Canonical murder 3: LB vol.2. The Reports Beneath the Hood

Canonical murder 3: Lake Berryessa vol.2.

The Reports Beneath the Hood

 

Testimony, Dialogue, Physical Evidence, and the Formation of a Canonical Crime

By Space; By Time; By Motif

 

Archive and Methodological Note

This study is based upon the surviving Lake Berryessa police-report collection assembled under Napa County Sheriff’s Department case number 105907. The present research archive contains fifty-eight documentary scan units:

• twenty-one Initial Report scans;
• twenty-five Witness Report scans;
• twelve Evidence Report scans.

The collection includes the initial incident reports, the thirteen-page Narlow–Lonergan investigative narrative, witness and rescuer statements, Officer David Slaight’s account of the Napa telephone call, Detective Sergeant John Robertson’s interview materials, Bryan Hartnell’s signed dialogue reconstruction, Sergeant Hal Snook’s physical-evidence report, vehicle and footwear inquiries, and the subsequent California Department of Justice and Federal Bureau of Investigation latent-impression examinations.

The archive is “complete” in the limited documentary sense used here: it contains the full police-report corpus defined by the assembled collection. It does not contain every item generated by the investigation. Missing or unlocated materials include the original Hartnell audio recording, the complete photographic inventory, the aerial photographs, the plaster casts, dispatch recordings, all laboratory worksheets, the original car door, and the full questioned-document report relating to the writing on the vehicle.

This volume does not attempt to identify the offender. It does not convert every person investigated in 1969 into a retrospective suspect. Individuals who were cleared, incompletely investigated or recorded only as preliminary leads are treated according to the disposition documented in the reports.

The purpose is narrower and, in another sense, deeper:

To examine how the Lake Berryessa event moved through wounded memory, emergency testimony, police reporting, physical measurement, institutional comparison and eventual Zodiac canon.

 

1. Introduction: Beneath the Received Narrative

Lake Berryessa is usually remembered through a compressed sequence.

A hooded man approached a young couple beside the lake. He bound them. He stabbed them. He wrote dates on their car. He telephoned the police.

That sequence is not false. It is simply incomplete.

The police archive reveals a much larger event. It contains an afternoon observation field involving three young women and an unidentified man near a light-blue Chevrolet; a prolonged conversation in which Bryan Hartnell attempted to negotiate survival; a robbery story that collapsed under the offender’s own conduct; uncertain and changing recollections of a Montana prison; a rescue conducted by boat, truck and radio; a path of distinctive footprints between shoreline and vehicle; tire impressions, soil samples and plaster casts; a payphone receiver left off the hook; forty-five latent lifts delivered to the California Department of Justice; federal comparison with prints from another Zodiac-related file; and a full passenger door physically removed from the Karmann Ghia and entered into evidence.

The reports also preserve something less tangible but equally important: the gradual construction of a canonical account.

The canonical Lake Berryessa narrative was not produced at the shoreline in one complete form. It emerged through multiple documentary transformations:

The value of the archive lies not in pretending that these layers agree perfectly. Its value lies in showing where they converge, where they diverge and where the evidence stops.

Figure 1. Initial Incident Classification and Modus Operandi. The first page of the Collins–Land incident report records the shoreline setting, hooded armed assailant, binding and multiple stabbing of Bryan Hartnell and Cecelia Shepard. The report represents the case at its first formal administrative stage, before later witness, forensic and interagency expansion.
Source: LB-NCSD-IR-002. Original file: [01 Initial Report 02] Initial report by Collins + Land P1.jpg.

 

2. The Documentary Architecture

The completed archive naturally divides into six interrelated modules.

Table 1. Documentary Architecture of the Lake Berryessa Police-Report Archive

Module	Principal records	Documentary function
LB-A — Event Formation	Collins–Land initial incident reports	Attack, first rescue, vehicle and immediate chronology
LB-B — Investigative Expansion	Narlow–Lonergan supplementary report	Witnesses, scene preservation, autopsy, footwear and interagency development
LB-C — Voice and Memory	Slaight report, Robertson materials, Hartnell transcript and signed dialogue	Offender speech, survivor memory and documentary transformation
LB-D — Peripheral Witness Field	Three-woman reports, ranger and rescuer statements	Possible pre-attack observation and rescue geography
LB-E — Physical Corridor	Snook evidence report, vehicle report, casts, soil and property records	Material route between shoreline, vehicle and telephone
LB-F — Institutional Comparison	California DOJ and FBI reports	Latent-print examination, questioned documents and negative comparison results

 

These modules do not form a single evidentiary hierarchy in which one automatically supersedes the others.

A direct physical measurement may be stronger than a later estimate of distance. A surviving witness’s detailed recollection may be stronger than an officer’s compressed paraphrase. Yet a contemporaneous emergency report can preserve a statement made closer to the event than a later formal interview.

The correct method is therefore comparative rather than eliminative.

 

3. September 27, 1969: The Day Before the Archive

Bryan Hartnell and Cecelia Shepard had known one another for some time. Shepard had returned to the area for the weekend, and the two spent part of Saturday together. Their afternoon did not begin with a fixed plan for Lake Berryessa.

According to Hartnell’s hospital interview, they visited friends, stopped at a rummage sale in St. Helena and considered several possibilities for the remainder of the afternoon. They might take a walk. They might travel toward San Francisco. They might simply find somewhere quiet to talk. By the time their earlier errands were completed, the afternoon was already advanced. Hartnell chose Lake Berryessa. [5], [7]

The Narlow–Lonergan investigation recorded the approximate sequence as follows:

 

Table 2. Approximate Chronology Recorded in the Narlow–Lonergan Investigation

Approximate time	Event
1:00 p.m.	Hartnell and Shepard leave Pacific Union College
Early afternoon	St. Helena rummage-sale stop
About 2:00 p.m.	Hartnell returns to collect Shepard
About 2:45–4:30 p.m.	Three other young women occupy a separate Lake Berryessa area and observe an unidentified man
About 5:15 p.m.	A college acquaintance reports seeing Hartnell and Shepard arrive
About 6:30 p.m.	Attack time later recorded in the police reports and on the vehicle door

 

The couple parked near Knoxville Road and walked down toward the shoreline. Hartnell described the terrain as peninsula-like, partly shaped by the seasonal level of the lake. They selected one of two large trees and rested beneath the larger, more exposed tree near the water.

The setting was peaceful enough that Hartnell later described it as beautiful.

