, Fred Perkerson, and John Shields. Their accounts describe alleged efforts to obtain false evidence, but they said they refused.
When a witness refused, the historically careful phrase is attempt to procure false testimony, or attempted subornation where the actual historical proceeding used such a charge. In a broader explanatory sense, the alleged conduct can also be discussed as possible witness tampering because the alleged objective was to manipulate what a witness would say.
The words must follow the evidence.
A sworn accusation is a sworn accusation.
An admission is an admission.
A newspaper report is a newspaper report.
An indictment is an indictment.
An acquittal is an acquittal.
A court finding is a court finding.
Keeping those categories separate makes the evidence against questionable investigative methods stronger because no rhetorical shortcut is needed.
Naming the recurring methods
Several recurring methods deserve names so readers can recognize them when they reappear.
The Pretext Interview means approaching a witness through a false story, concealing the actual purpose of the questioning. Ivey Jones is the strongest example because J. W. Wrenn admitted using the false railroad-accident story himself.
The Post-Signature Rewrite describes Albert McKnight’s accusation claiming important language appeared in his affidavit after he had signed it.
The Inflated Affidavit describes a paper containing a stronger or more specific claim than the witness later said she actually supplied. Mary Rich provides the clearest example.
Money, Employment, and Witness Expenses covers several different categories which must not be confused. Ordinary investigator wages are one category. Legitimate travel expenses are another. Alleged rewards tied to testimony are a different category. Job opportunities allegedly connected with a desired affidavit are another. Visible cash appearing during an effort to obtain a highly specific story belongs in still another category. The importance lies in whether a benefit was simply incidental to investigation or was intended to influence what a witness said.
The Manufactured Confession describes an allegation involving an effort to create a confession story instead of discovering a genuine confession. Reese, Perkerson, Ragsdale, and Conley himself appear within this pattern.
The Substitute Witness describes looking for another person after the first person refuses, fails, or will not support the desired story. The unnamed Negro associated with the Ragsdale affair and the Negro woman sought through John Shields fit this category.
The Illiterate Signature Attempt describes Frank Reese’s account of being asked to authorize a document despite his inability to read or write.
The Interstate Witness Hunt describes the paid search reaching Indiana, Chicago, Virginia, and the chain-gang system for Negroes who might provide useful information against Conley.
The Pressure Interview describes aggressive confrontation of a witness who repeatedly denies possessing the information investigators seek. William Calhoun provides the clearest example.
Removal After Cooperation describes Annie Maud Carter being moved outside Georgia after she supplied an affidavit useful to Frank.
The Manufactured Correspondence Accusation refers to the State’s formal allegation claiming the Carter-Conley letters had been forged and manufactured through George Wrenn. The defense denied this accusation and produced handwriting evidence.
The Confession Dragnet describes the broader pattern linking Carter, Allen, Ragsdale, Reese, Perkerson, an unnamed Negro corroborator, and Conley himself. Several episodes centered upon finding someone who could say Conley confessed.
The Affidavit Repudiation Cycle describes another recurring sequence. An affidavit useful to Frank appeared. The supposed witness later disputed, narrowed, or repudiated it. Frank’s side then produced counter-affidavits or testimony defending the original paper.
None of these labels substitutes for evidence.
They simply allow the evidence to be compared across witnesses.
June 1913
Minola McKnight: detention, questioning, a repudiated affidavit, and a later reported reversal
Minola McKnight was the Negro cook in the Selig household, where Leo and Lucile Frank lived.
Her story begins on Monday, June 2, 1913.
The Atlanta Georgian reported her detention under the headline “Negro Cook at Home Where Frank Lived Held by the Police.” The paper said information leading to her detention was reported to have originated with her husband, Albert McKnight. It described intensive questioning in Solicitor General Hugh M. Dorsey’s office.
On Tuesday, June 3, The Atlanta Journal reported Minola had been questioned by Dorsey and detectives Campbell and Starnes. The paper described her as distressed and frightened. Attorney George Gordon was retained on her behalf and reportedly considered seeking habeas corpus.
The Atlanta Constitution independently reported her detention on June 3 under “Leo Frank’s Cook Put Under Arrest.” It described her crying over the possibility of punishment for something she said she knew nothing about. (AC, 3)
Minola eventually signed the statement known as State’s Exhibit J.
The exhibit survives in the Georgia Supreme Court record. It bears the date June 3, 1913. (Georgia Supreme Court, 4)
The statement contained highly damaging claims involving conversations at the Selig home after Mary Phagan’s death.
Yet Minola did not leave the statement unchallenged.
During the August 1913 trial, she repudiated it.
The trial record preserves her saying:
“None of the things in that statement is true. It’s all a lie.”
She said she signed because she feared remaining locked up. She accused investigators of pressuring her and said she had not told her husband the things he attributed to her. (Georgia Supreme Court, 5)
The State presented a different account of how the statement had been obtained. Detective Starnes testified he instructed Minola not to sign if the statement was untrue. Other evidence concerned attorney Gordon’s presence and the circumstances of the interrogation. The resulting record is internally disputed.
Minola therefore left two sharply different sworn positions.
The first was her June 1913 signed statement.
The second was her August 1913 courtroom repudiation.
Then a possible third position entered the history, but it did not arrive from Minola directly.
On Sunday, April 19, 1914, while Albert McKnight was publicly repudiating his own Burke affidavit, The Atlanta Constitution reported Albert saying Minola had privately told him after Frank’s trial her original statement was true. (AC, 6)
This claim must be described with precision.
It was not a third sworn statement from Minola.
It was Albert McKnight’s account of what he said Minola told him privately after trial.
Its source context also matters. Albert was making the claim while publicly defending his own credibility. He was saying Burke’s later affidavit did not represent the truth and insisting his trial testimony had been accurate. A claim claiming Minola privately moved back toward her original version also supported Albert’s own position.
This does not prove Albert invented the story.
It tells us why attribution is essential.
Minola’s record therefore contains three layers:
June 1913, State’s Exhibit J.
August 1913, Minola’s sworn repudiation in court.
April 19, 1914, Albert’s newspaper-reported claim concerning what Minola supposedly told him privately after trial.
Any article collapsing those three into a single settled version destroys the real historical problem.
Minola’s story also supplies an important counterpoint for the post-trial defense controversy. When Burns later argued Negro witnesses might need protection from Atlanta authorities, Minola’s detention and subsequent allegations gave such concerns historical context.
Pressure did not belong exclusively to one side.
But Minola’s experience does not excuse questionable defense conduct later.
It proves something simpler.
A poor Negro witness could become trapped between investigators, employers, attorneys, family members, competing affidavits, and public controversy.
December 7, 1913
Annie Maud Carter and the Poison-for-Profit allegation
Annie Maud Carter was confined in the Fulton County Jail while Jim Conley was also being held there.
Carter eventually became an important witness in Frank’s post-trial attack on Conley.
Her most disturbing accusation concerned an unidentified jail visitor.
Carter placed the event on the first Sunday in December 1913, which fell on December 7.
She said a black-haired man approached her and asked whether she wanted to “get rich right quick.”
According to Carter, the man produced a small vial and wanted her to put some of its contents into Conley’s food. She said money was promised.
Carter said she refused.
The accusation surfaced publicly during Frank’s active post-trial litigation.
On Tuesday, May 5, 1914, The Atlanta Journal reported Carter describing a “black-haired man whom she doesn’t know” asking whether she wanted to get rich quickly and instructing her to put “a drop or two in Conley’s food.” (AJ, 7)
The story reached national readers the following day when The New York Times reported the episode under the subheading “Poison Plot Against Conley?”
The evidentiary limitations are substantial.
Carter could not identify the man by name.
No surviving chemical test identifies the contents of the vial.
No criminal judgment established a poisoning conspiracy.
For those reasons, the episode must remain Carter’s sworn Poison-for-Profit allegation.
Its importance comes partly from timing.
This was not folklore created fifty or seventy years later. Carter made the accusation while Frank’s litigation remained active, and Atlanta newspapers reported it immediately.
A further point deserves attention. Carter placed Leo Frank and other visitors nearby during the episode, but she did not say Frank personally handed her the vial.
