That Accusation was Rejected by the Trial Judge. The Georgia Supreme Court Upheld the Judge’s Decision.

Atticus Haygood Henslee and Marcellus Johenning: The Post-Verdict Affidavit Campaign, the Defense Investigation, the Credibility of the Accusers, Judge Leonard S. Roan’s Decision, and the Georgia Supreme Court’s Rejection of the Bias Claims

Revised and expanded October 15, 2026

By Brian Nelson

The Defense Went After the Jury

On Monday, August 25, 1913, twelve men found Leo M. Frank guilty of murdering thirteen-year-old Mary Phagan at the National Pencil Company in Atlanta. The following day, Judge Leonard S. Roan sentenced Frank to death. Within weeks, Frank’s attorneys were building another case, this time challenging the qualifications of two men who had helped convict him.

The jurors were Atticus Haygood Henslee, a thirty-six-year-old traveling salesman, and Marcellus Johenning, a forty-six-year-old industrial employee. Frank’s lawyers claimed both men had expressed firm opinions of his guilt before hearing the trial evidence. Some witnesses attributed violent and anti-Jewish statements to Henslee, while three members of one Atlanta family accused Johenning of declaring Frank guilty months before the trial began.

Both jurors denied entering the jury box prejudiced. Henslee acknowledged discussing the murder in some of the places identified by his accusers, but disputed the damaging words or the dates assigned to them. Johenning admitted speaking with two of his accusers while insisting they misunderstood remarks about unfavorable newspaper coverage.

The defense investigation developed into a substantial legal and newspaper controversy. It involved traveling salesmen, merchants, hotel clerks, business orders, expense records, reluctant witnesses, sworn affidavits, depositions, and investigators tracing Henslee’s travels across Georgia. The State answered with the jurors’ denials, character evidence, documentary challenges, and affidavits from men who had served beside Henslee and Johenning for nearly a month.

Judge Roan understood how serious the accusations could become. During the new-trial hearing on Thursday, October 23, 1913, he indicated that proof of concealed juror prejudice could make further argument nearly unnecessary. After examining the competing material, he denied Frank another trial on Friday, October 31.

The Georgia Supreme Court reviewed the attacks upon the two jurors individually on Tuesday, February 17, 1914. Regarding Henslee, the Court acknowledged evidence tending to show strong prejudice and also acknowledged his sworn denial. It concluded that Roan had not abused his discretion in resolving the conflicting evidence. The Court then upheld his determination that Johenning was not prejudiced or biased (Frank v. State, 1).

Those rulings provide the legal foundation for this study. The accusations were genuine, some were serious, and they deserved careful examination. Their existence did not automatically establish their truth. Understanding why the courts rejected the juror-bias grounds requires examining the affidavits, how the witnesses were found, the accuracy of their dates, the independent evidence supporting them, and the testimony offered in response.

The Question the Courts Had to Decide

The question was whether Henslee and Johenning could serve as fair and impartial jurors.

Atlanta had been saturated with coverage of Mary Phagan’s murder since Saturday, April 26, 1913. Newspapers described the investigation daily, and people discussed the case in offices, stores, hotels, streetcars, and homes. A prospective juror could hardly have avoided hearing about it.

Having read newspaper reports did not necessarily establish disqualifying prejudice. A person might believe the evidence looked bad for Frank and still be willing to hear both sides. A person who had already decided to convict regardless of the testimony presented a much more serious problem.

Frank’s attorneys claimed Henslee and Johenning belonged in the second category. The jurors maintained that they had qualified honestly and reached their verdict from the evidence heard during trial.

The factual disagreement turned upon remembered conversations, the meaning of particular words, the reliability of witnesses, and whether the jurors had formed opinions so firmly that they could no longer judge the case impartially.

Georgia law placed responsibility for resolving this post-verdict factual dispute with the trial judge. Roan could receive affidavits, examine contradictory statements, consider reputation evidence, and decide whether the defense had established grounds requiring another trial. The Georgia Supreme Court then reviewed his decision to determine whether he had abused his discretion (Frank v. State, 1).

This difference in judicial responsibility becomes central to understanding the final ruling. An affidavit could establish that a serious accusation had been sworn, but the oath did not independently verify every word. Likewise, a juror’s denial did not automatically defeat the accusation. Roan had to evaluate the conflict.

Who Was Atticus Haygood Henslee?

Atticus Haygood Henslee was born on Sunday, June 3, 1877, and died on Thursday, June 14, 1956, according to Lamar County genealogical records. Those records associate him with Greenwood Cemetery in Barnesville, Georgia, and identify his wife as Elizabeth May Franklin Henslee. A Lamar County marriage index records Atticus Henslee and May Franklin marrying in 1900 (Lamar County Records, 22).

At the beginning of Frank’s trial, Henslee was thirty-six years old, married, and head salesman for the Franklin Buggy Company of Barnesville. Contemporary jury descriptions placed his Atlanta residence at 74 Oak Street in the West End neighborhood. The 1912 Atlanta directory independently confirms that address, while the 1914 directory describes him as a traveling salesman who had moved to Barnesville (The Atlanta Constitution, 21) (The Atlanta Journal, 20) (Atlanta City Directory, 23).

Henslee’s occupation explains much of the later controversy. He traveled around Georgia selling buggies and related equipment, visiting merchants, riding trains, staying in hotels, and keeping business accounts. Like many people during the spring and summer of 1913, he discussed the Phagan murder. His travels left behind both recollections and paperwork, allowing Frank’s investigators to follow his route and find people who remembered conversations.

That is why the juror investigation eventually reached Albany, Sparta, Monroe, Blakely, Atlanta, and other communities. The same business trail that helped the defense locate witnesses also produced records that could test their claims. Hotel registrations, signed orders, and expense entries became important because they could establish where Henslee had been, even when they could not establish what he said.

Henslee’s Jury Summons

Henslee swore that he first learned he had been summoned for jury service at approximately five in the afternoon on Friday, July 25, 1913. He said the notice was at his residence, 74 Oak Street, and that he remained in Atlanta through the weekend before reporting to court on Monday, July 28 (Henslee Affidavit, 4).

This date became important because witnesses in Sparta later claimed Henslee had referred to being summoned or drawn as a juror during a conversation they placed on June 27, nearly a month earlier.

The original scan also demonstrates the danger of relying on automated transcriptions. One modern transcription of Henslee’s affidavit incorrectly gives a different address and date. The reproduced image identifies 74 Oak Street, while the surrounding chronology supports the July 25 notice. The scanned document should control where its text can be read clearly.

Henslee the Ventriloquist

Some contemporary newspapers described Henslee as the ventriloquist on the Frank jury. The description originated in a humorous Atlanta Journal account published Sunday, August 10, 1913.

According to the newspaper, Henslee entertained the sequestered jurors by throwing his voice around their quarters. Bailiff Charles F. Huber reportedly believed he heard a woman speaking nearby and searched for the source, eventually discovering that Henslee had produced the voice (The Atlanta Journal, 11).

The episode has no bearing on the allegation of prejudice, but it helps explain an unusual nickname appearing in later coverage. It also offers a glimpse of the ordinary people serving on the jury during their lengthy confinement.

Henslee’s Later Life

Henslee remained connected with the buggy business after the trial. In 1916 he ran for the Georgia Prison Commission, and Thomas E. Watson’s Jeffersonian used his Frank jury service in political publicity (The Atlanta Georgian, 43) (The Jeffersonian, 44).

His later political activity demonstrates how the Frank controversy continued to shape public life. It does not establish what he believed before serving on the jury in July 1913.

Who Was Marcellus Johenning?

Marcellus Johenning was approximately forty-six years old when Frank’s trial opened. Contemporary jury profiles described him as married, employed in shipping or factory work, and living at 161 Jones Avenue in Atlanta (The Atlanta Constitution, 21) (The Atlanta Journal, 20).

The Atlanta city directories provide a fuller picture. The 1912 directory identifies Marcellus Johenning, with Annie named in the household, as a foreman at 271 Marietta Street. The 1914 directory connects him with Hodges Broom Works at the same residence. The 1916 directory uses Alfred M. Johenning, while the 1919 directory uses A. Marcellus Johenning and describes him as a timekeeper (Atlanta City Directory, 23).

The repeated address and employer strongly suggest that the juror’s full name may have been Alfred Marcellus Johenning. A direct birth, marriage, census, or death record would provide firmer confirmation. Until such evidence is identified, Marcellus Johenning remains the safest published form.