Nothing in the surviving reports establishes that the couple chose the location because of any prior ritual, anniversary or association, Zodiac-related or otherwise. It was an afternoon destination reached after other plans had remained open.

 

4. Before the Hood: The Three-Woman Observation Field

Several hours before the attack, three Pacific Union College students entered the broader Lake Berryessa area from Angwin by way of Pope Valley.

Their accounts were preserved both individually and through Deputy Ray Land’s consolidated summary. [8]

At approximately 2:45 p.m., the women pulled into a parking area. Before they could leave their vehicle, another car reportedly passed them and backed into a position close beside or behind them.

The car was described as:

• a Chevrolet;
• approximately a 1966 model, although estimates varied;
• a two-door, full-sized sedan;
• silver-blue, ice-blue or light blue;
• carrying California plates;
• conservative in appearance;
• with rear glass described by one witness as unusually dark.

The driver allegedly sat with his head lowered as though reading. The women suspected that he might be pretending.

They proceeded to the shoreline at approximately 3:00 p.m. At about 3:30 p.m., they observed what they believed to be the same man watching them.

The combined description included:

• white male;
• approximately twenty-five to thirty-five years old;
• about six feet to six feet two inches;
• approximately 200–230 pounds;
• dark, straight, parted hair;
• no glasses observed;
• dark blue trousers;
• a black short-sleeved sweater or knitted shirt;
• white material or a T-shirt visible at the rear;
• cigarettes;
• a clean-cut or comparatively pleasant appearance.

One report placed him within approximately twenty feet during part of the observation.

When the women returned to their car at approximately 4:30 p.m., the Chevrolet was gone.

This is a significant investigative lead. It is not an identification.

The reports establish that three women observed a man behaving unusually within the Lake Berryessa recreational field on the same afternoon. They do not establish that he later became the hooded assailant.

The distinction is essential:

The witnesses saw the man’s uncovered head and ordinary clothing. Hartnell and Shepard later encountered a costumed figure. Any connection requires more than proximity, build and suspicious behaviour.

No report in the supplied archive physically joins the light-blue Chevrolet to the tire impressions later cast near the Karmann Ghia.

 

Figure 2. The Three-Woman Observation Field. Deputy Ray Land’s consolidated record of the three Pacific Union College students’ observations preserves the light-blue Chevrolet, approximate chronology and shared description of an unidentified man. The document records a significant investigative lead, not a confirmed identification of the Lake Berryessa assailant.
Source: LB-NCSD-WR-022.Original file: [02 Witness Report 22] Land interview with 3 girls.png

 

5. Sound, Concealment and Emergence

Hartnell and Shepard were resting beneath the tree when Hartnell heard movement in nearby leaves. He asked Shepard whether she could see anything. She saw a man partly concealed behind a tree but initially could not determine what he was doing.

The situation changed when the man emerged. Shepard recognized the weapon and exclaimed that he had a gun. [7]

The approach therefore unfolded in stages:

 

The reports do not establish when the hood was first put on. Sergeant Snook later collected a soil sample from behind the tree where investigators believed the offender may have positioned himself while preparing or adjusting the costume. [10]

That sample proves that investigators considered the location significant. It does not itself prove that the offender dressed there.

 

6. The Hood as a Constructed Object

The public description “ceremonial hood” is memorable but interpretive.

Hartnell’s fuller interview described a physical construction:

• black cloth;
• bag-like or approximately four-cornered at the top;
• covering the entire head and shoulders;
• extending down the front toward the stomach;
• extending down the back;
• carrying eye openings;
• with clip-on sunglasses attached over those openings.

Hartnell described the mask as “ingeniously devised.” He could see a limited amount of hair through or around the eye openings. He believed it was dark brown, although changing light may have made parts of it appear greyish. [7]

The chest emblem was described as:

• white;
• displayed against the black garment;
• a circle crossed by symmetrical vertical and horizontal lines;
• the arms extending beyond the circumference;
• approximately three by three inches, possibly approaching four by four;
• formed with comparatively thick marks.

During the final pages of the hospital interview, Robertson momentarily reversed the colours. Hartnell corrected him: the field was black; the emblem was white.

This is not a trivial detail.

The costume did not merely conceal the face. It presented an authored surface.

The body became a field of communication before the car door became one.

 

Figure 3. Constructing the Hood: Material Form, Eye Openings, and Emblem. In the hospital interview, Hartnell described a black, bag-like covering extending over the head and shoulders, with eye openings, attached clip-on sunglasses and a white symmetrical crosshair emblem. His recollection is primarily material and geometric; the frequently repeated description “ceremonial” originated as an interpretive classification rather than Hartnell’s complete physical description.
Sources: LB-NCSD-WR-012 and LB-NCSD-WR-013. Original files: [02 Witness Report 12] 1969-09-28 Bryan Hartnell - John Robertson Interview-04.jpg; [02 Witness Report 13] 1969-09-28 Bryan Hartnell - John Robertson Interview-05.jpg.

 

7. The Script Before the Knife

The assailant initially presented the encounter as robbery.

He claimed to need money and transportation. He said that his own car was “hot.” He referred to escape from a prison in Montana, violence against a guard or guards and travel toward Mexico. [5]–[7]

Hartnell responded not with silence but with an improvised survival strategy.

He attempted to:

• cooperate;
• appear non-threatening;
• keep the man talking;
• offer the little money he possessed;
• offer a check;
• offer a telephone number;
• suggest that he could provide further assistance;
• and delay the offender’s movement toward irreversible violence.

Hartnell explained that his interests in sociology, history and psychology led him to engage the man conversationally. This was not academic curiosity detached from danger. It was an attempt to remain human in the eyes of an armed stranger.

The offender and Hartnell were therefore conducting two competing scripts.

Table 3. Competing Interaction Scripts: Offender Control and Hartnell’s Survival Strategy

Offender’s script	Hartnell’s counter-script
“I am a desperate escaped prisoner.”	“I believe you and will cooperate.”
“I need money and a car.”	“Take what I have; I can help further.”
“I have killed before.”	“There is no need for more violence.”
“Do not resist.”	“I will keep you speaking.”
“You are under my physical control.”	“I will seek psychological room for survival.”

 

The signed dialogue reconstruction preserves this strategy with unusual clarity. Hartnell repeatedly offered cooperation and assistance, even suggesting continued contact.

The offender rejected genuine assistance.

He did not ultimately take:

• the offered money;
• the wallet;
• the loose change;
• or the car keys.

The robbery story therefore performed a control function that the actual taking of property did not fulfil.