The distinction is important.
The article cannot transform proximity into proof of participation.
Yet Carter’s story reveals how completely Conley had become the center of the post-trial struggle. By late 1913 and early 1914, multiple lines of investigation were moving toward damaging him, discrediting him, extracting confessions from him, or finding others who could claim he confessed.
January through April 1914
Albert McKnight: persistent pressure, a Pullman job, and an alleged Post-Signature Rewrite
Albert McKnight had testified for the prosecution at Frank’s 1913 trial.
His testimony dealt in part with events at the Selig home and with information associated with his wife, Minola.
On January 19, 1914, C. W. Burke obtained a new affidavit from McKnight which weakened portions of his trial testimony.
McKnight later described Burke’s persistence in blunt language.
The Georgia Supreme Court record preserves him saying:
“Mr. Burke keeps after me.”
During another exchange, McKnight insisted his trial testimony had been truthful. Burke, according to the record, answered with profanity and accused him of lying. (Georgia Supreme Court, 8)
This was not a single interview followed by a spontaneous recantation.
McKnight described repeated pursuit.
Three months later, he went public.
On Sunday, April 19, The Atlanta Journal, page 1, column 6, published “Albert M’Knight Now Repudiates His Recent Story.”
McKnight said Burke asked whether he wanted a Pullman job paying $40 monthly plus tips, enough, McKnight believed, to bring his total monthly income close to $100.
McKnight’s own words were memorable:
“That was the way he got around me.”
(AJ, 9)
McKnight’s reported ordinary wages were about $28 per month.
A possible $100 income would have represented roughly 3.6 times his normal earnings.
For perspective, a worker earning $28 each month would need more than three and a half months to earn $100. A job capable of producing $100 every month was not a small favor.
If McKnight’s account were accurate, the potential job represented a major economic improvement.
The same Sunday, The Atlanta Constitution, page 1, column 6, published “Testimony He Gave at Trial Was True, Declares M’Knight.”
McKnight said:
“I signed that affidavit that Mr. Burke brought me just because I wanted to get rid of him.”
He insisted his trial testimony had been truthful.
Then came an even more serious accusation.
McKnight said language blaming former employers Roy Craven and C. C. Pickett for framing his trial testimony had been added after he signed the affidavit. (AC, 6)
If true, this would move far beyond persistent questioning.
Knowingly adding substantive words to a signed sworn statement would manufacture evidence under a real witness’s name.
In modern language, conduct of this type would raise obvious witness-tampering and obstruction concerns because the witness’s statement itself would be altered to affect a legal proceeding.
The historical evidence does not establish the allegation as a judicial finding.
Burke denied it.
Burke also denied promising McKnight a job or reward in return for the affidavit.
The defense produced substantial rebuttal material.
Minola McKnight swore Burke had made no promise of reward to her husband.
Dan Lehon, William Burns, Dr. Thomas Hancock, Charles Phillips Jr., Fred Lynn, and others provided statements supporting the position claiming McKnight had voluntarily reaffirmed Burke’s affidavit.
Burke did acknowledge later trying to help McKnight obtain employment, including contact concerning a job at a guano factory.
His explanation was straightforward.
The employment discussion happened after the affidavit and was not payment for it. (Georgia Supreme Court, 10)
This is why the McKnight episode cannot honestly be summarized as “Burke bribed McKnight.”
No such conviction exists.
The historically sound conclusion is more precise.
McKnight accused Burke of using the prospect of dramatically better employment as leverage while pursuing a changed affidavit. He separately accused Burke of post-signature alteration. Burke denied both forms of misconduct and produced supporting witnesses.
Yet even after the denials are included, one fact survives.
Burke repeatedly pursued a Negro State witness for a changed statement and openly told him his trial testimony was false.
The validity of the new affidavit then became a major controversy of its own.
That is a recurring feature of Frank’s post-trial campaign.
The effort to create new evidence repeatedly produced new disputes about how the new evidence had been created.
February 1914
Ivey Jones and the admitted Pretext Interview
Ivey Jones provides the strongest direct evidence of deliberate deception because the essential conduct was admitted by Frank’s own investigators.
Jones was a Negro truck driver and State witness.
The Atlanta Georgian later identified him as:
“Ivey Jones, a negro and cousin of Jim Conley.”
(AG, 11)
C. W. Burke instructed J. W. Wrenn to locate him.
Wrenn did not introduce himself honestly as a man gathering evidence for Leo Frank.
His own affidavit explains what happened.
Wrenn represented himself as investigating a railroad accident.
He invented an injured person who supposedly had identified Jones as a witness.
The fictional accident allowed Wrenn to ask Jones where he had been on April 26, 1913, and whom he had seen, without initially telling Jones the Frank case was the real subject.
Wrenn did not deny this deception.
He explained it.
His sworn affidavit says:
“This pretense of looking into an accident was adopted”
because he believed Jones would tell the truth if his attention were moved away from the Frank case. (Georgia Supreme Court, 12)
The word “pretense” is Wrenn’s own.
This is important.
The deception does not depend on Dorsey’s rhetoric.
It does not depend on Jones’s accusation.
It does not depend on a hostile newspaper interpretation.
The investigator himself admitted using a fictitious cover story.
Burke’s role makes the episode even more significant.
Burke acknowledged directing Wrenn to Jones.
Then, according to Burke’s own affidavit, he used Wrenn’s notes to dictate the affidavit later attributed to Jones. (Georgia Supreme Court, 12)
The chain therefore ran from Burke to Wrenn, from Wrenn’s false railroad story to Jones, from Jones’s answers into Wrenn’s notes, and from those notes into an affidavit dictated by Burke.
Jones later repudiated the affidavit attributed to him.
He said it was false.
The defense answered claiming Jones heard the affidavit read, corrected a point, swore to it, and signed voluntarily.
The forgery accusation remains disputed.
The original deception does not.
For a modern reader trying to understand witness tampering, this episode provides an important example of why misleading conduct matters.
An investigator may interview a witness.
An investigator may ask difficult questions.
An investigator may test a witness’s memory.
But once an investigator intentionally lies about his identity or purpose in order to obtain evidence concerning pending litigation, the integrity of the resulting statement becomes a legitimate issue.
Modern Georgia law expressly recognizes “misleading conduct” undertaken with intent to influence testimony as potentially criminal witness interference. Again, this modern law cannot be retroactively imposed upon Wrenn. It helps explain why the conduct deserves scrutiny.
The defense’s justification was essentially utilitarian.
Wrenn believed deception would produce truth.
Yet the historical difficulty is obvious.
An investigator who decides deception is acceptable because it may produce truth also becomes the person deciding which deception is permissible, which answers are useful, which notes are carried forward, and how the final affidavit is written.
Jones’s case is therefore more than a curiosity.
It exposes a basic weakness in the post-trial evidence-gathering operation.
The defense wanted evidence against Conley badly enough to hide its purpose from a Negro State witness while questioning him.
Then the defense turned the resulting interview into a sworn paper.
March through May 1914
Mary Rich and the Inflated Affidavit
Mary Rich operated a food stand near the National Pencil Company factory.
Her potential value to Frank was obvious.
If Rich could place Jim Conley near the factory’s rear alley around 2:15 or 2:30 in the afternoon on April 26, 1913, the sighting could interfere with the State’s chronology.
When Frank’s extraordinary motion became public on Friday, March 27, 1914, The Atlanta Journal, page 1, column 2, published “Frank’s Motion Is Made Public: Many of State’s Witnesses Change Their Testimony.”
The paper carried the subheading:
“Negro Woman Testifies She Saw Conley.”
The newspaper reported the defense affidavit as claiming Conley came running out of the alley behind the factory around 2:15, bought a twenty-cent lunch, ate it, and returned toward the alley. (AJ, 13)
This was strong evidence if accurate.
The problem was what happened next.
On Wednesday, April 22, The Atlanta Journal reported Rich’s deposition.
She still remembered Conley buying a twenty-cent lunch between approximately 2 and 2:30.
But the paper reported she did not remember where Conley came from or where he went afterward. (AJ, 14)
The newspaper itself contrasted the deposition with the stronger claim already appearing in Frank’s motion.