At age forty-six in July 1913, Johenning was probably born in 1866 or 1867. An exact birth or death date has not yet been securely established.

The Hodges Broom Works Connection

Johenning’s employment became relevant because one of his later accusers, H. C. Loevenhart, said he knew Johenning through Hodges Broom Works.

The directory evidence supports the claimed workplace connection. That strengthens the possibility that the men knew one another, although it does not confirm the alleged conversation about Frank’s guilt.

The difference matters. A directory can identify an address, occupation, or employer, but it cannot tell us what one person said to another several months earlier. Similar limitations apply throughout the Henslee and Johenning investigation.

Johenning Was Observed During Trial

Johenning appears directly in contemporary reporting from Frank’s trial. On Friday, August 1, 1913, The Atlanta Journal described him asking for further explanation of the National Pencil Company time-clock system during N. V. Darley’s testimony.

Darley approached the jury box and explained the punches and gaps on the factory time slips more fully. This demonstrated that Johenning was actively seeking clarification of evidence rather than remaining entirely passive during the proceedings (The Atlanta Journal, 24).

A juror who possessed an earlier opinion could still ask questions, so the incident cannot prove impartiality by itself. It nevertheless belongs in an assessment of how Johenning behaved while hearing the case.

Another contemporary account described Johenning displaying emotion during Frank’s lengthy unsworn statement. That observation is also limited, since emotion does not establish a juror’s private judgment.

Tarleton Collier’s August 24 Portrait

On Sunday, August 24, 1913, one day before the verdict, journalist Tarleton Collier published a personality study of the twelve jurors in Hearst’s Sunday American. The article was titled “Frank Jurors an Enthralling Study in Human Emotion.”

Collier portrayed Johenning as “plainly indecisive and nervous,” apparently responding to evidence from both sides and weighing conflicting testimony (Hearst’s Sunday American, 25).

This was a reporter’s impression rather than sworn evidence. Collier could not know Johenning’s private thoughts from facial expressions. The timing nevertheless gives the account historical interest because it was published before the verdict, when no post-verdict affidavit campaign was underway.

A reporter studying Johenning during trial did not portray him as visibly demonstrating a judgment fixed months earlier. That observation provides context for the later accusations without settling their truth.

The Defense Accepted Both Jurors During Selection

Henslee and Johenning qualified for jury service when Frank’s trial opened Monday, July 28, 1913.

Dorsey later emphasized the substantial investigative resources available to Rosser and Arnold and questioned why the defense had accepted Henslee if damaging information existed about him. He also noted that the defense still possessed unused challenges when Henslee was accepted (The Atlanta Journal, 15).

The defense answered through sworn statements from Frank and his attorneys.

Luther Z. Rosser, Morris Brandon, Reuben R. Arnold, Herbert J. Haas, and Frank himself maintained that they had not known of the alleged disqualifying statements during jury selection or trial. They claimed the information was discovered afterward and would have been presented immediately had they known (Motion for New Trial, 5).

The affidavits addressed an important legal question: why the defense had not challenged the men earlier. Frank’s lawyers maintained that they had investigated after the verdict and acted promptly upon learning the allegations.

Those assertions concerning diligence did not independently prove the alleged conversations. They explained the timing of the challenge and the defense’s claim that the information was newly discovered.

Monday, August 25, 1913: The Doubtful Ballot

One of the most interesting parts of the controversy concerns the jury’s voting after Judge Roan completed his instructions.

The newspapers disagreed immediately after the verdict.

The Atlanta Constitution reported that only one ballot had been taken. The Atlanta Georgian described Frank as guilty on the first formal ballot. The Atlanta Journal, however, reported an initial vote of ten for conviction, one blank, and one doubtful, followed approximately one hour later by unanimity (The Atlanta Constitution, 18) (The Atlanta Georgian, 19) (The Atlanta Journal, 17).

This conflict existed on Tuesday, August 26, the morning after the verdict. The doubtful-ballot story was therefore not first created in October to answer allegations against Henslee.

Henslee Said He Was the Doubtful Voter

During the new-trial proceedings, Henslee identified himself as the man who cast the doubtful ballot. He said the other jurors who had expressed opinions were favoring guilt and that he wanted the evidence discussed thoroughly because the verdict could result in Frank’s execution.

Foreman F. E. Winburn and other jurors supported significant portions of this account. F. V. L. Smith also described Henslee expressing caution concerning the strength of evidence necessary for conviction while withholding his final opinion during the trial (Georgia Supreme Court New-Trial Record, 6).

The later affidavits do not perfectly match the Journal’s original report. The precise voting mechanics remain uncertain, although there is contemporary support for an initial expression of doubt followed by a unanimous verdict.

Why the Ballot Helps Henslee

The doubtful ballot creates difficulty for the strongest version of Frank’s defense theory.

Reuben Arnold eventually portrayed Henslee as a man who entered the jury with a concealed determination to convict. Yet according to Henslee and his fellow jurors, his doubtful ballot delayed unanimity and required further discussion.

If the other eleven men already favored guilt, a guilty vote from Henslee would have produced an immediate unanimous decision. His claimed doubtful vote had the opposite effect.

This does not prove Henslee never made prejudicial remarks in May or June. Someone can speak confidently in casual conversation and become more cautious when actually deciding another person’s fate.

The ballot nevertheless makes the simplest secret-plot theory more difficult to sustain.

Rosser Called the Ballot a Trick

During closing arguments on Wednesday, October 29, Luther Rosser attacked Henslee’s explanation. The Atlanta Journal reported Rosser suggesting the doubtful ballot was a trick upon the other jurors and calling it a “miserable, dirty farce” (The Atlanta Journal, 16).

The accusation required evidence explaining how the supposed trick worked. No fellow juror described Henslee admitting a secret plan, and no affidavit established that he pretended doubt to manipulate the others.

Rosser could fairly argue that the doubtful ballot did not erase earlier statements attributed to Henslee. His stronger claim that the ballot itself formed part of a concealed scheme remained unproved.

The timing also raises a problem with any theory of advance self-protection. Henslee cast the ballot on August 25, while the extensive investigation of his prior conversations developed afterward. The record does not establish that he knew he would become one of the principal targets of Frank’s amended motion.

The Defense Begins Hunting for Juror Evidence

Frank’s attorneys conducted an extensive post-verdict investigation.

They reconstructed Henslee’s business travels, visited communities where he sold buggies, questioned merchants and acquaintances, obtained affidavits, and used legal procedures when witnesses resisted involvement.

Stiles Hopkins assisted with depositions. C. W. Burke was publicly identified as an agent working for Rosser, and contemporary newspapers described investigators searching towns along Henslee’s sales territory (The Atlanta Georgian, 12).

There was nothing inherently improper about examining possible juror misconduct. Frank faced execution, and his lawyers had every reason to investigate credible grounds for another trial.

The reliability of the resulting affidavits still required examination.

The investigation was directed toward finding statements useful to Frank’s defense. It was not a survey of every person who had ever spoken with Henslee, including those who remembered neutral comments or no comments at all.

This creates a limitation in treating the affidavit total as proof. A targeted investigation may uncover truthful evidence while also collecting one side of a man’s conversational history.

Dorsey’s “Fine Tooth Comb” Argument

Dorsey attacked the investigative method directly during his October 28 presentation.

He described Rosser and Arnold as possessing unusually extensive resources for investigating prospective jurors and argued that the defense had searched Georgia with a “fine tooth comb” for anything damaging to a juror (The Atlanta Journal, 15).

Dorsey was arguing for the State, but his criticism was consistent with contemporary descriptions of investigators tracing Henslee’s route.

His point was that the number of affidavits could not substitute for testing their accuracy. The judge needed to consider dates, identification, corroboration, witness reputation, and the circumstances in which testimony had been obtained.

Dorsey also questioned conversations allegedly heard by several people when not all those listeners had provided affidavits. Missing corroborators did not prove the first witness false, but their absence limited the independent confirmation available.

Newspaper Evidence of Pressure on Reluctant Witnesses

The newspaper record contains direct evidence showing that some defense witnesses did not eagerly volunteer their information.

C. P. Stough provides the clearest example. According to The Atlanta Georgian of Friday, October 3, 1913, information concerning his alleged conversation with Henslee reached the defense through another person. Stough said two men working for Rosser approached him repeatedly for approximately two weeks while he declined to give information (The Atlanta Georgian, 7).