The reports strongly support the following conclusion:

The offender’s robbery narrative was behaviourally inconsistent with ordinary robbery and functioned principally as an instrument of compliance.

The reports do not establish the full psychological motive that replaced it.

 

8. Deer Lodge: Memory, Prompting and Canon

Few details from Lake Berryessa became as famous as “Deer Lodge.”

Yet the reports preserve more than one form of the prison story.

 

Table 4. Documentary Variants of the Prison-Escape Narrative

Source	Recorded version
Collins–Land initial narrative	Escape from Deer Lodge, Montana; stolen car; nothing to lose
Ranger White’s relayed account	Ex-convict associated with Colorado; travelling toward Mexico
Robertson’s summary	Possibly Deer Lodge State Prison; a guard killed during escape
Full taped-interview transcript	Montana prison and uncertain compound name; Robertson supplies “Lodge”
Hartnell’s signed dialogue	“Deer Lodge in Montana”; “a couple of guards” killed

 

During the taped interview, Hartnell remembered Montana and remembered that the prison name sounded compound or doubled. He struggled with the exact wording. Robertson offered the word “Lodge,” which Hartnell accepted. [7]

This does not prove that the name was invented by the interviewer. Other near-contemporaneous sources already preserve “Deer Lodge,” and Hartnell independently remembered Montana and a compound name. It does, however, prove something methodologically important: The exact canonical wording was not retrieved in one smooth, unaided statement during the hospital interview.

The transformation was gradual:

The number of allegedly killed guards also changed:

• one guard in one account;
• a couple of guards in another.

The stable core is narrower:

• prison escape;
• Montana;
• violence against guard personnel;
• claimed movement toward Mexico.

The archive should preserve that core without pretending that every spoken detail remained fixed.

 

Figure 4. Memory Before Canon: From an Uncertain Montana Prison to “Lodge.” Consecutive pages of the Hartnell interview preserve his uncertainty over the compound prison name and Robertson’s introduction of the word “Lodge.” The exchange does not disprove the Deer Lodge account; it documents one stage through which the now-canonical wording became stabilized.
Sources: LB-NCSD-WR-010 and LB-NCSD-WR-011. Original files: [02 Witness Report 10] 1969-09-28 Bryan Hartnell - John Robertson Interview-02.jpg; [02 Witness Report 11] 1969-09-28 Bryan Hartnell - John Robertson Interview-03.jpg.

 

9. The Offender’s Voice

Hartnell described a distinctive way of speaking.

He rejected a simple Southern “drawl” classification. What he remembered was closer to:

• an unusual cadence;
• something slightly sing-song;
• a moderate pitch;
• neither notably high nor low;
• a manner he thought he might recognize but could not reproduce.

Hartnell did not believe the man sounded highly educated, but he did not consider him illiterate. He associated speech, clothing and presentation with what he called a lower-class impression.

That was a witness judgment, not an established social or educational classification.

Officer David Slaight later described the telephone caller as:

• male;
• young-sounding;
• possibly in his early twenties;
• calm;
• and barely audible during the final confession. [3]

The two voice descriptions are not identical, but they are not inherently incompatible. The listening conditions were radically different.

 

Table 5. Comparative Voice Impressions: Bryan Hartnell and David Slaight

Listener	Setting	Reported impression
Bryan Hartnell	Prolonged face-to-face encounter under threat	Distinctive cadence; moderate pitch; estimated age 20–30
David Slaight	Brief telephone call through switchboard	Young-sounding; possibly early twenties; calm

 

The reports later show that voice comparison was used to eliminate at least one preliminary suspect. In that case, the person’s speaking manner was regarded as inconsistent with Hartnell’s description, and his movements and lack of vehicle access were corroborated. [2]

Voice was therefore used as an investigative filter, not as a scientific identification.

 

10. Binding, Shaking Hands and Escalating Control

The binding sequence is preserved most clearly in Robertson’s summary and the complete interview. [5]-[7]

The sequence was approximately:

1. Shepard tied Hartnell’s hands.
2. The offender inspected the binding.
3. He tightened or retied Hartnell.
4. He tied Shepard’s hands.
5. He connected or tied the victims’ ankles.
6. He ordered them to lie face down.
7. He completed a form of hog-tying.

Hartnell recalled considering an attempt to reach the gun. He whispered to Shepard but abandoned the plan because her life as well as his own would be placed at immediate risk.

Multiple Hartnell-derived records agree that the offender’s hands appeared to shake.

They do not preserve identical dialogue.

Robertson’s summary presents the offender as saying:

“I’m getting nervous.”

The signed reconstruction instead has Hartnell ask whether the offender is nervous, with the offender agreeing.

The full transcript preserves a broader behavioural impression: shaking hands, increasing tension, a more serious manner and Hartnell’s belief that the man was becoming rattled.

The safe conclusion is:

Hartnell perceived visible shaking and increasing nervousness during the binding process. The precise wording of the exchange varies among the surviving documentary forms.

The phenomenon is relatively stable.

The quotation is not.

 

11. The Gun Demonstration

After the victims had been restrained, Hartnell asked whether the gun was loaded.

The offender manipulated the weapon or its magazine and displayed a cartridge.

Hartnell did not possess enough firearms knowledge to identify the calibre confidently. He later described a brass- or gold-coloured cartridge or projectile that seemed comparatively short and blunt.

On October 1, investigators showed Hartnell a range of automatic pistols, including examples in .45, .380, .32, 9 mm and .25 calibre. [13]

He could not identify any pistol as the weapon used at Lake Berryessa.

Later that day he was shown several kinds of ammunition. He tentatively believed the displayed cartridge may have resembled a .45 automatic cartridge with brass-coloured components.

The report expressly records his uncertainty.

The evidence therefore supports two different conclusions:

The frequent public shorthand “the offender carried a .45” is stronger than the surviving identification evidence.

 

12. The Weapon Transition

Hartnell remembered seeing the offender put the gun away.

The knife then appeared.

This produced a transition from threatened violence to intimate physical violence:

Hartnell described the knife as:

• long;
• comparable to a bread knife;
• with a blade he estimated near twelve inches;
• relatively narrow;
• a hardwood handle;
• brass rivets;
• cotton surgical tape wrapped around part of the handle;
• and a sheath or case he thought might have been wooden. [5]

No knife was recovered in the report archive.

The description remains a victim account, not a physical identification.

 

13. When Language Ends

The signed dialogue reconstruction stops before the stabbing.

Hartnell marked the final remembered verbal exchange as the moment when the offender demonstrated that the gun was loaded.