This creates a revealing chronology.
First came the strongest version.
Conley emerged from the factory alley and returned toward it.
Then came a narrower version.
Conley bought lunch at approximately the same time, but Rich did not remember his direction of travel.
Then, on May 5, came direct repudiation.
The Atlanta Constitution reported Rich saying:
“The affidavit was not true, and I refused to sign it.”
She added:
“I could not tell a lie.”
(AC, 15)
Rich said Lucile Frank, Rabbi David Marx, and others had visited her regarding the affidavit.
The Georgia Supreme Court record preserves Rich’s denial in nearly the same form.
E. H. Wellborn supplied a supporting affidavit concerning part of the encounter. (Georgia Supreme Court, 16)
The defense did not accept Rich’s repudiation.
Rabbi Marx swore Rich had acknowledged the substance of the affidavit and had corrected only the time, moving it from approximately 2:15 to approximately 2:30. (AJ, 17)
The two versions cannot both be complete.
Either Rich had substantially confirmed the alley story and later denied it, or the defense’s affidavit contained a stronger version than Rich was willing to authorize.
The April 22 deposition is what makes this case especially important.
It sits between the two extremes.
It does not contain the full alley claim.
The weakening therefore did not begin only when Rich appeared against the defense in May.
A narrower account already existed in the documentary sequence.
For a study of possible witness tampering, Rich’s case is important because it illustrates a recurring danger in affidavit-driven litigation.
The statement most useful to the party collecting it can become stronger on paper than the witness’s later account.
This does not prove intentional fabrication.
It demands careful examination of who drafted the paper, what questions were asked, what the witness actually said, whether the witness signed, and what later corrections followed.
Rich’s own words make the issue impossible to avoid.
She said she would not sign because she “could not tell a lie.”
The defense said she had already confirmed the substance.
The source record preserves both.
April 23 and April 25, 1914
Annie Maud Carter’s affidavit session and the printed date
Annie Maud Carter eventually supplied evidence useful to Frank’s attack on Conley.
The dating of her affidavit has caused confusion, but the legal record itself allows a more precise reconstruction.
Georgia Supreme Court page 1764 contains the sworn statement of the men present with Carter in the Haas offices.
Their statement places the affidavit session on Thursday, April 23, 1914. (Georgia Supreme Court, 18)
The Atlanta Journal independently uses April 23 for the session. (AJ, 7)
Georgia Supreme Court page 1765, however, gives April 25, 1914 as the printed date appearing on the resulting affidavit. (Georgia Supreme Court, 19)
Those two pieces should not be forced into one date.
The strongest chronology is:
April 23 was the actual Haas-office affidavit session according to the sworn witnesses and contemporary newspaper chronology.
April 25 was the printed date appearing on the resulting paper.
A sworn account of who was physically present on a particular day carries greater evidentiary weight for reconstructing the session than a bare dateline printed on the completed document.
The distinction matters because chronology is central to Carter’s subsequent movements and the question of when money, travel arrangements, and investigative contacts occurred.
The witnesses present during the session also swore Carter was asked directly whether she had received money or any promised reward.
According to their statement, Carter said she had not. (Georgia Supreme Court, 19)
This defense evidence must remain visible.
Carter later accused people around Frank’s investigation of giving her money and inducing testimony.
The earlier denial means her later allegation cannot simply be inserted into the chronology as though no contrary evidence exists.
One possible explanation is timing.
Carter could have truthfully denied being paid for the April 23 session and later received money or benefits.
Another possibility is she later changed her account.
The evidence does not conclusively resolve the issue.
April 26 through May 2, 1914
The Carter-Conley letters become a public weapon
By Sunday, April 26, the supposed Carter-Conley correspondence had become a major part of William J. Burns’s public campaign against Conley.
The Atlanta Constitution published “Says Love Letters Written by Conley Prove Him Guilty.”
Burns described the writings as:
“proof conclusive”
and argued they connected Conley to the murder notes and Mary Phagan’s murder. (AC, 20)
The Atlanta Georgian published similar claims the same day under “New Conley Notes Prove Him Guilty, Says Burns.” (AG, 21)
This was not quiet investigative material waiting for judicial evaluation.
Burns was publicly presenting the alleged correspondence as powerful proof of Conley’s guilt.
That makes the later dispute over authorship and handling much more important.
On Friday, May 1, The Atlanta Georgian reported Carter describing a large volume of correspondence. The defense ultimately offered twenty-two sheets attributed to Carter and twenty-seven sheets attributed to Conley. (AG, 22)
Carter later disputed vulgar content attributed to Conley’s original letters.
Then the State made a formal accusation going far beyond disagreement over interpretation.
The Manufactured Correspondence Accusation
Georgia Supreme Court page 1495 contains the State’s counter-showing.
The State accused people around the defense of presenting notes which were:
“forged and manufactured”
through a conspiracy involving George Wrenn. (Georgia Supreme Court, 23)
This was the State’s accusation.
It was not a judicial finding.
The defense denied fabrication and produced handwriting testimony supporting authenticity.
No ruling cited here establishes the letters were forged.
Yet surrounding evidence explains why the allegation became a serious part of the litigation.
Frank Reese swore he had seen George Wrenn and Carter speaking frequently.
Reese also described one occasion when Wrenn threw Carter a note and Carter carried it toward Conley’s cell wing. Reese did not know what the note contained. (Georgia Supreme Court, 24)
Willis J. Gilbert, an inside jail guard, independently swore Carter and George Wrenn talked “very frequently,” generally in the medicine room.
Gilbert also said Carter requested permission to enter Conley’s cell wing and he refused. (Georgia Supreme Court, 25)
Then William Burns supplied another connection.
On Saturday, May 2, The Atlanta Georgian reported Burns saying Burke had furnished him the notes and George Wrenn made the first translation. (AG, 26)
None of these facts proves forgery.
A note passed near a jail cell is not necessarily a forged letter.
Frequent conversation is not proof of conspiracy.
Making a translation is not proof of fabrication.
But these facts create a chain of custody which was anything but remote.
Carter was providing the evidence.
Wrenn was in repeated contact with Carter.
Reese placed Wrenn passing paper to Carter near Conley’s cell area.
Burke supplied Burns the notes.
Wrenn made the first translation.
Burns then presented the correspondence publicly as “proof conclusive” against Conley.
The State accused Wrenn of participating in manufacturing the evidence.
The defense denied the charge.
For a study of witness tampering and evidence manipulation, the episode is highly significant even without a finding of forgery.
The dispute concerns not simply what a witness remembered.
It concerns the creation, handling, translation, promotion, and authenticity of physical written evidence introduced to damage the prosecution’s principal Negro witness.
Spring 1914
Aaron Allen and the Confession Dragnet
Aaron Allen’s importance arose from a very specific possibility.
Investigators believed he might say Jim Conley confessed to him.
Allen said he did not know Conley.
Burns’s own testimony establishes how much effort went into pursuing him.
On Saturday, May 2, The Atlanta Georgian reported Burns acknowledging an employee named O’Neill in Chicago had gone into Indiana to obtain Allen while Allen was sick. Burns denied abusing him, denied paying him, and denied buying a false story. (AG, 26)
The significance is not a train ticket.
The significance is the objective.
An investigator was dispatched across state lines to obtain a Negro prospective witness because Frank’s investigators wanted information concerning a Conley confession.
Allen’s sworn statement preserves an extraordinary exchange.
Allen asked Burns:
“Mr. Burns, do you want me to tell you the truth or to tell you a story?”
Burns answered:
“Mr. Allen, we don’t ask anybody to tell us lies, but the plain truth.”
(Georgia Supreme Court, 27)
This is important defense evidence.
If the record ended there, Burns’s response would support the position claiming he sought honest information.
The record does not end there.
C. A. Isom’s affidavit gives the episode another dimension.
Isom said Allen did not know Conley and had never spoken with him.
According to Isom, he reported this to Burns.
Burns responded:
“you can make him come across.”
(Georgia Supreme Court, 28)
Isom returned to Allen.
Allen then said:
“I will make that affidavit, but it will be a lie.”