He eventually received a court summons requiring testimony before a commissioner.

This is evidence of pressure to participate. It does not establish that Stough was pressured to lie.

The Sparta Witnesses Were Also Reluctant

Shi Gray, the Johnson witness, and John M. Holmes similarly described reluctance.

In a joint public letter published Monday, October 6, the men explained that some had known Henslee and his family for years and did not want to injure him. They acknowledged declining voluntary affidavits and eventually gave testimony through formal procedures (The Atlanta Constitution, 9).

Their reluctance works against portraying them as eager volunteers seeking publicity. It does not prove their dates or remembered words were accurate.

The Blakely Witnesses Were Compelled and Then Hurt the Defense

A particularly revealing episode occurred in Blakely, Georgia.

A rumor claimed Henslee had said he would remain on the jury until Christmas rather than release Frank. The defense pursued L. E. Black and Walter Thomas, who resisted voluntary involvement and were eventually required to answer through legal process.

Their testimony did not support the damaging rumor (The Atlanta Constitution, 8).

Both acknowledged discussing newspaper accounts with Henslee, but neither confirmed the alleged declaration of fixed prejudice.

The episode demonstrates that an investigative lead could sound powerful and then collapse when the supposed witnesses were formally questioned. It also shows compulsory process did not automatically produce testimony favorable to Frank.

The defense followed damaging leads aggressively, but some failed.

The Affidavit Campaign Was Built Over Several Weeks

The dates of the affidavits show how the defense case expanded after the verdict.

H. C. Loevenhart and Miriam Loevenhart executed their affidavits against Johenning on Tuesday, September 2.

R. L. Gremer and Mack Farkas executed their Albany affidavits on Thursday, September 4, before L. L. Ford of Dougherty County.

Julian A. Lehman executed his first affidavit in September.

The Sparta depositions were taken Friday, September 26, before J. W. Lewis.

J. J. Nunnally and Dr. W. L. Ricker gave their Monroe testimony on Saturday, September 27.

C. P. Stough’s testimony was taken Monday, September 29, before a commissioner identified in contemporary records as Teitlebaum.

Samuel Aron executed his affidavit Friday, October 3, and Leon Harrison’s affidavit followed Wednesday, October 8.

Nunnally and Ricker reaffirmed their accusations October 10, while Lehman executed his supplemental affidavit Monday, October 13 (Motion for New Trial, 5).

This was an expanding evidentiary file rather than a completed set of statements collected at the same time.

An Important Correction to the Loevenhart Dates

The original exhibits establish that H. C. and Miriam Loevenhart signed their affidavits on September 2, while Jennie G. Loevenhart signed on September 26.

All three were notarized by C. W. Burke (Loevenhart Affidavits, 2).

The later date of Jennie’s affidavit should remain visible. The surviving documents do not establish why her statement was executed twenty-four days after those of the other two family members.

This correction is important because earlier consolidated versions assigned September 2 to all three.

How Many People Accused Henslee?

Contemporary newspapers gave changing figures.

Early October reporting identified eight principal accusers. On October 23, The Atlanta Georgian referred to more than twenty affidavits connected with the juror attack. Later reporting on October 30 identified eleven direct Henslee accusers and said the State had attempted to impeach three (The Atlanta Georgian, 12) (The Atlanta Georgian, 38).

The safest reconstruction identifies eleven principal direct accusers: R. L. Gremer, Mack Farkas, Julian A. Lehman, Samuel Aron, C. P. Stough, Shi Gray, the Johnson witness, John M. Holmes, J. J. Nunnally, Dr. W. L. Ricker, and Leon Harrison.

The larger affidavit figure may have included supplemental statements, character evidence, expected papers, or overlapping documents. The surviving newspaper report does not explain the difference conclusively.

The raw number also requires care. Gray, Johnson, and Holmes described one Sparta discussion. Nunnally and Ricker described one Monroe encounter. Gremer and Farkas described a shared Albany setting.

Multiple witnesses to one event can provide corroboration, but they cannot transform one conversation into several independent events.

The Albany Accusations: Gremer and Farkas

Gremer and Mack Farkas accused Henslee of expressing belief in Frank’s guilt while visiting Albany.

Gremer described an emphatic statement but could not supply the exact date. His later identification involved newspaper photographs of the jury. Mack Farkas described the same general setting but acknowledged uncertainty about the exact wording (Motion for New Trial, 5).

These accounts supported the possibility that Henslee had discussed Frank unfavorably while conducting business.

They were not equally precise. Farkas could not independently authenticate every word Gremer attributed to Henslee.

The July 8 Records Contradicted Henslee

The defense’s strongest documentary discovery concerned Henslee’s statement that he had not visited Albany between an early-June trip and September 18.

A Franklin Buggy Company order dated Tuesday, July 8, 1913, identified Henslee as salesman. Mack Farkas, Sam Farkas, and B. W. Simon authenticated the business transaction.

W. W. Little, head clerk of the New Albany Hotel, also identified Henslee’s registration for July 8 and his assignment to room 79 (Motion for New Trial, 5).

These records showed Henslee’s broad travel chronology was inaccurate.

That was an important weakness in his sworn account, and Frank’s attorneys were entitled to use it against his credibility.

What the Albany Documents Could Prove

The business order established a transaction associated with Henslee on July 8. The hotel register established his presence in Albany.

Neither document recorded what he said about Frank.

The records strengthened the opportunity for an alleged conversation and contradicted Henslee’s recollection of his travels. They did not independently establish Gremer’s strongest wording or prove a concealed intention to convict.

Roan could accept the Albany contradiction while remaining unconvinced that the alleged prejudicial statements had been established.

The Buggy Order and Frank Hooper’s Argument

The original order recorded four buggies at $62.50 each and two rubber sets at $15 each, totaling $280.

During argument, State attorney Frank Hooper was reported as referring to eight buggies (The Atlanta Journal, 15).

The original exhibit should control the quantity.

Hooper argued that a salesman trying to obtain business from a Jewish customer would have little reason to insult him through hostile remarks about Frank.

This was an argument about probability rather than proof. A business transaction cannot establish the absence of offensive language, and the Albany statements should not be combined with the separate Elks Club accusation to create one imagined conversation.

Julian Lehman and the June 2 Travel Evidence

Julian A. Lehman alleged Henslee made strongly prejudicial statements during travel encounters. His original account associated one encounter with Monday, June 2, 1913, and a route involving Atlanta and Experiment, Georgia.

Henslee denied the accusation.

Joe Murray, a clerk at the New Albany Hotel, swore that Henslee registered there for lodging on June 2 and spent the night in Albany. Henslee’s expense material also placed him in southwestern Georgia (Georgia Supreme Court New-Trial Record, 6).

This created a significant problem for Lehman’s original date and route.

An overnight registration does not establish where Henslee spent every earlier hour. A complete all-day alibi would require the alleged time of the encounter, railroad schedules, and realistic travel arrangements.

The combined evidence nevertheless weakened the precision of Lehman’s original account.

Lehman Becomes Less Certain

Lehman executed a supplemental affidavit on Monday, October 13.

He continued to allege two prejudicial conversations but became less definite about their dates. One was placed near June 20, while the other was described as occurring early in June, near June 2, without certainty about the exact day (Lehman Supplemental Affidavit, 3).

This does not establish dishonesty.

It does weaken the apparent precision of the first affidavit because an allegation tied firmly to June 2 can be checked against that day’s itinerary, while a general early-June allegation is harder to test.

The travel evidence therefore cut both ways. July 8 damaged Henslee’s own chronology, while June 2 weakened Lehman’s original date.

Henslee Confronts Lehman

The dispute also became personal.

The Atlanta Georgian of Tuesday, October 7 reported that Henslee had confronted Lehman and threatened legal action over the accusation that made Henslee appear dishonest (The Atlanta Georgian, 11).

Lehman did not withdraw his allegation.

This episode is valuable because pressure existed on more than one side. Defense investigators pursued witnesses, while Henslee confronted an accuser. Neither event independently established the truth of the disputed conversation.

Samuel Aron and the Elks Club Accusation

Samuel Aron made the most inflammatory individual accusation against Henslee.

He alleged that shortly after Frank’s indictment, Henslee expressed satisfaction with the indictment in explicitly anti-Jewish terms while at the Elks Club. Aron also alleged Henslee supported lynching and said he would vote to hang Frank if selected for the jury (Motion for New Trial, 5).