The violence then began without a clear verbal transition.

Hartnell initially experienced the first impact almost as a blow before understanding that he was being stabbed. He described repeated wounds to his back.

He pretended to be dead.

The offender then moved toward Shepard.

Hartnell remembered:

• repeated stabbing to her back;
• Shepard turning over;
• further stabbing to the front and lower abdomen;
• screaming;
• and conduct he perceived as more violent or “wild” toward Shepard than toward him. [7]

Words such as “frenzy” appear in officer summaries of Hartnell’s perception. They should be attributed to the witness account rather than treated as clinical determinations of the offender’s mental state.

The wound counts also vary.

Table 6. Recorded Estimates of Bryan Hartnell’s Stab Wounds

Source	Hartnell wound estimate
Collins–Land initial reconstruction	Approximately ten
Hartnell interview	Approximately six, partly based upon what medical personnel told him
Surgical record	Not included in the archive

 

These are not necessarily irreconcilable. A victim stabbed in the back cannot reliably count every wound during the assault. Initial officers were also reconstructing events while the victims were receiving emergency care. The medical record, where available, must remain authoritative.

 

14. Cecelia Shepard’s Surviving Voice

Cecelia did not survive to give the prolonged formal interview later obtained from Hartnell. Her surviving voice is dispersed across the accounts of rescuers, rangers, medical personnel and investigating officers. Through those relayed accounts, she is reported as describing the hood, its dark clip-on sunglasses, a black pistol and the offender’s demand for money or transportation. Separate report fields record that property was not ultimately taken. Any statement concerning sexual assault should be attributed to its exact reporting source rather than folded automatically into Cecelia Shepard’s direct voice.

She reportedly said or communicated that:

• the hood prevented her from seeing the offender’s face;
• dark clip-on sunglasses covered the eye openings;
• the offender carried a black pistol;
• he asked for money;
• no money was ultimately taken;
• and she had not been sexually assaulted. [2]

These are not all direct police quotations.

Their transmission frequently followed this path:

Each surviving statement must therefore be identified by speaker, listener, recorder and date.

The incompleteness of Cecelia Shepard’s documentary voice should not cause her to disappear behind Bryan Hartnell’s surviving testimony.

She recognized the gun first. She participated in the binding sequence under coercion. She attempted to help free Hartnell after the attack despite her own wounds. She remained conscious long enough to communicate fragments that became part of the official record.

15. Survival After the Offender

After the assailant left, Hartnell began the second struggle of the evening: remaining alive and obtaining help. [5].[7],[9]

He and Shepard attempted to free themselves. Hartnell recalled trying to help Shepard while also struggling with his own bindings and loss of blood.

A boat remained offshore.

Hartnell remembered:

• the boat moving or circling;
• coming within roughly one hundred yards;
• switching off its motor;
• remaining for an extended interval;
• and failing to approach immediately despite the victims’ calls.

He suggested that the occupants may have feared that an armed person was still nearby.

Eventually he began moving toward the road.

His account records:

• repeated collapse;
• temporary loss of visual awareness;
• short periods of movement followed by sitting or falling;
• severe blood loss;
• religious faith;
• thoughts of his parents;
• and the determination that he had to keep himself alive.

This is not ancillary material.

The offender’s self-authored mythology tends to dominate the public memory of Lake Berryessa. Hartnell’s movement toward the road restores a different centre of agency.

16. The Rescue Network

The rescue was not one single arrival.

It developed through reports, boat travel, road travel and radio coordination.

Ronald Fong heard calls coming from the lake and eventually verified that a man and woman required help. He proceeded to Rancho Monticello Resort and alerted others.

Archie and Elizabeth White travelled toward the scene with Ranger William White and Fong. Shepard was found near the shoreline, bloodied and in severe pain. Mrs. White attempted to comfort her while help was being summoned. [2]

Park Ranger Dennis Land’s affidavit provides the clearest direct account of Hartnell’s recovery near the road. [9]

Land reported receiving a radio call at approximately 7:00 p.m. while patrolling near Pope Creek Bridge. The response was divided:

• one responder proceeded by boat;
• Land proceeded by vehicle;
• another unit was requested.

Land encountered Hartnell in the grass beside the road, calling for help.

Hartnell reportedly told him: “Don’t bother with me, my girlfriend is hurt much worse.”

Land helped Hartnell into the truck. Hartnell pointed toward the lakeside tree where Shepard remained. After Land moved the truck toward the location, Hartnell collapsed. Land then coordinated radio traffic and directed officers and the ambulance toward the remote area.

The rescue sequence was therefore spatially distributed:

The recorded times vary slightly among ranger and sheriff sources. Approximately 6:55 p.m. and 7:00 p.m. appear in different reports.

Those figures should be treated as separate recorded times, not artificially harmonized.

 

Figure 5. The Survivor Leads Rescuers Back. Park Ranger Dennis Land’s affidavit records his discovery of Hartnell beside the road, Hartnell’s insistence that Cecelia was hurt worse and his attempt to direct rescuers toward the shoreline tree before collapsing. The affidavit restores the rescue as a survivor-directed process rather than a passive discovery.
Sources: LB-NCSD-WR-024 and LB-NCSD-WR-025. Original files: [02 Witness Report 24] Dennis Land statement P1.jpeg; [02 Witness Report 25] Dennis Land statement P2.jpeg.

 

17. The Telephone Call

At 7:40 p.m., Napa Police Officer David Slaight received a call on line one of the department switchboard. [3]

He described the caller as young-sounding, calm and possibly in his early twenties.

The caller said:

“I want to report a murder, no, a double murder.”

He identified the location as approximately two miles north of Park Headquarters and referred to a white Volkswagen Karmann Ghia.

When Slaight asked where the caller was, the man replied in a voice described as barely audible:

“I’m the one that did it.”

The receiver was then placed down without being hung up.

The self-correction is significant:

At the time of the call, both victims were alive.

The caller was not reporting a medically established fact. He was imposing his intended classification upon the event.

The call functioned simultaneously as:

• location report;
• authorship claim;
• narrative correction;
• and attempted control of the crime’s meaning.

17.1 The open line

Slaight could hear traffic passing near the abandoned telephone. He also tentatively recalled possible feminine voices in the background, but immediately qualified this impression because he was simultaneously contacting the Sheriff’s Office and transmitting by radio.

The possible voices are therefore a weak, qualified auditory impression—not proof that women stood beside the caller.