(Georgia Supreme Court, 28)
This exchange is among the strongest pieces of evidence in the entire Negro-witness record.
The pressure is not described solely by Allen.
An investigator participating in the operation preserved it.
Isom also swore Burke instructed him, while working on Allen, to use dinners, cigars, entertainment, and whatever money was necessary. (Georgia Supreme Court, 29)
The distinction between ordinary expenses and improper inducement is essential.
Buying dinner for a witness does not prove corruption.
Giving an investigator expense money does not prove corruption.
The key question is purpose.
Was the spending designed simply to make contact with a witness and learn what he knew?
Or was the purpose to move a man who denied knowing Conley toward a statement claiming Conley had confessed?
Allen’s own account adds more.
On Tuesday, May 5, The Atlanta Journal reported Allen saying Isom told him there was:
“a lot of money in it for him”
if he supplied the wanted information. (AJ, 7)
Allen also described being left alone in a room containing a large pile of greenbacks and silver.
He said he looked at the money but did not touch it.
Allen did not say Burns explicitly offered him the entire pile in exchange for perjury.
Burns denied the account.
A careful article cannot transform visible cash into a proved bribe.
But Allen’s story does not depend on the pile of cash alone.
Its force comes from the combination.
The defense was seeking a specific Conley-confession story.
Allen denied knowing Conley.
Isom said Allen did not know Conley.
Isom said Burns instructed continued pressure.
Allen told Isom the wanted affidavit would be a lie.
Burke authorized spending during the effort.
Allen alleged money was mentioned.
Burns denied improper payment.
This is exactly why witness tampering is a useful analytical concept.
The question is not whether investigators were allowed to speak with Allen.
Of course they were.
The question is whether the investigation kept pursuing a desired statement after its own investigator knew the proposed witness had no factual basis for giving it.
If Isom’s account is accurate, the effort ceased being an open search for whatever Allen knew.
It became an attempt to make Allen “come across” with a story investigators wanted.
Spring 1914
The Interstate Witness Hunt: Calhoun, Wilson, and Whatley
The search did not stop with Allen.
C. A. Isom’s affidavit describes a paid effort to locate other Negro prospective witnesses.
Isom said Burke paid him three dollars per day and provided another $100 for expenses connected with Chicago. (Georgia Supreme Court, 29)
Isom reported Mark Wilson had gone to Virginia.
He located William Calhoun at 4238 Wabash Avenue in Chicago.
He learned Whatley had been in the chain gang and went looking for him after release.
The figures help show scale.
At three dollars per day, a thirty-day period of continuous work would equal ninety dollars.
The additional $100 Chicago expense fund exceeded thirty-three days of Isom’s daily pay.
This was a financed investigative operation, not casual conversation among acquaintances.
William Calhoun was eventually questioned in Chicago.
Isom swore he overheard Burns confronting him with profanity:
“You are a damn liar, you bastard, you!”
(Georgia Supreme Court, 30)
Calhoun denied knowing Conley.
Attorney Stiles Hopkins later acknowledged an affidavit had been taken from:
“a negro named Calhoun”
but said its contents were of doubtful relevance and did not use it in the extraordinary-motion hearing. (AJ, 17)
Calhoun’s case should not be inflated into a bribery accusation.
No surviving statement reviewed here shows Calhoun accusing investigators of offering him money for false testimony.
Mark Wilson and Whatley likewise left no known personal complaints.
Their importance is structural.
The search for Negro evidence extended across state lines, workplaces, neighborhoods, and institutional systems.
Investigators were actively hunting witnesses capable of producing something useful against Conley.
Calhoun’s statement was obtained and then discarded as insufficiently useful.
The episode illustrates how much of Frank’s post-trial evidence came through a filtering process.
Investigators looked for people who might help.
They interviewed them.
They gathered statements.
Statements lacking value to the motion could be set aside.
Statements carrying value moved forward.
This is normal to some degree in legal investigation.
It becomes problematic when a desired outcome begins driving the questioning itself.
May 5 and May 6, 1914
Frank Reese and Fred Perkerson: the alleged Manufactured Confession
Frank Reese and Fred Perkerson were Negro trusties at the Fulton County Jail.
Both accused Dr. George Wrenn of proposing a scheme centered directly upon Jim Conley.
According to their sworn statements, Wrenn wanted them to enter Conley’s cell wing, leave later, and claim Conley had confessed to murdering Mary Phagan.
Reese said Wrenn discussed supplying cigars and obtaining:
“lots of money”
for the story. (Georgia Supreme Court, 24)
Perkerson described substantially the same proposal.
He said the men could enter Conley’s wing and later report a confession, with substantial money available if they cooperated. (Georgia Supreme Court, 31)
Both men said they refused.
George Wrenn denied the accusations.
On Wednesday, May 6, The Atlanta Journal reported Wrenn:
“denied positively”
ever offering the Negroes a reward to obtain a Conley confession. (AJ, 17)
The evidentiary situation is therefore straightforward.
Two Negro trusties supplied parallel allegations.
The accused man denied both.
No court ruling established the alleged scheme as fact.
Yet the similarities between Reese’s and Perkerson’s accounts make the accusations historically consequential.
Both described the same basic objective.
Enter Conley’s area.
Leave.
Say he confessed.
Receive benefits or money.
A genuine confession would have been enormously important evidence for Frank.
A fabricated confession would have been enormously useful false evidence.
This explains why the allegation, if true, would represent a particularly serious form of witness tampering and attempted subornation.
The objective was not to influence a minor detail.
It was to manufacture the central piece of evidence capable of shifting the murder itself onto Conley.
The alleged proposal reaches Jim Conley
Reese and Perkerson said the alleged effort did not stop with them.
Perkerson swore Wrenn told Conley the sensible course, after receiving his existing sentence, would be to accept responsibility for Mary Phagan’s murder and thereby free Frank.
According to Perkerson, money and limited additional punishment entered the alleged proposal. (Georgia Supreme Court, 32)
Reese independently described hearing Wrenn tell Conley he could take Mary Phagan’s murder upon himself and free Frank. (AC, 15)
Wrenn denied the story.
No judicial finding established such a proposal.
Still, two separate Negro jail trusties described the same ultimate objective.
This point changes the scale of the allegation.
Investigators were not, according to Reese and Perkerson, simply searching for somebody who had heard Conley confess.
The alleged effort eventually reached Conley himself.
If their accounts were true, Conley was being asked to create the confession the defense needed.
That would be witness tampering in its most direct conceptual form.
A person connected with the defense investigation would be trying to induce the prosecution’s central Negro witness to accept guilt for a murder and thereby change the outcome of Frank’s case.
Again, the article does not present this as a proved crime.
It presents two sworn allegations and a sworn denial.
The reader should see both.
Frank Reese and the Illiterate Signature Attempt
Reese described another episode after leaving jail.
He said Wrenn arrived at his home carrying a long white paper and wanted him to sign.
Reese said:
“I cannot read or write.”
(Georgia Supreme Court, 33)
He refused until he understood the paper.
According to Reese, Wrenn explained the document as concerning notes between Carter and Conley rather than a confession.
Wrenn acknowledged seeking a written statement but disputed Reese’s explanation of its purpose.
The contact therefore is not seriously disputed.
Wrenn wanted a statement from Reese.
The controversy concerns what the paper said and whether its purpose was accurately represented to a man unable to read it himself.
The power imbalance is obvious.
One man arrived with a prepared written document.
The other could not independently read a single line.
Reese’s refusal prevented his name from being placed onto a statement he did not understand.
Even if Wrenn’s purpose was entirely innocent, the episode demonstrates why signature practices involving illiterate witnesses demand unusually careful safeguards.
The Fulton County Jail
The network around Frank, Burke, Carter, and the Wrenns
The jail evidence is important because it establishes access independently of the disputed allegations.
Willis J. Gilbert identified himself as an inside guard.
Gilbert swore he saw Jimmie Wrenn and C. W. Burke enter together to visit Frank in his cell wing.
He said Burke spoke privately with George Wrenn.
Gilbert also said George Wrenn visited Frank’s wing frequently. (Georgia Supreme Court, 25)
The Georgia Supreme Court record separately identifies George Wrenn as Jimmie Wrenn’s brother and identifies Jimmie as a man assisting C. W. Burke. (Georgia Supreme Court, 34)
These contacts do not prove conspiracy.