If accurately recalled, this would have been powerful evidence of fixed prejudice, violent hostility, and an intention to convict before trial.

Precisely because the accusation was so serious, Aron’s credibility became important.

The Identification Problem

Aron said he had not originally known the speaker by name and identified him afterward as Henslee.

The affidavit does not document a complete independent identification made at the time of the conversation.

This does not prove Aron was mistaken. It leaves a reasonable question concerning how securely he connected the remembered words with the photographed juror.

Henslee’s Denial

Henslee acknowledged being at the Elks Club during the relevant period but denied Aron’s statements and denied hostility toward Jewish people.

He referred to Jewish acquaintances and club members as part of his answer (Henslee Affidavit, 4).

Those relationships do not make prejudice impossible, but they formed part of the sworn material available to Roan.

The State Attacks Aron’s Credibility

The State introduced reputation affidavits from H. R. Pitts, W. L. Lyle, T. W. Webb, John R. Flournoy, Lou Castro, and others challenging Aron’s credibility.

The Atlanta Constitution also reported Aron was free on bond in connection with a pending larceny accusation (The Atlanta Constitution, 34).

A pending accusation is not a conviction. An unfavorable reputation affidavit cannot prove an alleged conversation never occurred.

Roan nevertheless had reasons to examine the credibility of the witness responsible for the most explosive anti-Jewish allegation.

C. P. Stough and the Streetcar Conversation

C. P. Stough accused Henslee of expressing belief in Frank’s guilt and a desire to help inflict punishment while riding into Atlanta on a College Park streetcar.

Henslee denied the damaging words.

Stough’s account was important because of its severity, but its history is also unusually revealing. According to Stough, the information reached Rosser through a friend, investigators pursued him repeatedly, and he eventually testified after receiving a court summons (The Atlanta Georgian, 7).

The allegation did not emerge as an affidavit voluntarily delivered to the defense immediately after the verdict.

Stough also acknowledged he had expressed strong opinions during the conversation, while the deposition did not reproduce his entire side of the exchange. That limits how completely the later reader can understand the context of Henslee’s alleged reply.

The State challenged Stough’s reputation through witnesses including Lou Castro and H. L. Bennett (Georgia Supreme Court New-Trial Record, 6).

Those challenges did not establish falsehood. They gave Roan additional conflicting evidence about the witness.

The Sparta Witnesses: Gray, Johnson, and Holmes

The Sparta controversy involved Shi Gray, John M. Holmes, and a Johnson whose initials appear differently in contemporary records.

The formal deposition identifies T. M. Johnson, while the newspaper letter names S. M. Johnson. The records concern the same Sparta accusation, but the discrepancy should remain visible until another original document settles the identification (Sparta Depositions, 45).

The three witnesses described a discussion in the Walker & Holmes office during the period before Frank’s trial.

They alleged Henslee expressed belief in Frank’s guilt positively and firmly. Holmes placed the conversation on Friday, June 27, while the others described the latter part of June.

Their accounts offered corroboration concerning one general office discussion rather than three separate incidents.

The Leading Questions

The written deposition questions reveal how Frank’s lawyers framed the testimony.

Witnesses were asked to examine a newspaper photograph and identify Henslee, then questioned about statements allegedly made between Mary’s death and the trial.

One question introduced the proposition that Henslee had wagered or wished to wager money that he would serve on the Frank jury.

The witnesses’ answers did not fully adopt that proposition.

Holmes described Henslee saying he had been summoned, while Johnson described him saying he had been drawn and might have to serve (Sparta Depositions, 45).

This is a concrete weakness in inflated retellings.

The question contained a proposed wager. The recorded answers concerned jury service. A lawyer’s question is not the witness’s testimony, and similar answers to prepared questions do not establish that every suggested detail occurred.

The questioning method does not prove coaching or false testimony. It requires a careful separation between the attorney’s suggestion and the witness’s actual answer.

The Jury-Summons Problem

The Sparta testimony contained a serious chronological difficulty.

Holmes placed the discussion on June 27.

Holmes and Johnson remembered Henslee speaking as though he had already been drawn or summoned for Frank’s jury.

Henslee swore that he first learned of his summons on Friday, July 25, at 74 Oak Street (Henslee Affidavit, 4).

If both accounts are taken literally, the chronology does not fit.

The problem affects an important detail because the alleged reference to jury service connected Henslee’s supposed prejudgment with his future role in the trial.

The contradiction does not prove the office conversation never occurred. Henslee acknowledged being in Sparta during the period, but the date and jury-summons detail were disputed.

Henslee’s September 2 Explanation

Henslee offered an answer more specific than a general denial.

He said he returned to Sparta on Tuesday, September 2, after Frank’s conviction, and admitted discussing Frank strongly during that later visit.

By then he had heard the trial evidence and participated in the guilty verdict. Expressing belief in Frank’s guilt after the trial had a different legal meaning from announcing a fixed opinion beforehand.

Henslee maintained that the Sparta witnesses had attached his post-verdict remarks to an earlier visit.

His position appeared in The Atlanta Constitution on Tuesday, October 7 (The Atlanta Constitution, 10).

The Sparta witnesses rejected his explanation and continued to insist that the conversation occurred before trial (The Atlanta Georgian, 31).

Roan therefore faced a direct credibility dispute. Henslee said the witnesses moved a later conversation backward, while the witnesses said Henslee was moving an earlier conversation forward.

The Sparta Witnesses Were Reluctant

The Sparta witnesses also acknowledged reluctance to participate.

Their public letter explained that some had known Henslee and his family for years and did not want to injure him. They initially declined voluntary affidavits and eventually gave testimony through formal procedures (The Atlanta Constitution, 9).

This works against portraying them as eager volunteers seeking publicity.

It does not establish the accuracy of their chronology.

Their strong local reputations were relevant, but reputation cannot independently settle whether a conversation occurred in June or September.

Monroe: Nunnally and Ricker

The Monroe evidence is important because Henslee admitted discussing the Frank case there.

J. J. Nunnally described Henslee as unusually forceful and alleged he argued Frank’s guilt for an extended period. Nunnally also attributed an alarming prediction about Frank’s safety after acquittal.

Dr. W. L. Ricker attributed an explicitly anti-Jewish statement to Henslee (Motion for New Trial, 5).

Henslee admitted the discussion but denied the damaging language and maintained that he had expressed only the view that whoever murdered Mary deserved hanging.

The Monroe Date

Henslee’s affidavit places the visit on Sunday, June 22, 1913 (Henslee Affidavit, 4).

The Sunday date is unusual enough to flag, but it should not be silently changed. It could represent a genuine Sunday visit, a mistaken date, or incomplete context.

Unlike an alleged conversation Henslee denied ever having, Monroe involved an admitted encounter. The dispute concerned duration, intensity, and wording.

The Limits of Corroboration

Ricker was present for only part of the discussion and could corroborate only the portion he personally heard.

Nunnally described a longer conversation, so the two accounts did not necessarily authenticate every statement or minute attributed to Henslee.

The witnesses later reaffirmed their accusations while acknowledging uncertainty about the exact timing (The Atlanta Georgian, 30).

Their persistence deserves consideration. Their uncertain chronology remains a limitation.

Virgil Harris Was Not an Accuser

Virgil V. Harris was incorrectly associated with the Monroe evidence in early reporting.

He publicly corrected the record, denying that he had made an affidavit and explaining that he was away at the relevant time.

Nunnally’s deposition also placed Harris absent (The Atlanta Journal, 29).

Harris should not be counted among the direct witnesses accusing Henslee.

His correction demonstrates why newspaper witness lists must be checked against the original sworn papers.

Leon Harrison

Leon Harrison described another alleged pretrial conversation involving Henslee.

He claimed to have overheard an exchange on Peachtree Street and later identified Henslee through newspaper photographs.

Harrison attributed a belief in Frank’s guilt and an intention to convict if selected for the jury (Motion for New Trial, 5).

The account raised questions concerning identification, how much of the conversation Harrison heard, and the accuracy of a recollection recorded months afterward.

The other speaker was not identified in the affidavit.

The Religious Language Was Not Henslee’s

Harrison’s account included a reference to Frank’s Jewish identity.

The attribution is important.

The religious reference belonged to the other participant in the conversation, while Harrison’s accusation against Henslee concerned prejudgment.

Harrison should therefore not be counted as another witness personally attributing explicit anti-Jewish language to Henslee.

The clearest direct anti-Jewish allegations remained those of Aron and Ricker.