The telephone operator reportedly stated that:

• the call originated from a twenty-five-cent payphone;
• and the caller had refused to provide the originating number.

Slaight inferred that the man may have contacted the operator and requested connection to the Police Department.

That calling route is an inference from the operator’s information, not something Slaight directly witnessed.

 

18. The Payphone as a Crime Scene

Sergeant Hal Snook reached the Napa Car Wash payphone at Main and Clinton Streets after officers had secured the area. [10]

The receiver was found:

• off the hook;
• resting on the booth shelf;
• with the mouthpiece beneath the telephone unit;
• and the earpiece projecting into the booth.

The physical position corresponded with what Slaight had heard: the line remained open because the receiver had been placed down rather than hung up.

The booth and surrounding area were photographed.

Snook collected thirty-five latent lifts from the telephone area. Heavy moisture remained on the surfaces when the booth was processed.

No clear tire or footwear impressions were recovered from the blacktop, and Snook observed no physical indication that a vehicle had departed at high speed.

That negative observation does not establish whether the caller:

• walked;
• drove slowly;
• entered a vehicle elsewhere;
• or had no vehicle nearby.

It establishes only that the processed blacktop did not preserve a clear high-speed departure trace.

 

Figure 6. Telephone Authorship: The Heard Call and the Recovered Receiver. Officer David Slaight’s account preserves the caller’s self-corrected report of “a murder—no, a double murder” and the barely audible authorship claim. Sergeant Hal Snook’s evidence report then records the receiver resting off the hook on the booth shelf. Spoken account and physical scene converge: the line remained open because the receiver had been placed down rather than hung up.
Sources: LB-NCSD-WR-001 and LB-NCSD-ER-001. Original files: [02 Witness Report 01] Slaight report on Zodiac phone call.jpg; [04 Evidence Report 01] Snook report P1.jpg.

 

19. The Karmann Ghia and the Written Claim

The victims’ white Karmann Ghia stood in the roadside parking area.

On the passenger door, investigators found:

• the crosshair symbol;
• “Vallejo”;
• 12-20-68;
• 7-4-69;
• Sept. 27, 69 – 6:30;
• and “by knife.”

The writing was made in black ink with what Snook estimated to be a felt-type instrument approximately three-sixteenths of an inch wide at the tip. [10]

The car door did more than record the current attack.

It retrospectively assembled a sequence:

The inscription linked a remote Napa County stabbing to two earlier Vallejo-area attacks before full interagency comparison had been completed.

The offender turned the vehicle into an index.

The telephone call then converted that index into a spoken claim.

19.1 The whole door became evidence

The car-door inscription was not preserved only through photography.

On September 29, the vehicle was transferred to the Napa County Corporation Yard. Snook removed the entire passenger door and entered it into evidence.

The preservation sequence was:

The present location or ultimate disposition of the original door is not established in the supplied reports.

 

Figure 7. The Vehicle as Index: Symbol, Dates, and Method. Snook’s physical-evidence report transcribes the crosshair, “Vallejo,” the two earlier crime dates, the Lake Berryessa date and time, and “by knife.” The inscription did not merely claim the current attack; it organized three events into an offender-authored sequence.
Source: LB-NCSD-ER-002. Original file: [04 Evidence Report 02] Snook report P2.jpg.

 

 

20. The Physical Corridor

At approximately 12:20 a.m. on September 28, Snook arrived at the parking area.

A foot stile crossed the fence approximately fifteen feet south of the victims’ vehicle.

Footprints were observed leading from the stile toward the passenger side of the Karmann Ghia, where the inscription had been written.

One principal impression measured approximately:

• thirteen inches in length;
• four and one-half inches at the widest part of the sole;
• with the tread inset from the outer edge.

Sergeant Butler photographed the impression and documented the interpreted direction of travel. Snook produced a plaster cast.

At daylight, officers traced related impressions:

• between the parking area and hillside;
• near the bottom of the first hill;
• at a trail and dirt-road junction;
• near the location where Hartnell had been found;
• and in sand approximately one hundred feet from the attack site.

The prints were photographed, measured and cast.

Cardboard covers were positioned over some impressions, with orientation preserved to indicate the suspected direction toward the shoreline.

This creates a material corridor:

 

The reports strongly indicate that investigators believed a continuous or related tread pattern linked the lakeside area to the victims’ vehicle.

The archive does not include the full technical comparison report necessary to prove that every preserved impression came from one shoe.

 

21. Tire Impressions

Snook documented at least two tire-impression patterns approximately twenty feet behind the Karmann Ghia.

One measured approximately four and one-half inches in width and showed a comparatively parallel design.

Another measured approximately five and one-half inches and contained a central straight element with herringbone-like features.

A distance of approximately fifty-two inches was recorded between corresponding tire positions. The impressions were photographed and cast. Their existence does not automatically connect them to the offender. The parking area was a public or semi-public roadside location. Pre-existing tracks were possible.

Nor can the impressions presently be connected to the light-blue Chevrolet observed by the three women.

The two fields must remain separate:

Only a later vehicle comparison could join them. No such conclusive comparison appears in the supplied archive.

 

22. Soil, Bottle and Contextual Evidence

Snook collected soil from three locations:

 

Table 7. Soil-Sample Collection Points in the Lake Berryessa Evidence Corridor

Sample	Location
1	Point where the attack occurred
2	Behind the tree where the offender was believed to have prepared or adjusted the hood
3	Parking area beside the Karmann Ghia

 

The collection shows that investigators attempted to preserve the spatial transitions of the event. The reports do not contain the results of a laboratory comparison among the three samples.

A green bottle was also recovered near a stump on the suspected approach route. It was processed for latent impressions and entered into evidence. Its location justified collection. Its collection does not prove that the offender handled it.

This distinction applies throughout the archive:

An object found within a suspected route is contextually relevant, but it does not become offender-derived evidence without additional attribution.

 

23. The Aerial View

Snook recorded that aerial photographs of the crime scene were ordered.

Napa Register photographer Robert McKenzie took the images from an aircraft piloted by Special Deputy Harold Moskowitz of the Napa County Sheriff’s Department Aerial Squadron. [10]

Those photographs are not included in the assembled corpus. Their documented existence is nevertheless significant. They may have recorded:

• the parking area;
• the foot stile;
• hillside and trail;
• shoreline tree;
• attack point;
• route toward the road;
• and relationship between water, road and terrain.

The aerial series would be among the highest-priority visual recoveries for any later edition of this volume.

24. The Footwear Investigation

The first footwear inquiry was local.