They prove opportunity and familiarity.
The people appearing in the disputed Carter correspondence, the alleged confession scheme, and the affidavit controversies were not isolated strangers whose names happened to appear independently.
They moved through the same jail environment.
Burke was working for Frank.
Jimmie Wrenn assisted Burke.
George Wrenn was Jimmie’s brother.
George visited Frank’s wing.
Gilbert saw Burke and Jimmie Wrenn enter the wing together.
Gilbert saw Burke speak privately with George.
Carter and George Wrenn spoke frequently according to Gilbert.
Reese said he saw Wrenn pass Carter a note.
Reese and Perkerson later accused George Wrenn of seeking a manufactured confession.
None of this proves their accusations.
It establishes a real network of access within which the accusations arose.
This matters because the defense response cannot simply treat each allegation as an unrelated invention by an unreliable Negro witness.
The people named in those accusations actually had contact with one another.
The physical opportunity existed.
The investigative relationships existed.
The jail access existed.
The remaining historical question concerns what those contacts were used for.
May 9 through May 26, 1914
Annie Maud Carter and Removal After Cooperation
Carter’s handling after she supplied evidence useful against Conley became a courtroom controversy in its own right.
On Saturday, May 9, The Atlanta Journal reported Judge Benjamin H. Hill questioning Carter about her departure from Atlanta.
Asked who told her to leave, Carter answered:
“Mr. Burns told me to go.”
(AJ, 35)
Carter said a Burns operative traveled with her.
She said her expenses were covered.
She said she received five dollars in spending money.
Her journey moved through several cities and eventually outside Georgia.
Judge Hill required Burns and Lehon to explain why a witness connected with pending litigation had been removed from his jurisdiction.
The defense explanation was protection.
Burns said Atlanta detectives might pressure Carter into changing her statement.
Minola McKnight’s earlier experience explains why such fear could not simply be dismissed as imaginary.
Yet the legal issue did not disappear because Burns claimed a protective motive.
Carter was a witness whose evidence favored Frank.
She had supplied an affidavit.
Then she was moved out of Georgia under defense-associated arrangements.
Judge Hill considered the removal serious enough to intervene.
The final ruling deserves exact treatment.
On Tuesday, May 26, The Atlanta Journal reported Judge Hill found Burns and Lehon had been in:
“technical contempt of court”
for removing Carter after her affidavit.
But Carter had been returned to Georgia and was once again subject to legal process.
Hill therefore held the men had purged themselves of contempt and dismissed the rule. (AJ, 36)
This is much more relevant to the Negro-witness story than any unrelated private-detective licensing case.
A court actually addressed the handling of Carter as a witness.
The court found her removal technically contemptuous.
The problem was cured by returning her.
The proceeding was dismissed.
That is the complete legal outcome.
Carter’s later money accusation
Carter later accused Lehon of giving her small amounts of money and inducing her to swear against Conley.
This accusation remained disputed.
It should not be confused with a judicial finding.
The defense possessed earlier sworn statements claiming Carter denied receiving money or a promised reward when her affidavit was obtained.
The chronology therefore leaves several possibilities.
Carter may have received no payment for the April 23 affidavit and received money later.
Her later accusation may have referred to ongoing support rather than purchase of the original affidavit.
She may have changed her story.
The evidence does not allow certainty.
What can be said is more limited.
A Negro witness favorable to Frank was moved outside Georgia after supplying useful evidence.
Defense-associated men financed the movement.
A court found the removal technically contemptuous until the witness was returned.
The witness later accused one of the men involved of giving her money and inducing testimony against Conley.
This sequence belongs directly in any study of witness tampering because control over a cooperating witness’s location, expenses, and access to legal process can affect the integrity of pending proceedings even when the stated purpose is protection.
Spring 1914
John Shields and the search for a Substitute Witness
John Shields’s race is not established in the sources reviewed here.
He is not counted among the Negro witnesses.
His evidence belongs here because it directly concerns an alleged attempt to recruit a Negro woman against Conley.
Shields said he knew Burke and Jimmie Wrenn and saw them around the National Pencil Company while they worked on Frank’s case.
According to Shields, Wrenn repeatedly tried to persuade him to sign an affidavit accusing Conley of a degrading sexual act.
Shields said:
“I refused every time because it was a lie.”
(Georgia Supreme Court, 37)
He also said Wrenn offered to:
“dress me up and send me to Cincinnati or anywhere else I wanted to go.”
(Georgia Supreme Court, 38)
When Shields continued refusing, he said Wrenn asked whether he knew:
“a negro woman”
who could provide similar testimony.
Shields said he did not.
Jimmie Wrenn denied the episode.
The historical value lies in the alleged sequence.
A prospective witness refused a story because he said it was false.
According to him, the investigator then asked for another person, specifically a Negro woman, who could supply similar evidence.
If Shields’s account is accurate, this is almost a textbook description of a substitute-witness search.
The desired accusation existed first.
The witness was then sought.
This reverses the healthy order of investigation.
Proper investigation asks a person what he knows.
Improper evidence construction begins with what the case needs and then searches for a person willing to say it.
No unnamed Negro woman appears later in the surviving record, so there is no evidence claiming such a woman was actually recruited.
The significance is Shields’s sworn allegation concerning the request.
May 1914 through February 1915
Ragsdale, Barber, and the unnamed Negro corroborator
The Reverend C. B. Ragsdale affair became the most formal legal test of accusations surrounding the post-trial investigation.
Ragsdale supplied Frank’s defense with an affidavit claiming he had overheard one of two Negro men confess to Mary Phagan’s murder.
R. L. Barber supplied supplementary identification intended to connect one of the men with Jim Conley.
Ragsdale later repudiated his affidavit.
His later account described another Negro prospective witness.
According to Ragsdale, attorney Arthur Thurman said a Negro had been found to corroborate the confession story.
Then:
“The Negro, however, fell down.”
Another person was supposedly sought afterward. (AC, 15)
The unnamed Negro left no known affidavit of his own.
We do not know his name.
We do not know what he was asked.
We do not know why he “fell down.”
This evidence is therefore secondhand.
Its importance comes from the alleged method.
A desired confession story existed.
A corroborating Negro was supposedly found.
He failed.
Another person was allegedly sought.
Ragsdale’s account went farther.
He claimed Thurman told him he could make money from the Frank case if he framed a statement useful to the defense.
Ragsdale said the first version was considered weak in places and claimed Thurman said weak sections could be strengthened.
Ragsdale alleged $200 for himself and $100 for Barber. (AC, 15)
These accusations were serious enough to reach a grand jury.
On Friday, May 22, The Atlanta Journal reported Fulton County indictments.
Dan S. Lehon, Carlton C. Tedder, and Arthur Thurman were indicted for subornation of perjury.
Ragsdale and Barber were indicted for perjury. (AJ, 39)
Here the phrase subornation of perjury is not an analytical label created for this article.
It was the actual historical criminal charge.
An indictment is not a conviction.
The case moved to trial.
The Ragsdale subornation trial
On Tuesday, January 26, 1915, The Atlanta Journal reported Lehon, Thurman, and Tedder had gone to trial for subornation of perjury arising from the Ragsdale-Barber affidavits. (AJ, 40)
During the trial, evidence concerning investigative money became part of the record.
On Friday, January 29, The Atlanta Journal reported testimony concerning $500 Herbert Haas had provided as part of funding connected with the investigation.
Testimony described $250 as Tedder’s monthly salary and another $250 as travel expenses used while locating witnesses.
The defense maintained these were legitimate investigative expenditures and had no connection with purchasing Ragsdale’s affidavit. (AJ, 41)
On Sunday, January 31, The Atlanta Constitution, page 1, column 1, published “Fate of Burns’ Agents Now in Hands of Jury; No Verdict Until Today.”
Judge Hill reduced the jury’s task to two principal questions.
Was the Ragsdale affidavit false?
Had Ragsdale been procured through persuasion, promised money, or payment?
The jury deliberated for nearly sixteen hours. (AC, 42)
Then arrived the verdict.