The Three Loevenhart Affidavits Against Johenning

The case against Johenning was considerably narrower than the one developed against Henslee.

H. C. Loevenhart described one conversation.

Jennie and Miriam Loevenhart described another encounter together.

Miriam attributed the strongest alleged declaration of guilt (Loevenhart Affidavits, 2).

The affidavits were executed relatively soon after the verdict, but they concerned conversations said to have occurred months earlier. Jennie’s affidavit also followed the other two by twenty-four days.

The family relationship did not establish dishonesty or collusion. It limited the degree of independence among the accounts.

Burke’s Role as Notary

C. W. Burke administered the oaths on all three Loevenhart affidavits while also participating in the defense investigation.

This fact deserves attention because it identifies part of the process by which the statements entered the record.

A notary certifies the administration of an oath, not the historical truth of the narrative.

Burke’s participation therefore does not prove misconduct, although his later investigative controversies justify examining the original papers carefully.

Johenning’s Counter-Affidavit

Johenning denied H. C. Loevenhart’s separate conversation but admitted speaking with Jennie and Miriam.

His account was materially different.

He said he discussed how unfavorable the newspapers made Frank’s case look and added that he had only read the papers and knew nothing firsthand.

He maintained that the coming trial would determine guilt or innocence (Johenning Counter-Affidavit, 47).

The conflict was straightforward. The family remembered a settled declaration of guilt, while Johenning remembered an unfavorable newspaper impression accompanied by a willingness to await the evidence.

Johenning’s Early Public Denial

Johenning denied prejudice near the beginning of the public controversy.

The Atlanta Georgian reported him insisting that he had entered the trial impartial and required the State to prove its case (The Atlanta Georgian, 26).

On Sunday, October 5, The Atlanta Journal reported him saying his verdict rested on the evidence presented in court (The Atlanta Journal, 27).

These early statements do not establish that he was correct, but they show his explanation existed before the State’s full counter-affidavit campaign.

No comparable explicit anti-Jewish statement has been identified in the Loevenhart accusations against him. His controversy concerned alleged prejudgment rather than the particular religious language attributed to Henslee.

Character Evidence Supporting the Defense Witnesses

Frank’s lawyers obtained substantial reputation testimony for their accusers.

Nunnally and Ricker received support from Monroe residents, including R. C. Knight, H. G. Howell, O. Roberts, J. B. Shelnutt, and Judge Alonzo C. Stone.

The Sparta witnesses received extensive support from local citizens, including W. H. Burwell, Henry H. Little, Frank L. Little, T. M. Hunt, H. D. Chapman, Thomas F. Fleming, H. L. Middlebrooks, G. W. Brier, R. E. Wheeler, D. E. Wiley, A. H. Birdsong, E. A. Rozier Sr., J. D. Bennett, and Hancock County Sheriff T. B. Hightower.

Lehman and Stough also had supporting character witnesses (Georgia Supreme Court New-Trial Record, 6).

These affidavits showed the defense anticipated attacks on its witnesses’ credibility.

They did not independently establish the disputed conversations.

A person with an excellent reputation can still remember a date incorrectly, just as a person with a poor reputation can accurately describe an event.

The State’s Character Evidence Supporting the Jurors

The State responded with reputation evidence of its own.

Affiants supporting Henslee included W. M. Howard, J. C. Collier, T. W. Cochran, P. L. Cordy, J. E. Howard, J. D. Lockridge, and C. O. Summers.

W. N. Weaver, O. H. Puckett, T. W. McGarrity, and W. C. Robinson supplied favorable character evidence concerning Johenning (Georgia Supreme Court New-Trial Record, 6).

These affidavits helped address credibility.

They could not establish what Henslee said in Albany or what Johenning said to the Loevenharts.

Their importance lay in giving Roan additional information when evaluating conflicting sworn accounts.

The State’s Impeachment Witnesses

Dorsey also attacked selected defense accusers.

M. G. Staten and T. S. Hawes challenged Gremer’s reputation.

H. R. Pitts, W. L. Lyle, T. W. Webb, John R. Flournoy, and others challenged Aron.

H. L. Bennett and Lou Castro challenged Stough.

Other impeachment evidence concerned W. P. Neill, whose separate allegation involved outside contact with a juror (Georgia Supreme Court New-Trial Record, 6).

The juror-bias contest therefore involved much more than eleven accusations opposed by two denials.

Some witnesses had strong character support, some faced impeachment, some possessed documentary corroboration of location, and some had dates weakened by independent records.

The Other Jurors Defended Henslee and Johenning

Several fellow jurors executed affidavits supporting the accused men.

The record includes Foreman F. E. Winburn, D. Townsend, M. S. Woodward, A. L. Wisbey, W. M. Jeffries, J. T. Ozburn, F. V. L. Smith, C. J. Bosshardt, W. F. Medcalf, and others (Georgia Supreme Court New-Trial Record, 6).

They generally described Henslee and Johenning as conscientious during their twenty-nine days together.

Some supported the doubtful-ballot account.

Others said they had not observed either man reveal a fixed intention to convict before deliberations.

The affidavits deserve weight because the witnesses spent weeks alongside the two men during trial.

Their limits are equally clear. They could not know what Henslee said during private conversations months earlier in other towns, nor could they know precisely what Johenning said on Forsyth Street.

The Fellow Jurors Also Had an Interest

The fellow jurors had participated in the verdict.

An allegation that two of their number concealed prejudice reflected upon the integrity of the jury as a whole.

That shared interest does not make their affidavits false, but it means their evidence should receive the same scrutiny applied to the defense witnesses.

A careful assessment must consider what each person could know rather than accepting or rejecting testimony solely because of the side supporting it.

Crowd Pressure Was a Separate Question

Frank’s lawyers also alleged cheering, demonstrations, and improper outside contact influenced the jury.

Those allegations should remain separate from the claim of fixed pretrial prejudice.

A juror could enter court impartial and later face intimidation. Another juror could possess a prior opinion without being affected by a crowd.

The State answered with affidavits from jurors and court officers describing custody, supervision, movements, and alleged communications.

The Georgia Supreme Court addressed the crowd allegations separately and held the evidence did not require overturning the verdict (Frank v. State, 1).

That ruling does not establish that Atlanta’s atmosphere was calm.

Judge Roan himself acknowledged serious concern about the possibility of violence surrounding the conclusion of the trial (The Atlanta Constitution, 13).

He could recognize a hostile public atmosphere while finding insufficient proof that Henslee or Johenning had entered the jury legally disqualified.

Arnold and Dorsey Debate Religious Prejudice

Reuben Arnold incorporated the allegations against Henslee into a broader argument concerning anti-Jewish prejudice.

On Saturday, October 25, The Atlanta Georgian reported Arnold comparing the atmosphere in Atlanta with the contemporary Mendel Beilis prosecution in Kiev (The Atlanta Georgian, 35).

The religious-prejudice question was therefore part of Frank’s own legal arguments in October 1913.

Dorsey answered it directly.

Dorsey’s “Gentile Race” and “Jewish Race” Rebuttal

Under a newspaper heading concerning the denial of racial feeling, Dorsey opened his October 28 argument by invoking the Gentile community and members of the Jewish community.

He denied that opposition to Frank arose because Frank was Jewish and argued public hostility came from belief in Frank’s guilt (The Atlanta Journal, 15).

His statement belongs beside Arnold’s accusation.

Dorsey’s denial does not prove anti-Jewish prejudice was absent from Atlanta.

Arnold’s argument does not independently prove the two jurors were disqualified.

The original record establishes that religious prejudice was expressly alleged and expressly denied during the hearing.

Dorsey Distinguished Anger at the Murder From Prejudice Against Frank

Henslee acknowledged making strong statements about punishment for whoever murdered Mary Phagan.

Dorsey argued that condemning the unknown murderer was different from declaring Frank guilty.

A citizen could believe Mary’s killer deserved hanging while remaining uncertain about who committed the crime.

The defense therefore needed evidence showing that Henslee had firmly identified Frank as the murderer and closed his mind to contrary evidence.

That is why the exact language attributed by Aron, Stough, Nunnally, Ricker, Harrison, and the other accusers became important.

Newspaper Opinion Versus Fixed Prejudice

Dorsey also argued that an unfavorable impression derived from newspapers or rumors did not automatically establish a fixed legal prejudice.

This argument was especially relevant to Johenning, who admitted discussing newspaper coverage but maintained that he had reserved judgment for trial.