On September 29, an officer canvassed twelve Napa shoe businesses in an effort to identify the tread design. [12]

The preliminary retail-level opinions were that:

• the shoe or boot might have been foreign-made or imported;
• the estimated size was approximately 10½ EE;
• and the sole and heel might have been neoprene.

These were working impressions, not laboratory conclusions.

The investigation later shifted toward Air Force footwear.

A retired Air Force master sergeant showed investigators a chukka-style or “wing-walker” shoe whose sole design was considered similar to the plaster-cast pattern.

Investigators travelled to Travis Air Force Base, where:

• base security assisted;
• the Office of Special Investigations participated;
• a comparison shoe was located;
• and procurement and distribution procedures were examined. [2]

The reports established that the footwear was associated with Air Force issue and aircraft-related work.

They also widened the availability field:

• civilian base employees could wear such footwear;
• surplus sales had distributed large numbers of pairs;
• and complete size-by-purchaser records were not available.

Therefore:

The footwear supports possible institutional access.

It does not prove military status, base assignment or a specific wearer.

 

Figure 8. From Unknown Tread to Air Force–Associated “Wing-Walker” Footwear. The investigation moved from a local retail canvass—where the impression was tentatively described as a possible imported, neoprene-soled shoe of approximately size 10½ EE—to an Air Force-associated chukka or “wing-walker” design. Subsequent inquiry also documented civilian and surplus availability, preventing the tread from functioning as proof of active-duty military status.
Sources: LB-NCSD-ER-006, LB-NCSD-IR-018, LB-NCSD-IR-019. Original files: [04 Evidence Report 06] Investigation into boot prints.jpg; [01 Initial Report 18] Initial report by Narlow P10.jpg; [01 Initial Report 19] Initial report by Narlow P11.jpg.

 

 

25. Evidence Custody

The reports allow several evidence chains to be reconstructed.

25.1 Shoreline property

Ranger Dennis Land initially received or secured items left at the scene, including:

• blanket;
• Hartnell’s shoes;
• wallet;
• eyeglasses;
• playing cards;
• plastic clothesline;
• and other property.

These were transferred to Sergeant Snook for preservation.

25.2 Hospital property

Items transported from Queen of the Valley Hospital included:

• Hartnell’s shoes and socks;
• trousers, shirt, T-shirt, shorts and belt;
• wallet, glasses, key and playing cards;
• Shepard’s dress, slip, panties and bra.

These passed through officers Allan Brambrink and William Munk into the Sheriff’s evidence locker and then into Snook’s custody.

25.3 The vehicle

The Karmann Ghia was:

• examined at the roadside;
• photographed;
• processed for latent impressions;
• moved to a secured building at Park Headquarters;
• transferred to the Napa County Corporation Yard;
• and partially dismantled when the passenger door was removed for evidence.

The vehicle’s movement requires ordinary chain-of-custody scrutiny. The surviving reports document security precautions, but do not preserve every handling detail. The reports, however, document security precautions and do not by themselves establish mishandling.

 

26. The Latent-Impression Field

Snook collected latent impressions from several sites.

The numbering appears to include:

• payphone lifts approximately 1–35;
• Karmann Ghia lifts approximately 36–43;
• bottle lifts approximately 44–48.

The California Department of Justice later recorded receiving thirty-five cards bearing forty-five latent lifts. [14]

The difference between the local numbering reaching at least forty-eight and the forty-five lifts delivered to the state laboratory is unresolved.

Possible explanations include:

• some numbers unused;
• some lifts rejected locally;
• multiple lifts mounted on one card;
• or selected submission rather than full submission.

The reports do not establish which explanation is correct.

 

27. California Department of Justice Examination

California DOJ personnel conducted a formal field and laboratory review in early October. [14]

The materials included:

• thirty-five cards bearing forty-five latent lifts;
• a photograph of a latent impression;
• inked comparison prints from several named persons;
• a photograph of the automobile door;
• printed comparison material;
• and a card described as a “Charlie Brown” card bearing printing.

The latter item requires caution.

The report does not state:

• who authored the card;
• where it originated;
• whether it was questioned evidence or known comparison writing;
• or whether investigators considered it a Zodiac communication.

It must not be conflated with later Zodiac greeting-card material.

The safe conclusion is only this: A “Charlie Brown” card bearing printing formed part of the October 1969 material transferred to the California DOJ Questioned Documents Unit.

The door photograph and printing samples were delivered to questioned-document examiner Sherwood Morrill. The full Morrill report is not included in this archive.

 

28. State Latent-Print Results

The California DOJ examiner compared usable latent impressions with the submitted comparison prints. [14]

No identification was accomplished with the submitted comparison prints.

Several impressions appeared to consist of portions of palms. Standard records did not contain all corresponding palm areas, preventing complete comparison.

The result should therefore not be paraphrased as:

“All suspects were conclusively excluded.”

The more accurate formulation is:

No submitted individual was identified as the source of the usable impressions with the comparable print material then available; some palm and peripheral finger areas could not be conclusively compared.

This distinction between no identification and complete exclusion is essential.

 

29. The FBI Examination

On October 23, the FBI Identification Division reported upon five photographs of selected latent impressions. [15]

The FBI found:

• seven latent fingerprints;
• three latent palm prints;
• and one impression that could be either a fingerprint or a partial palm print;

all possessing sufficient value for possible identification.

This is not equivalent to saying that the impressions came from the offender.

It means that the photographs contained usable ridge detail.

The FBI compared the Lake Berryessa material with:

• available fingerprints from several named individuals;
• and comparable latent impressions in the Vallejo Times-Herald extortion case.

No identification was effected.

Again, the FBI noted limitations involving:

• fingertips;
• sides of fingers;
• lower finger-joint areas;
• and palm regions not contained in ordinary fingerprint records.

Additional exemplars would be required for conclusive comparisons in some cases.

 

30. The Vallejo Times-Herald Comparison

The FBI compared the Lake Berryessa latent impressions with comparable latent material in the case described as [15]:

“Unknown Subject; Vallejo Times Herald, Vallejo, California, Extortion.”

No identification was made.

That result means:

The selected Lake Berryessa impressions were not identified as originating from the same source as the comparable Times-Herald impressions.

It does not establish that different offenders produced the two crimes.

Either latent field could contain prints from:

• legitimate users;
• employees;
• responders;
• vehicle occupants;
• journalists;
• or unrelated members of the public.

Partial prints may also fail to overlap in comparable areas even when produced by the same person.

Negative latent comparison is not equivalent to offender separation.