On Monday, February 1, The Atlanta Constitution reported:
“not guilty.”
Lehon, Thurman, and Tedder were acquitted. (AC, 43)
This ending is essential.
The Ragsdale affair cannot honestly be described as a proved criminal conspiracy to suborn perjury.
The defendants stood trial.
The jury did not convict them.
But acquittal does not erase the historical events preceding the verdict.
Ragsdale did repudiate his affidavit.
He did make accusations involving money.
A grand jury did indict.
An unnamed Negro corroborator did appear in Ragsdale’s account as somebody who supposedly failed to support the desired story.
The criminal prosecution did occur.
The jury deliberated for nearly sixteen hours.
Then the defendants were acquitted.
The complete history contains all of those facts at once.
The Confession Dragnet
One of the most striking features of the entire post-trial investigation is the repeated appearance of a Conley-confession narrative.
Several otherwise separate episodes move toward the same destination.
Carter supplied claims about statements Conley supposedly made in jail.
The Carter-Conley correspondence was used to attack Conley and portray him as deceptive or incriminating.
Aaron Allen was pursued because investigators hoped he could say Conley confessed.
Allen denied knowing Conley.
Ragsdale supplied a confession story and later repudiated it.
An unnamed Negro was allegedly sought to corroborate Ragsdale.
Reese and Perkerson said they were asked to claim Conley confessed.
Those same two Negro trusties said Conley himself was asked to take responsibility for Mary Phagan’s murder.
No single episode proves a unified criminal plan.
The repetition still has historical significance.
A Conley confession was an extraordinarily valuable piece of evidence for Frank.
If genuine, it could transform the case.
For this reason, investigators had every legitimate reason to investigate credible reports of a confession.
The problem arises when witnesses say the confession never occurred.
At such a point, continued pursuit of the desired narrative becomes something different from ordinary investigation.
Allen said he did not know Conley.
Isom said Allen did not know Conley.
Yet, according to Isom, Burns said Allen could be made to “come across.”
Reese and Perkerson said they were asked to manufacture a confession.
Shields said a false accusation was proposed to him and, after he refused, another Negro witness was sought.
Ragsdale said one Negro corroborator “fell down” and another person was sought.
This recurring structure deserves attention.
The investigation did not always begin with a witness arriving spontaneously and reporting a story.
In several disputed episodes, the desired story appears first.
Then investigators search for a person capable of carrying it.
That is precisely where a search for evidence can become evidence construction.
The Affidavit Repudiation Cycle
Another pattern appears across the 1914 record.
An affidavit favorable to Frank surfaced.
The supposed witness later challenged it.
The defense then produced rebuttal affidavits defending the original paper.
Albert McKnight signed a statement useful to Frank.
Later he said a superior job prospect helped induce it and alleged words were inserted after his signature.
Burke denied improper inducement or alteration.
Minola McKnight and other defense witnesses supported Burke.
Ivey Jones became the source of a defense affidavit after an admitted deceptive interview.
Jones later repudiated the paper.
The defense said he had heard it read and signed voluntarily.
Mary Rich’s strongest alley story entered Frank’s motion.
Her subsequent deposition contained a narrower version.
She later denied the full story.
Rabbi Marx said she had confirmed it except for a time correction.
Carter supplied important evidence against Conley.
Her affidavit session developed a date issue between the April 23 sworn session record and the April 25 printed date.
She was then moved outside Georgia.
She later accused an investigator of giving her money and inducing testimony.
Defense witnesses said she had denied receiving a reward during the original session.
None of these episodes proves every defense affidavit was false.
The pattern reveals something more subtle.
Frank’s extraordinary motion increasingly depended upon papers whose validity had to be defended after the people named on those papers challenged how the statements were obtained, written, or understood.
A sound affidavit process should reduce ambiguity.
The Frank post-trial campaign repeatedly generated more ambiguity.
Who asked the questions?
Who wrote the paper?
Did the witness read it?
Could the witness read at all?
Was the statement signed?
When was it signed?
Was the witness offered employment?
Was money discussed?
Were words added?
Did the witness later recant?
Did the recantation arise through State pressure?
The litigation over the evidence-gathering process began to rival the litigation over the evidence itself.
Why the admitted conduct carries the most weight
Some accusations in this history will probably remain impossible to resolve conclusively.
Witnesses died more than a century ago.
Original documents are incomplete.
Affidavits conflict.
Newspaper accounts differ.
For this reason, the strongest argument should rest on what Frank’s own side admitted whenever possible.
Burke directed Wrenn to Ivey Jones.
Wrenn admitted the fictitious railroad-accident pretext.
Burke admitted preparing Jones’s affidavit from Wrenn’s notes.
Burns acknowledged an employee went from Chicago into Indiana to obtain Aaron Allen.
Isom acknowledged being paid while searching for Negro prospective witnesses and acknowledged receiving travel funds.
Hopkins acknowledged the Calhoun affidavit existed.
Burke acknowledged later trying to help Albert McKnight obtain employment while denying any job-for-affidavit arrangement.
George Wrenn acknowledged seeking a written statement from Frank Reese while disputing Reese’s explanation of its purpose.
Burns acknowledged Burke furnished him the Carter-Conley notes.
Burns said George Wrenn produced the first translation.
Carter was deliberately sent outside Georgia after giving evidence useful to Frank.
A judge found the removal technically contemptuous until she was returned.
These admissions establish much of the machinery.
One does not need to believe every allegation of bribery or forgery to see an unusually aggressive investigative process.
The remaining dispute concerns how far the machinery crossed from aggressive investigation into corrupt influence.
Witness tampering as a framework
The phrase witness tampering should not be used as a magic label capable of converting every questionable interview into a felony.
It is useful because it asks the correct questions.
Was somebody trying to discover what a witness honestly knew?
Or was somebody trying to change what the witness would say?
Was a job offered simply as kindness after cooperation?
Or was the job used as leverage for cooperation?
Was travel paid because a witness needed transportation?
Or was transportation part of controlling the witness?
Was money ordinary subsistence?
Or was money tied to a desired story?
Was an affidavit drafted from a witness’s own account?
Or did the affidavit become stronger than the witness’s account?
Was a false cover story used simply to avoid rehearsed answers?
Or was it used to manipulate a witness into generating evidence without informed awareness of the real investigation?
Was a confession discovered?
Or was somebody being asked to invent one?
Was a second witness sought because the first witness genuinely pointed investigators toward another source?
Or was a replacement witness being sought because the first refused the desired story?
These questions appear repeatedly in the Frank post-trial record.
Modern witness-tampering law places great weight on intent for exactly this reason.
An ordinary payment can become improper if corruptly tied to testimony.
An ordinary conversation can become improper if designed to intimidate.
An investigator can ask a witness to correct false testimony.
An investigator cannot properly ask a witness to replace truthful testimony with a lie.
The line is not the existence of contact.
The line is corrupt purpose.
This historical record cannot always prove purpose.
But several witnesses described it.
McKnight said the job was how Burke “got around” him.
Rich said she would not sign because she could not tell a lie.
Allen said the wanted affidavit would be false.
Reese and Perkerson said they were asked to claim Conley confessed.
Shields said he refused because the proposed accusation was false.
These statements are why “witness tampering” belongs in the historical discussion even when the article carefully stops short of declaring unproved crimes.
The Negro witnesses who refused
Perhaps the most important part of this entire history is not the investigators.
It is the witnesses who said no.
Albert McKnight eventually repudiated the affidavit Burke obtained from him.
Mary Rich said:
“I could not tell a lie.”
Aaron Allen denied knowing Conley and said the desired affidavit would be false.
Frank Reese could not read or write and refused to authorize a document until he understood it.
Fred Perkerson said he rejected the alleged confession scheme.
John Shields said:
“I refused every time because it was a lie.”
Jim Conley, according to Reese and Perkerson, refused to accept responsibility for Mary Phagan’s murder.
Ivey Jones repudiated the affidavit attributed to him.
Annie Maud Carter said she refused the alleged vial and promised money.
These Negro witnesses were not passive pieces on somebody else’s chessboard.
Their resistance created much of the documentary record we can study today.