The Loevenharts said he expressed certainty.

Roan had to decide which account more accurately described the conversation.

Arnold’s “Cold Plotter” Theory

On Monday, October 27, The Atlanta Georgian carried the headline “Henslee Is Attacked as Cold Plotter.”

Arnold portrayed Henslee as a man who had concealed a fixed opinion and deliberately entered the jury to help convict Frank (The Atlanta Georgian, 36).

This was the defense theory in its strongest form.

It alleged intention and concealment beyond the simple claim that Henslee had discussed a sensational murder in public.

The theory remained dependent upon the reliability of the affidavits supporting it and had to account for the doubtful ballot and the favorable observations of fellow jurors.

Dorsey’s answer challenged the dates, missing corroborators, selected witnesses’ reputations, and the legal meaning of newspaper-based opinions.

The disagreement became a substantial contest over credibility rather than a straightforward tally of statements.

October 29: Dorsey Defends Both Jurors

On Wednesday, October 29, The Atlanta Georgian reported Dorsey devoting extensive attention to defending Henslee and Johenning.

He strongly praised Johenning’s conscientious character and maintained that the two jurors had performed their duty (The Atlanta Georgian, 37).

This deserves inclusion because later histories often devote far more attention to Henslee’s accusations than to the State’s answer concerning Johenning.

The defense responded by emphasizing the number of witnesses, Henslee’s incorrect Albany chronology, the doubtful ballot, and the State’s inability to impeach every accuser.

Roan heard both positions.

October 30: Eleven Henslee Affiants

On Thursday, October 30, The Atlanta Georgian reported eleven direct affiants against Henslee and said the State had attempted to impeach three.

Rosser and Arnold argued that Dorsey would have challenged the remaining eight if he had found evidence allowing him to do so (The Atlanta Georgian, 38).

This was a meaningful defense argument concerning reputation.

Its limitation is that unimpeached reputation does not establish an exact historical conversation.

The affidavits still required examination of dates, identification, wording, corroboration, and documentary conflicts.

Judge Roan Recognizes the Stakes

During the hearing on Thursday, October 23, Judge Roan recognized the potential legal effect of a proved juror-bias claim.

The Atlanta Constitution reported his statement the following day, emphasizing that proof of the accusations against Henslee and Johenning could make continued examination of the other grounds nearly unnecessary (The Atlanta Constitution, 13).

This is one of the strongest facts supporting Roan’s serious treatment of the issue.

He did not dismiss the legal importance of juror prejudice.

The question was whether the alleged facts had been sufficiently established.

Roan Personally Examines the Affidavits

The Atlanta Georgian reported that the juror-bias grounds were held while Roan personally examined the defense affidavits and the State’s counter-affidavits (The Atlanta Georgian, 14).

He had the accusations, denials, documentary travel evidence, fellow-juror statements, character affidavits, impeachment evidence, and the lawyers’ arguments.

The record does not reveal his private assessment of every witness, but it establishes that the material was presented for judicial consideration.

Friday, October 31, 1913: Roan Denies Another Trial

Roan denied Frank’s motion on Friday, October 31, 1913.

The written order was brief and did not contain separate credibility findings for each affidavit witness (The Atlanta Journal, 39).

A historical account should therefore avoid claiming Roan specifically branded Aron dishonest, accepted every word of Henslee’s Sparta explanation, or personally rejected each Loevenhart witness for a particular reason.

The formal result is clear. Roan considered the motion and conflicting evidence and denied relief.

Roan’s Personal Uncertainty About Frank’s Guilt

Roan also expressed personal uncertainty concerning Frank’s guilt.

This is historically important but should remain separate from the juror qualification issue.

A judge could regard the overall murder case as close while deciding that a particular post-verdict allegation had not been sufficiently proved.

The Georgia Supreme Court later treated Roan’s formal denial as the controlling judicial action despite his oral uncertainty (Frank v. State, 1).

Trial Judge and Appellate Court Had Different Jobs

Roan was the trial judge. He could receive the affidavits, examine conflicting evidence, and resolve factual disputes.

The Georgia Supreme Court reviewed his ruling.

Its task was not to conduct a new live hearing in which Aron, Holmes, Henslee, Johenning, and the Loevenharts testified again. It had to decide whether Roan abused the discretion Georgia law gave him.

Appellate deference did not mean the Court ignored the evidence.

It meant the Court reviewed the trial judge’s resolution of the conflict under the applicable legal standard.

On Saturday, November 1, The Atlanta Constitution discussed another Georgia appellate decision explaining the trial judge’s responsibility in conflicting juror-qualification evidence (The Atlanta Constitution, 40).

The principle was established law rather than a rule invented for Frank’s case.

February 17, 1914: The Georgia Supreme Court Reviews Henslee

The Georgia Supreme Court directly addressed Henslee in Frank v. State, 141 Ga. 243.

Ground 73 alleged he had expressed a decided opinion of Frank’s guilt and entered the trial prejudiced.

The Court acknowledged evidence tending to show strong bias before trial. It also acknowledged Henslee’s sworn denial and assertion that he had remained impartial.

The Court explained that the trial judge became the finder of fact when a juror’s impartiality was challenged through a motion for a new trial.

It found no abuse of discretion on the conflicting evidence (Frank v. State, 1).

This is the controlling legal description of Henslee’s case.

The Supreme Court did not say no evidence existed against him, and it did not declare every defense affiant dishonest. It recognized the conflict and sustained Roan’s authority to resolve it.

The Supreme Court Reviews Johenning Separately

The Court then addressed Johenning and upheld Roan’s determination that he was not prejudiced or biased (Frank v. State, 1).

This was a specific judicial ruling concerning the second juror.

The case against Johenning was narrower than the case against Henslee. Three family members supplied the principal evidence, two described the same encounter, and Johenning admitted speaking with the women while giving a detailed alternative account.

The State added character evidence and observations from fellow jurors.

Roan resolved the factual conflict in Johenning’s favor, and the Supreme Court found no abuse of discretion.

The Four-to-Two Decision Was Not a Split Over the Two Jurors

The Georgia Supreme Court affirmed Frank’s conviction by a vote of four to two.

Chief Justice William H. Fish and Justice Marcus W. Beck dissented over a separate evidentiary issue involving testimony concerning alleged prior sexual conduct.

Their published dissent did not establish that Henslee or Johenning was disqualified.

The division should therefore not be described as a four-to-two disagreement over the two accused jurors (Frank v. State, 1).

The Court also independently held the trial evidence legally sufficient to uphold the guilty verdict. That ruling should remain separate from juror impartiality, since sufficient evidence cannot cure a genuinely disqualified juror.

C. W. Burke and the Later Investigative Controversies

C. W. Burke connects the 1913 juror investigation with later disputes concerning defense witness-handling methods.

In October 1913 he was identified as an agent for Rosser and served as notary on the Loevenhart affidavits.

Later his name appeared in several disputed investigative episodes.

Those later events do not prove the original juror affidavits false, but they provide reasons for examining document provenance and investigative methods carefully.

Burke and Ira Fisher

Burke helped bring Ira W. Fisher from Birmingham after Fisher made a sensational accusation involving another man in the Phagan case.

Frank’s attorneys soon distanced themselves from Fisher’s story (The Atlanta Journal, 41).

The episode demonstrates how an investigative lead could produce dramatic allegations that later lost evidentiary value.

It does not prove Burke invented the account.

Burke, George Epps, and “Kelly”

George Epps later signed a statement altering parts of his earlier account and then repudiated it.

Disputes arose concerning investigators using assumed identities.

N. A. Garner later identified a man described as “Kelly” as C. W. Burke (Georgia Supreme Court Appeals Record, 49).

Burke and defense representatives disputed important portions of the episode.

Because it occurred after Roan ruled in October 1913, it cannot explain his original decision.

Burke and Albert McKnight

Albert McKnight later signed an affidavit favorable to Frank and subsequently repudiated it.

He accused Burke of discussing employment and housing, while Burke denied any agreement to purchase testimony (The Atlanta Journal, 41).

The surviving material establishes a controversy concerning possible assistance and witness handling, but does not conclusively prove a completed exchange of false testimony for employment.

Later Forgery Accusations

During the 1914 extraordinary-motion litigation, the State also challenged other Burke-connected affidavits, including papers attributed to Ivey Jones and Ruth Robinson or Robison.

These were contested accusations.