31. Elimination Problems

The FBI could not locate identification records for Hartnell or Shepard. [15]

Unless complete local elimination prints were separately obtained, the federal file could not determine whether particular impressions belonged to:

• Hartnell;
• Shepard;
• a prior telephone user;
• a mechanic;
• a ranger;
• an officer;
• a passenger;
• a resort employee;
• or the offender.

This is the central limitation of the latent evidence.

A relevant-location print is not automatically an offender print.

Attribution requires:

1. usable ridge detail;
2. location and deposition context;
3. elimination of legitimate users;
4. and a matching exemplar.

The archive proves that some impressions satisfied the first condition.

It does not prove that every other condition was satisfied.

 

32. The Reports as Layers of Knowledge

The Lake Berryessa archive repeatedly demonstrates that statements must be classified by how they entered the file.

Direct observation

Examples include:

• Slaight hearing the caller;
• Land finding Hartnell beside the road;
• Snook observing the receiver off the hook;
• Snook measuring footprints;
• officers seeing the inscription.

Direct witness statement recorded by an officer

Examples include:

• Hartnell’s taped interview;
• individual statements from the three women;
• responder affidavits.

Relayed statement

Examples include:

• Cecelia’s words reported through rescuers;
• Hartnell’s emergency comments repeated by rangers;
• medical information relayed to investigators.

Officer synthesis

Examples include:

• Collins and Land’s initial reconstruction;
• Robertson’s three-page summary;
• Narlow and Lonergan’s compiled chronology.

Investigator inference

Examples include:

• interpreting footprint direction;
• associating a tree with costume preparation;
• judging one observed man unlikely to have crossed the terrain on foot;
• considering a shoe design military-associated.

These levels must not be silently collapsed.

 

33. Memory Before Canon

The complete Hartnell interview opens with Hartnell himself expressing concern about repetition. Many people had already asked him to recount the attack. He understood that repeated telling could create small variations. This is one of the most important methodological statements in the archive. The canonical account did not arise from bad faith. It arose from the normal processes through which traumatic experience becomes documentary memory:

• sedation;
• pain;
• repeated questioning;
• interviewer prompts;
• correction;
• summary;
• later reconstruction;
• public quotation.

The Hartnell materials exist in at least three major forms.

Table 8. Three Documentary Forms of Hartnell’s Recollection

Source	Character
Twelve-page interview transcription	Extended, hesitant, self-correcting, interviewer-visible
Robertson’s three-page report	Compressed administrative summary
Signed dialogue reconstruction	Structured attempt to recreate offender–victim speech

 

Each has a different value.

The transcript best preserves uncertainty.

The summary best preserves investigative organization.

The signed dialogue best preserves conversational order as Hartnell later reconstructed it.

None is a stenographic recording of the shoreline encounter.

 

Figure 9. Three Documentary Forms of One Memory: Transcript, Officer Summary, and Signed Reconstruction. Hartnell’s account survives in three distinct documentary forms: an extended interview preserving hesitation and correction; an officer’s compressed narrative summary; and a structured dialogue reconstruction. Each preserves a different property of memory, and none is a stenographic recording of the shoreline encounter.
Sources: LB-NCSD-WR-009, LB-NCSD-WR-004 and LB-NCSD-WR-007. Original files: [02 Witness Report 09] 1969-09-28 Bryan Hartnell - John Robertson Interview-01.jpg; [02 Witness Report 04] Initial interview with Hartnell P1.jpg; [02 Witness Report 07] Hartnell + Zodiac dialogue transcript P1.jpg.

 

34. Source Variations

The archive contains genuine variations that should remain visible.

 

Table 9. Principal Source Variations and Evidentiary Treatment

Issue	Version A	Version B	Evidentiary treatment
Prison location	Deer Lodge, Montana	Colorado in Ranger White’s relayed account	Detailed Hartnell sources favour Montana/Deer Lodge; preserve variation
Exact prison name	Hartnell struggles during interview	Signed dialogue gives “Deer Lodge” cleanly	Canonical wording stabilizes across documentary stages
Guards killed	One guard	A couple of guards	Number unresolved
Nervousness	Offender says he is nervous	Hartnell asks; offender agrees	Behaviour stable, wording variable
Hartnell wounds	Approximately ten	Approximately six	Medical record required
Gloves	Reportedly mentioned to rescuers	Hartnell cannot directly remember	Possible, not confidently observed
Firearm	Black automatic pistol	No later pistol identified	Type broad, model unidentified
Calibre	Possible .45	Hartnell uncertain	Tentative ammunition resemblance only
Vehicle year	Narrative descriptions vary	Tow form contains an apparently inconsistent entry	Registration documentation required
Three-woman suspect	Investigatively significant	Not physically linked to offender	Preserve as possible observation field

 

The aim is not to select whichever version appears most dramatic.

The aim is to identify:

• what remains stable;
• what changes;
• and why the documentary route matters.

 

35. From Shoreline Crime to Multi-Agency File

Lake Berryessa did not remain an isolated Napa County stabbing investigation.

The transformation began with the car door. The dates and Vallejo reference linked the attack to Lake Herman Road and Blue Rock Springs.

The telephone call claimed authorship.

Within days, Napa investigators were exchanging information with Vallejo Police Department personnel concerning the July 4 attack, suspects and the earlier coded communication. [2]

The investigation expanded through:

• Napa County Sheriff’s Department;
• Napa Police Department;
• park and ranger personnel;
• Queen of the Valley Hospital;
• California Highway Patrol;
• Travis Air Force Base;
• Air Force Office of Special Investigations;
• California Department of Justice;
• FBI Identification Division.

The event became institutional before it became fully public.

The canonical event was formed through this administrative expansion.

 

36. Funeral Surveillance

Narlow and Lonergan attended Cecelia Shepard’s funeral and graveside service.

Other officers positioned themselves outside the church and photographed persons entering and leaving, both individually and in groups. [2]

The reports do not state that the surveillance identified the offender or produced a viable suspect.

Its importance lies elsewhere.

The investigation had already moved:

• from shoreline;
• to hospital;
• to autopsy;
• to funeral;
• and into the observation of the community surrounding the victim.

This expansion is a reminder that the file contained real families, friends, classmates and mourners—not only offender evidence.

The absence of a documented result should remain explicit:

Funeral surveillance occurred; no investigative outcome is preserved in the supplied reports.

 

37. Cleared and Unresolved Leads

The reports record persons considered because of:

• prior relationships;
• alleged threats;
• physical resemblance;
• absence from institutions;
• military history;
• psychiatric history;
• or access to relevant locations.