If McKnight had remained silent, the Burke affidavit might have stood without public challenge.
If Rich had remained silent, the alley story might have stood without her repudiation.
If Allen had supplied the wanted confession story, investigators could have presented another witness claiming Conley confessed.
If Reese and Perkerson had accepted the alleged proposal, another Conley confession could have entered the record.
If Shields had signed the proposed accusation, Conley could have faced another damaging character claim.
Witness resistance is therefore not a side issue.
It is central.
The Negro men and women in this record repeatedly exercised independent judgment against powerful people attempting to shape a murder case.
Their own words deserve more attention than later theories about them.
The cumulative record
No single episode proves all the others.
McKnight changed stories.
His job allegation was denied.
His post-signature alteration accusation was never judicially established.
Rich and Marx supplied conflicting sworn accounts.
Jones’s forgery claim was denied.
Carter’s poison allegation rests heavily on Carter herself.
The State’s Carter-letter forgery accusation was never resolved as a judicial finding.
Reese and Perkerson were prisoners.
George Wrenn denied their accusations.
Calhoun did not accuse anyone of bribing him.
Wilson and Whatley left no surviving accusations.
Ragsdale was a deeply disputed witness.
The Ragsdale-related defendants were acquitted.
All of these weaknesses belong in the article.
Yet the cumulative record does not disappear after the weaknesses are acknowledged.
A Negro State witness said a job worth roughly three and a half times his usual monthly income accompanied an effort to obtain a changed affidavit.
The same witness alleged important language appeared after he had already signed.
Another Negro State witness was deliberately interviewed through a fictional railroad accident, with the deception admitted by the investigator.
The lead investigator then dictated the affidavit generated from notes produced through the deception.
A Negro woman’s strongest alley claim became weaker in her own deposition and was later repudiated.
A Negro man was pursued across state lines for a Conley-confession story he repeatedly denied.
An investigator inside the operation recorded Burns telling him the man could be made to “come across.”
The same investigator recorded Allen saying the affidavit would be a lie.
Two Negro jail trusties independently described an alleged plan to report a Conley confession which never occurred.
They said money or benefits were part of the proposal.
They also said Conley himself was approached.
An illiterate Negro trusty said a written paper was brought to his home for signature.
A Negro woman favorable to Frank was moved outside Georgia after giving her affidavit.
A judge found her removal technically contemptuous until she was returned.
The State formally accused people in the Frank investigative orbit of manufacturing Carter-Conley correspondence.
A witness said he was asked to find a Negro woman willing to provide testimony he himself had refused as false.
An unnamed Negro prospective corroborator was allegedly recruited to support another Conley-confession story, failed, and was followed by a search for somebody else.
One disputed confession-affidavit operation generated perjury and subornation indictments, followed by a criminal trial and acquittal.
The episodes differ.
The witnesses differ.
The evidentiary strength differs.
The repeated objective does not.
The defense needed evidence capable of destroying Jim Conley.
Again and again, investigators sought Negro witnesses whose words could do it.
Conclusion
The Negro witnesses belong near the center of the history of the Leo Frank case.
For more than a century, discussion of Negro participation has often narrowed itself to Jim Conley.
The surviving 1914 record is much larger.
Negro men and women became important because their words could alter Frank’s legal position.
They could place Conley somewhere else.
They could claim Conley confessed.
They could damage his character.
They could weaken previous testimony.
They could support an extraordinary motion for a new trial.
They could shift responsibility for Mary Phagan’s murder away from Frank.
Frank’s attorneys had every right to investigate the conviction.
They had every right to find legitimate new witnesses.
They had every right to challenge Conley.
They had every right to correct false testimony.
They had every right to pay reasonable investigative expenses.
They had every right to pursue a genuine confession.
Nothing in this article argues otherwise.
The problem begins where investigation becomes manipulation.
The admitted false pretext used against Ivey Jones belongs in this history.
Burke’s acknowledged preparation of Jones’s affidavit from notes produced through the deceptive interview belongs in this history.
McKnight’s detailed allegation involving employment and post-signature alteration belongs in this history, alongside Burke’s denial.
Mary Rich’s progressively narrowing account belongs in this history, alongside Marx’s rebuttal.
The interstate pursuit of Aaron Allen belongs in this history.
Isom’s statement claiming Allen did not know Conley belongs in this history.
Burns’s alleged instruction to make Allen “come across” belongs in this history.
Allen’s statement claiming the wanted affidavit would be a lie belongs in this history.
The paid interstate search for Calhoun, Wilson, and Whatley belongs in this history.
Reese and Perkerson’s parallel allegations concerning a fabricated Conley confession belong in this history.
Wrenn’s denial belongs beside them.
The alleged approach to Conley himself belongs in this history.
Reese’s refusal to sign a document he could not read belongs in this history.
The disputed Carter correspondence belongs in this history.
The State’s forgery accusation belongs in this history.
The defense handwriting evidence belongs beside it.
Carter’s removal from Georgia belongs in this history.
Judge Hill’s technical-contempt finding and subsequent dismissal after her return belong beside it.
The unnamed Negro corroborator in the Ragsdale affair belongs in this history.
The Ragsdale indictments belong in this history.
The 1915 acquittal belongs beside them.
None of these facts requires claiming every member of Frank’s defense committed a crime.
None requires claiming every accusation was true.
The incriminating feature is cumulative.
A post-trial investigation repeatedly found itself accused by Negro witnesses of trying to obtain statements they considered false, while some of the investigative machinery surrounding those accusations was admitted by the defense itself.
The strongest question is therefore not:
Were all of Frank’s investigators criminals?
The sources cannot support such a sweeping conclusion.
The stronger question is:
How often did the search for evidence favorable to Leo Frank move from discovering what Negro witnesses knew toward trying to produce what the defense needed them to say?
The surviving record gives enough examples to make the question unavoidable.
The term witness tampering belongs in this research because several episodes involve precisely the types of conduct modern law treats with suspicion: misleading approaches, alleged economic inducements, corrupt-persuasion allegations, attempts to obtain false statements, efforts to create confession testimony, witness removal from a jurisdiction, and disputed alteration or manufacture of evidence.
Historical discipline requires another sentence immediately afterward.
Not every allegation was proved.
The difference between accusation and judgment must remain visible.
But historical discipline does not require pretending the accusations never happened, pretending admitted deception was ordinary, or pretending a large group of Negro witnesses independently raising complaints about how affidavits and confession stories were being pursued deserves no scrutiny.
Some of those witnesses changed their stories.
Some contradicted themselves.
Some were prisoners.
Some were poor.
Some were illiterate.
Some faced pressure from the State as well as Frank’s investigators.
None of those facts makes their voices disposable.
In several episodes, the Negro witness’s refusal is the strongest surviving evidence.
“I could not tell a lie.”
“I will make that affidavit, but it will be a lie.”
“I refused every time because it was a lie.”
Those are not the words of passive people being carried along by famous lawyers and detectives.
They are the words of witnesses drawing a line.
Frank’s investigators needed evidence against Jim Conley.
Some Negro witnesses gave them useful evidence.
Some did not.
Some later repudiated it.
Some refused from the beginning.
And when the entire post-trial record is read together, one fact stands out.
The effort to overturn Leo Frank’s conviction repeatedly depended upon Negro men and women whose words could be turned against Jim Conley, and several of those people said the words being sought from them were not true.
That history deserves to be read in full.
References
- Georgia Code §16-10-93. “Influencing Witnesses.” Current Georgia statute.
http://www.law.justia.com/codes/georgia/title-16/chapter-10/article-5/section-16-10-93/ - Georgia Code §16-10-72. “Subornation of Perjury or False Swearing.” Statutory history includes Penal Code 1895 §§260–261 and Penal Code 1910 §§263–264.
http://www.law.justia.com/codes/georgia/title-16/chapter-10/article-4/section-16-10-72/ - The Atlanta Constitution. Tuesday, June 3, 1913. “Leo Frank’s Cook Put Under Arrest.”
- Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. State’s Exhibit J, affidavit of Minola McKnight, p. 1288 and following.
http://www.maryphagan.com/1288-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Minola McKnight trial testimony repudiating State’s Exhibit J, p. 1141.
http://www.maryphagan.com/1141-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Constitution. Sunday, April 19, 1914. “Testimony He Gave at Trial Was True, Declares M’Knight,” p. 1, col. 6.
http://www.maryphagan.com/1914-04-19-testimony-he-gave-at-trial-was-true-declares-m-knight-the-atlanta-constitution/ - The Atlanta Journal. Tuesday, May 5, 1914. “Many Sensations Sprung by Dorsey at Frank Hearing,” p. 1, col. 7, continuing on later pages.
http://www.maryphagan.com/1914-05-05-many-sensations-sprung-by-dorsey-at-frank-hearing-the-atlanta-journal/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Albert McKnight sworn statement, p. 1600.
http://www.maryphagan.com/1600-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Journal. Sunday, April 19, 1914. “Albert M’Knight Now Repudiates His Recent Story,” p. 1, col. 6.
http://www.maryphagan.com/1914-04-19-albert-m-knight-now-repudiates-his-recent-story-the-atlanta-journal/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Affidavits of C. W. Burke and Minola McKnight concerning Albert McKnight, p. 1741.
http://www.maryphagan.com/1741-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Georgian. Friday, March 27, 1914. “3 Swear to Alibi for Frank,” p. 6.
http://www.maryphagan.com/1914-03-27-3-swear-to-alibi-for-frank-the-atlanta-georgian/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. C. W. Burke and J. W. Wrenn affidavits concerning Ivey Jones, p. 1761.
http://www.maryphagan.com/1761-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Journal. Friday, March 27, 1914. “Frank’s Motion Is Made Public: Many of State’s Witnesses Change Their Testimony,” p. 1, col. 2.
http://www.maryphagan.com/1914-03-27-frank-s-motion-is-made-public-many-of-state-s-witnesses-change-their-testimony-the-atlanta-journal/ - The Atlanta Journal. Wednesday, April 22, 1914. “Frank Hearings Postponed; Burns Says Conley Guilty.”
http://www.maryphagan.com/1914-04-22-frank-hearings-postponed-burns-says-conley-guilty-the-atlanta-journal/ - The Atlanta Constitution. Tuesday, May 5, 1914. “Return of Negress Ordered by Judge Monday Morning,” p. 1, col. 5.
http://www.maryphagan.com/1914-05-05-return-of-negress-ordered-by-judge-monday-morning-the-atlanta-constitution/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Mary Rich and E. H. Wellborn affidavits, p. 1627.
http://www.maryphagan.com/1627-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Journal. Wednesday, May 6, 1914. “Judge Hill Denies Motion as Defense Closes Arguments,” pp. 1, 3.
http://www.maryphagan.com/1914-05-06-judge-hill-denies-motion-as-defense-closes-arguments-the-atlanta-journal/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Affidavit of Rabbi David Marx, Isaac Haas, Isaac Schoen, B. Wildauer, J. O. Knight, and Otto Schwab concerning the Annie Maud Carter affidavit session, p. 1764.
http://www.maryphagan.com/1764-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Annie Maud Carter affidavit continuation showing the April 25 printed date, p. 1765.
http://www.maryphagan.com/1765-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Constitution. Sunday, April 26, 1914. “Says Love Letters Written by Conley Prove Him Guilty,” p. 1, col. 2.
http://www.maryphagan.com/1914-04-26-says-love-letters-written-by-conley-prove-him-guilty-the-atlanta-constitution/ - The Atlanta Georgian. Sunday, April 26, 1914. “New Conley Notes Prove Him Guilty, Says Burns,” 4th ed., p. 1, col. 1.
http://www.maryphagan.com/1914-04-26-new-conley-notes-prove-him-guilty-says-burns-the-atlanta-georgian/ - The Atlanta Georgian. Friday, May 1, 1914. “Evidence Is Recorded Against Conley,” 3rd ed., p. 1, col. 8.
http://www.maryphagan.com/1914-05-01-evidence-is-recorded-against-conley-the-atlanta-georgian/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. State counter-showing concerning the Carter-Conley letters and alleged manufacture of the notes, p. 1495.
http://www.maryphagan.com/1495-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Frank Reese affidavit, p. 1661.
http://www.maryphagan.com/1661-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Willis J. Gilbert affidavit and jail testimony, pp. 1663–1664.
http://www.maryphagan.com/1663-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/
http://www.maryphagan.com/1664-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Georgian. Saturday, May 2, 1914. “Detective on Stand, Tells of Probe,” 3rd ed., p. 1, col. 8.
http://www.maryphagan.com/1914-05-02-detective-on-stand-tells-of-probe-the-atlanta-georgian/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Aaron Allen sworn statement, p. 1674.
http://www.maryphagan.com/1674-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. C. A. Isom affidavit concerning Aaron Allen, p. 1732.
http://www.maryphagan.com/1732-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. C. A. Isom affidavit concerning C. W. Burke’s instructions, Aaron Allen, William Calhoun, Mark Wilson, Whatley, investigator wages, and Chicago expenses, p. 1731.
http://www.maryphagan.com/1731-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. C. A. Isom affidavit concerning William Burns and William Calhoun, p. 1733.
http://www.maryphagan.com/1733-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Fred Perkerson affidavit concerning George Wrenn, p. 1724.
http://www.maryphagan.com/1724-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Fred Perkerson affidavit concerning George Wrenn and Jim Conley, p. 1725.
http://www.maryphagan.com/1725-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. Frank Reese affidavit continued, p. 1662.
http://www.maryphagan.com/1662-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. George Wrenn identified as Jimmie Wrenn’s brother and Jimmie Wrenn’s connection with C. W. Burke, p. 1723.
http://www.maryphagan.com/1723-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Journal. Saturday, May 9, 1914. “Burns and Lehon Are Summoned on Contempt Charge,” p. 1, col. 7.
http://www.maryphagan.com/1914-05-09-burns-and-lehon-are-summoned-on-contempt-charge-the-atlanta-journal/ - The Atlanta Journal. Tuesday, May 26, 1914. “Burns’ Men Fight Effort to Shut Up Offices in Atlanta,” p. 1, col. 1.
http://www.maryphagan.com/1914-05-26-burns-men-fight-effort-to-shut-up-offices-in-atlanta-the-atlanta-journal/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. John Shields affidavit, p. 1666.
http://www.maryphagan.com/1666-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - Georgia Supreme Court Case File. Leo M. Frank v. State of Georgia. John Shields affidavit continued, p. 1667.
http://www.maryphagan.com/1667-slide-georgia-supreme-court-leo-frank-appeals-documents-1913-1914/ - The Atlanta Journal. Friday, May 22, 1914. “Five Men Indicted on Perjury Charge in Leo Frank Case,” p. 1, col. 3.
http://www.maryphagan.com/1914-05-22-five-men-indicted-on-perjury-charge-in-leo-frank-case-the-atlanta-journal/ - The Atlanta Journal. Tuesday, January 26, 1915. “Trial of Burns Men On Before Judge Hill,” p. 2, col. 3.
http://www.maryphagan.com/1915-01-26-trial-of-burns-men-on-before-judge-hill-the-atlanta-journal/ - The Atlanta Journal. Friday, January 29, 1915. “Ragsdale Weak in Mind, Son Admits on Stand,” p. 1, col. 7.
http://www.maryphagan.com/1915-01-29-ragsdale-weak-in-mind-son-admits-on-stand-the-atlanta-journal/ - The Atlanta Constitution. Sunday, January 31, 1915. “Fate of Burns’ Agents Now in Hands of Jury; No Verdict Until Today,” p. 1, col. 1.
http://www.maryphagan.com/1915-01-31-fate-of-burns-agents-now-in-hands-of-jury-s-no-verdict-until-today-the-atlanta-constitution/ - The Atlanta Constitution. Monday, February 1, 1915. “Dan Lehon, Arthur Thurman and C. C. Tedder Are Given Their Freedom After Long Trial,” p. 1, col. 1.
http://www.maryphagan.com/1915-02-01-dan-lehon-arthur-thurman-and-c-c-tedder-are-given-their-freedom-after-long-trial-the-atlanta-constitution/