They do not prove the earlier Loevenhart affidavits false, but they strengthen the reason for inspecting original documents rather than relying exclusively on later transcriptions (Georgia Supreme Court Appeals Record, 49).

June 18, 1914: Burke Is Indicted

On Thursday, June 18, 1914, The Atlanta Journal reported a Fulton County grand-jury indictment accusing Burke of attempted subornation of perjury in connection with a Ferguson witness (The Atlanta Journal, 42).

An indictment is a formal accusation, not a conviction.

The materials examined here do not establish the final disposition of the charge.

Its timing is equally important. It occurred more than seven months after Roan denied Frank’s first new-trial motion, so Roan could not have considered it when deciding the original Henslee and Johenning claims.

Burke’s later controversies justify careful scrutiny without proving wholesale fabrication.

The Strongest Evidence Against Henslee

The defense had serious evidence against Henslee.

Multiple unrelated witnesses from several Georgia communities accused him of strong prejudgment. Aron and Ricker attributed explicitly anti-Jewish language, while some witnesses described conversations whose occurrence Henslee substantially admitted.

The July 8 Albany business and hotel records established a significant error in his sworn travel reconstruction.

The defense also produced extensive character evidence supporting many of its accusers.

The attack was not one unsupported rumor.

The Strongest Evidence Supporting Henslee

The State possessed substantial contrary evidence.

The Blakely allegation failed.

The June 2 hotel and travel records challenged Lehman’s original chronology.

The Sparta summons detail conflicted with Henslee’s account of first receiving notice on July 25.

Virgil Harris publicly denied giving an affidavit.

Several defense witnesses faced impeachment.

Henslee admitted some conversations rather than denying every encounter.

Numerous fellow jurors reported no observed prejudice, and several supported his doubtful-ballot account.

The doubtful vote also appeared in contemporary newspaper reporting immediately after the verdict.

Those facts explain why Roan could regard the evidence as genuinely conflicting.

The Strongest Evidence Against Johenning

The three Loevenharts made sworn accusations relatively soon after the verdict.

Miriam attributed a definite declaration of guilt.

Johenning admitted speaking with Jennie and Miriam.

The workplace connection with H. C. received independent support from city directories.

If the family accurately remembered a fixed pretrial judgment, its evidence was serious.

The Strongest Evidence Supporting Johenning

All three principal accusers belonged to one family, and Jennie and Miriam described the same general encounter.

Johenning denied H. C.’s separate conversation and gave a detailed alternative account of his discussion with the women.

His explanation tied the remarks to newspaper coverage and reserved final judgment for trial.

No explicit anti-Jewish statement was attributed to him.

Fellow jurors supported his conduct, character witnesses supported his reputation, contemporary reporting showed him engaging with evidence, and a pre-verdict reporter portrayed him as apparently indecisive.

Roan resolved the conflict in his favor, and the Georgia Supreme Court upheld that decision.

Why Judge Roan’s Decision Was Defensible

Roan presided over Frank’s trial, observed the jurors for weeks, knew the courtroom atmosphere, and received the defense affidavits and the State’s counter-showing.

He had the Albany contradiction, Henslee’s travel records, Lehman’s date problem, the Sparta summons conflict, the Monroe accusations, Aron’s affidavit, Stough’s deposition, the Loevenhart statements, character evidence, and the fellow-juror affidavits.

He heard the defense arguments and Dorsey’s answer.

Most importantly, he recognized that proved juror prejudice could justify another trial.

He examined the evidence and denied relief.

Supporting his ruling does not require claiming every defense witness lied. It requires recognizing that the record contained enough factual conflict for the trial judge to resolve the issue against Frank without exceeding his lawful discretion.

Why the Georgia Supreme Court’s Decision Was Defensible

The Supreme Court had a different job.

Georgia law placed primary responsibility for resolving conflicting post-verdict juror evidence with the trial judge. The appellate court reviewed whether Roan abused that discretion.

Regarding Henslee, the Court acknowledged evidence pointing toward prejudice, noted the sworn denial, and found no abuse.

Regarding Johenning, it expressly upheld Roan’s determination that the juror was not prejudiced or biased.

This was an established approach to reviewing factual conflicts.

It does not prove every defense witness wrong. It establishes that Frank failed to demonstrate that Roan’s ruling exceeded his lawful discretion.

An Archival Guide to the Original Documents

The Georgia Supreme Court appeal records reproduced by the Mary Phagan-Kean Legacy Project make it possible to examine many of the original documents discussed in this study.

The most useful starting points are Slide 0052 and following for Henslee’s sworn counter-affidavit and Slides 0043–0046 for Johenning’s answer. Those records contain the jurors’ responses, statements about their conduct during trial, and descriptions of earlier conversations.

Slides 0943–0945 contain principal early accusations against Henslee, including those involving Gremer, Mack Farkas, Lehman, and Aron. Slides 0946–0947 contain the Loevenhart affidavits against Johenning, while Slides 0262–0265 provide another reproduction useful for checking the dates and signatures.

Slides 0968 and following contain Stough’s testimony and commission material. Slides 0970–0972 concern the Sparta witnesses and their answers to written questions. Slides 0974 and following contain the Monroe testimony of Nunnally and Ricker.

Slide 0296 contains Leon Harrison’s accusation. Slides 0959–0961 concern the July 8 Franklin Buggy Company transaction, while Slide 0303 and related material concern W. W. Little’s identification of Henslee’s New Albany Hotel registration.

Slide 0962 contains Lehman’s supplemental affidavit and the Monroe witnesses’ reaffirmation. Slide 0069 and surrounding pages lead into State counter-affidavits. The original materials containing the defense lawyers’ claimed lack of prior knowledge appear around Slides 0945–0948.

Later documents concerning C. W. Burke, George Epps, and disputed affidavits appear around Slides 1493 and 1717. Those documents belong to the later investigative controversies and must not be confused with evidence available to Roan in October 1913.

This archival organization makes it possible to compare the first accusation, the juror’s answer, the supporting documents, and the State’s rebuttal without depending solely on newspaper summaries.

Final Conclusion

The post-verdict attack upon Atticus Haygood Henslee and Marcellus Johenning was one of the most serious parts of Leo Frank’s first motion for a new trial. It deserves careful examination because the allegations concerned the impartiality of two men who had helped return a capital conviction.

Henslee faced substantial accusations from several communities. Some witnesses alleged fixed prejudgment, two attributed explicitly anti-Jewish language, and documentary evidence exposed a genuine error in his Albany travel chronology. Those weaknesses in Henslee’s position should not be hidden.

The weaknesses in the defense affidavit campaign should not be hidden either. The Blakely lead failed, Lehman’s original June 2 chronology weakened, the Sparta testimony contained a serious summons problem, and Virgil Harris publicly corrected an inaccurate witness listing. Several affiants were reluctant or required legal process, some faced impeachment, and multiple affidavits described the same event rather than independent conversations.

The strongest documentary evidence against Henslee established his presence in Albany and contradicted his travel account. It did not independently establish the prejudicial words attributed to him. His doubtful ballot was reported immediately after the verdict and later supported by fellow jurors, creating a meaningful difficulty for the strongest secret-plot theory.

Johenning faced a narrower challenge. Three members of one family accused him of pretrial prejudgment. He admitted speaking with two of them and disputed the contents of their conversation. His version described newspaper impressions and reserved judgment for trial, while fellow-juror and character evidence supported his conduct.

The defense investigation was aggressive, organized, and directed toward a specific legal result. That was understandable for lawyers representing a condemned man, but it also means the resulting affidavits should be analyzed individually rather than accepted through their number. C. W. Burke’s later controversies justify close source criticism without proving the original juror affidavits fraudulent.

Dorsey’s rebuttal deserves equal attention alongside Arnold’s accusations. The State challenged dates, missing corroboration, witness credibility, the interpretation of newspaper opinions, and the broader religious-prejudice argument.

Judge Roan heard both sides. He understood that proved juror prejudice could justify another trial, personally examined the competing evidence, and rejected the juror-bias grounds.

The Georgia Supreme Court then reviewed Henslee and Johenning separately on Tuesday, February 17, 1914. Regarding Henslee, it recognized conflicting evidence and found no abuse of discretion. Regarding Johenning, it expressly upheld Roan’s determination that the juror was not prejudiced or biased.

The courts did not erase the accusations or declare every accuser dishonest. They resolved a contested evidentiary record under Georgia law.

The strongest historical argument supporting Judge Roan and the Georgia Supreme Court rests upon the documented weaknesses of the post-verdict affidavits, the substantial evidence answering them, and the trial judge’s legal responsibility to resolve conflicting testimony. Frank’s lawyers found allegations serious enough to investigate, but the Georgia Supreme Court held that the competing evidence did not require overturning Roan’s decision.

References

1. Supreme Court of Georgia. (1914, Tuesday, February 17). Frank v. State, 141 Ga. 243, 80 S.E. 1016. Ground 73, juror impartiality, the trial judge’s discretion, and the separate Johenning challenge.

2. Superior Court of Fulton County. (1913, September 2–26). Affidavits of H. C., Miriam, and Jennie G. Loevenhart concerning Marcellus Johenning. Exhibits F–H, reproduced archival record.

3. Lehman, J. A. (1913, Monday, October 13). Supplemental affidavit, Exhibit DD, printed record page 142.

4. Henslee, A. H. (1913, October). Sworn counter-affidavit concerning juror prejudice, business travels, jury summons, and deliberations. Reproduced appellate record, beginning Slide 0052.

5. Superior Court of Fulton County. (1913). State of Georgia v. Leo M. Frank: Motion for new trial, exhibits, depositions, certification, and order.

https://www.leofrank.com/library/trial/leo-frank-motion-for-new-trial-3-exhibits-charge-to-jury.pdf

6. Superior Court of Fulton County. (1913, October). New-trial record and State counter-affidavits. Henslee and Johenning responses, fellow-juror evidence, character testimony, impeachment evidence, and travel records.

https://www.leofrank.org/library/leo-frank-first-appeal-october-1913.pdf

7. The Atlanta Georgian. (1913, Friday, October 3). “Frank Trial Juror Denies Charge of Bias.” Stough’s account of repeated investigative approaches.

8. The Atlanta Constitution. (1913, Friday, September 26). “Did Not Discuss Guilt of Frank.” p. 5, col. 1. Blakely depositions of L. E. Black and Walter Thomas.

9. The Atlanta Constitution. (1913, Monday, October 6). “Sparta Citizens Insist Henslee Was Prejudiced.” p. 1. Public letter from the Sparta witnesses.

10. The Atlanta Constitution. (1913, Tuesday, October 7). “Made His Remarks About Leo Frank After Trial and Not Before, Says Juror Accused of Prejudice.”

11. The Atlanta Georgian. (1913, Tuesday, October 7). “Dorsey at Work to Combat Charge.” Henslee’s confrontation with Julian Lehman.

12. The Atlanta Georgian. (1913, Thursday, October 23). “Judge’s Admissions Help Frank’s Chance.” Defense investigation of the jurors and C. W. Burke’s role.

13. The Atlanta Constitution. (1913, Friday, October 24). “Proof of Charges Will Mean a New Trial, Says Court.” p. 1, col. 2. Reports Roan’s October 23 statement.

14. The Atlanta Georgian. (1913, Friday, October 24). “Disputes Block Frank Speech.” Roan’s examination of competing affidavits.

15. The Atlanta Journal. (1913, Tuesday, October 28). “Solicitor Dorsey Hammers Frank New Trial Motion.” Dorsey’s investigation, prejudice, corroboration, and religious-prejudice arguments.

16. The Atlanta Journal. (1913, Wednesday, October 29). “Frank New Trial Hearing to End This Afternoon.” Rosser’s doubtful-ballot argument.

17. The Atlanta Journal. (1913, Tuesday, August 26). “Twelve Peers of Leo M. Frank Who Found Guilty of Murder.” Early divided-ballot account.

18. The Atlanta Constitution. (1913, Tuesday, August 26). “Only One Ballot Taken by Jury and Fate of Frank Was Decided.”

https://www.leofrank.org/library/atlanta-constitution-issues/1913/atlanta-constitution-august-26-1913-tuesday-16-pages-combined.pdf

19. The Atlanta Georgian. (1913, Tuesday, August 26). “Frank Guilty on First Ballot.”

20. The Atlanta Journal. (1913, Monday, July 28). “Personnel of Frank Jury All Married Except One.”

21. The Atlanta Constitution. (1913, Tuesday, July 29). Trial-opening jury information identifying Henslee and Johenning.

https://www.leofrank.org/library/atlanta-constitution-issues/1913/atlanta-constitution-july-29-1913-tuesday-14-pages.pdf

22. Lamar County, Georgia, GaGenWeb. (n.d.). Burial and marriage records. Atticus Haygood Henslee and Elizabeth May Franklin Henslee.

23. Atlanta City Directory Company. (1912, 1914, 1916, 1919). Atlanta city directories. Entries for Henslee, Johenning, Annie Johenning, and Hodges Broom Works.

24. The Atlanta Journal. (1913, Friday, August 1). “Lawyers Battle Over Testimony of Frank’s Nervousness.” Johenning’s time-clock question.

25. Collier, T. (1913, Sunday, August 24). “Frank Jurors an Enthralling Study in Human Emotion.” Hearst’s Sunday American, p. 4A.

https://www.leofrank.org/enright-archive/atlanta-georgian-1913-08-24-full-pages.pdf

26. The Atlanta Georgian. (1913, Thursday, October 2). “Ask New Frank Trial on 115 Counts.” Johenning’s public denial.

27. The Atlanta Journal. (1913, Sunday, October 5). “Two Frank Jurors Firm in Denying Outside Pressure.”

28. The Atlanta Journal. (1913, Saturday, October 4). “Affidavits Attacking Frank Jurors Made Public.”

29. The Atlanta Journal. (1913, Wednesday, October 8). “Frank Hearing to Be Postponed Another Week.” Virgil Harris’s correction.

30. The Atlanta Georgian. (1913, Monday, October 13). “Attack Is Renewed on Frank Juror.” Nunnally and Ricker.

31. The Atlanta Georgian. (1913, Friday, October 17). “Sparta Citizens Attack Frank Trial Juror.”

32. The Atlanta Georgian. (1913, Monday, October 20). “Way Clear for Frank Battle.” Leon Harrison affidavit.

33. The Atlanta Georgian. (1913, Tuesday, October 21). “Thirty Affidavits Against New Frank Trial.”

34. The Atlanta Constitution. (1913, Wednesday, October 22). “Many Affidavits Defend the Jury.” p. 1, col. 4.

35. The Atlanta Georgian. (1913, Saturday, October 25). “Atlanta’s Prejudice as Bitter as Russia’s, Declares Attorney.”

36. The Atlanta Georgian. (1913, Monday, October 27). “Henslee Is Attacked as Cold Plotter.”

37. The Atlanta Georgian. (1913, Wednesday, October 29). “Negro’s Statement Legal Evidence, He Says; State Closes.”

38. The Atlanta Georgian. (1913, Thursday, October 30). “Frank Alibi Upheld by Rosser in Closing.” Eleven direct Henslee accusers.

39. The Atlanta Journal. (1913, Friday, October 31). “Leo Frank’s Lawyers Prepare for Supreme Court Fight.”

40. The Atlanta Constitution. (1913, Saturday, November 1). “Trial Judge Says Last Word on Bias of Jury, Declares Court of Appeals.”

41. The Atlanta Journal. (1914, Tuesday, May 5). “Many Sensations Sprung by Dorsey at Frank Hearing.” Later Burke, Epps, and McKnight disputes.

42. The Atlanta Journal. (1914, Thursday, June 18). “C. W. Burke Indicted for Suborning Perjury.” p. 1, col. 5.

43. The Atlanta Georgian. (1916, Sunday, June 4). Henslee’s Georgia Prison Commission candidacy.

44. The Jeffersonian. (1916, Thursday, July 6). Henslee campaign advertisement invoking the Frank jury controversy.

45. Superior Court of Fulton County. (1913, Friday, September 26). Sparta depositions of Shi Gray, Johnson, and John M. Holmes. Exhibit MM.

46. Superior Court of Fulton County. (1913). Franklin Buggy Company Albany business order and W. W. Little hotel affidavit.

47. Johenning, M. (1913, October). Counter-affidavit answering the Loevenhart family.

48. Superior Court of Fulton County. (1913). Original affidavits of Gremer, Farkas, Lehman, Aron, Stough, Nunnally, Ricker, and Harrison.

49. Superior Court of Fulton County. (1914). Later State answer and affidavit material concerning C. W. Burke, George Epps, Ivey Jones, Ruth Robinson or Robison, and N. A. Garner.