The existence of a name in a police file does not establish guilt.

At least one preliminary suspect was interviewed, checked against voice description, movement, alibi and vehicle access, and cleared.

A former acquaintance of Shepard was investigated after her parents reported a prior threat. The supplied pages do not connect that person to the Lake Berryessa attack.

The archive’s treatment of such individuals should follow a simple rule:

Cleared persons should remain cleared unless independent new evidence exists.

Unresolved leads should remain unresolved rather than being promoted through repetition.

 

38. What the Archive Establishes

The completed report corpus establishes with substantial documentary support that:

• Hartnell and Shepard were attacked near the shoreline on September 27, 1969;
• the offender used a hooded costume displaying a white crosshair emblem;
• a firearm was used to establish control;
• plastic clothesline was used for binding;
• the victims were stabbed after being restrained;
• Hartnell survived and moved toward the road to obtain help;
• Cecelia remained alive through rescue but later died;
• an inscription linked the attack to earlier Vallejo crimes;
• the passenger door was removed and preserved as evidence;
• a telephone caller claimed responsibility;
• the payphone receiver was recovered off the hook;
• dozens of latent impressions were collected;
• footprints and tire impressions were photographed and cast;
• soil and contextual objects were collected;
• aerial photographs were ordered and taken;
• military-associated footwear became an investigative lead;
• Hartnell could not identify a specific firearm;
• state and federal latent comparisons produced no identification;
• and no examined print conclusively identified the offender.

 

39. What the Archive Does Not Establish

The corpus does not establish:

• the offender’s identity;
• that the three women’s observed man was the assailant;
• that the light-blue Chevrolet belonged to the assailant;
• that the cast tire impressions came from that Chevrolet;
• that every footprint interpreted as related came from one shoe;
• that the offender was active-duty military;
• that the weapon was conclusively a .45;
• that the green bottle belonged to the offender;
• that any recovered latent impression belonged to the offender;
• that the Lake Berryessa prints matched another Zodiac file;
• that all named persons were conclusively excluded by prints;
• that the “Charlie Brown” card was offender-authored;
• or that every collected specimen received complete laboratory examination.

These are not weaknesses to be hidden.

They are the boundaries that keep the archive evidentiary rather than mythological.

 

40. The Formation of a Canonical Crime

Lake Berryessa became canonical through several different forms of authorship.

Costume authorship

The hood transformed the offender’s body into a controlled image.

Narrative authorship

The prison-escape and robbery story attempted to define the encounter for the victims.

Violence as transition

The knife revealed that the robbery story had been a mechanism rather than an end.

Inscription authorship

The car door placed the attack within a sequence of earlier crimes.

Telephone authorship

The caller declared the event a “double murder” and claimed responsibility.

Institutional authorship

Police reports, interviews, measurements, laboratory reports and interagency comparisons organized the event into a case.

Canonical authorship

Later retellings simplified these layers into a stable public narrative.

The final canon is therefore not merely the offender’s creation.

It is the result of an interaction among:

• offender performance;
• victim memory;
• witness observation;
• police procedure;
• forensic limitation;
• media repetition;
• and later historical selection.

 

41. By Space

The Lake Berryessa archive contains several different spaces that must remain analytically distinct.

The observation space

The area occupied by the three women and the unidentified man near the light-blue Chevrolet.

The resting space

The shoreline tree selected by Hartnell and Shepard.

The concealment space

The tree behind which the offender was first observed and may have prepared the hood.

The control space

The blanket and immediate area where the victims were bound.

The violence space

The location of the stabbing.

The survival space

The route Hartnell attempted to follow toward the road.

The physical-evidence corridor

The trail, hillside, stile, parking area and passenger door.

The authorship space

The Napa Car Wash telephone booth.

The institutional space

Hospital, sheriff’s office, corporation yard, state laboratory, FBI and allied agencies.

The event did not occur at one “scene.”

It moved through a chain of scenes.

 

42. By Time

The reports also contain several forms of time.

Event time

The afternoon movements, attack, rescue and telephone call.

Witness-estimated time

Durations remembered under fear, blood loss and shock.

Dispatch time

Recorded calls, assignments and arrivals.

Report time

The dates on which officers wrote or typed their accounts.

Laboratory time

The later processing, comparison and reporting of evidence.

Canonical time

The order in which later histories present the event.

These forms should not be confused.

The thirteen-page Narlow–Lonergan report, for example, describes events from September 27 through October 3 but was typed in stages on October 4 and 5.

It is a retrospective investigative synthesis, not a continuously written field diary.

Likewise, Hartnell’s signed dialogue is not the attack conversation itself. It is a later reconstruction of remembered speech.

 

43. By Motif

Several motifs emerge from the police archive.

Concealment and display

The hood concealed the face while displaying the emblem.

False purpose

The robbery story announced one motive while behaviour enacted another.

Binding and release

The victims were immobilized; Hartnell’s survival depended on reversing that condition.

Writing on transport

The vehicle used to reach the location became the surface upon which the attack was indexed.

Voice without face

The caller removed the costume but preserved anonymity through distance.

Footprint without identity

The earth retained form without naming the wearer.

Fingerprint without attribution

Ridge detail survived, but its depositor remained unknown.

Report without seamlessness

The archive accumulated knowledge without eliminating every contradiction.

 

44. Conclusion: The Reports Beneath the Hood

The hood is the central visual object of Lake Berryessa.

It is also a metaphor for the way the crime has been remembered.

The public narrative covers the file as the costume covered the offender: it creates a strong, simplified image while hiding the construction beneath it.

Beneath the hood are the reports.

Beneath “Deer Lodge” is a sequence of uncertain recollection, prompting, summary and later consolidation.

Beneath “the .45” is a black automatic pistol that Hartnell could not identify and ammunition he only tentatively associated with .45 calibre.

Beneath “the military boot” is a distinctive tread first thought possibly foreign or neoprene, later associated with Air Force footwear, civilian use and surplus sale.

Beneath “the fingerprints” are useful latent impressions that could not be attributed and could not always be conclusively compared.

Beneath “the suspect seen by three girls” is an important but unproven observation field.

Beneath “the double murder” are two victims who were still alive when the caller imposed that description upon them.

Beneath the offender’s performance is Hartnell’s counter-performance: conversation, delay, attempted cooperation, resistance to despair and movement toward the road.

Beneath the file is Cecelia Shepard, whose surviving voice exists in fragments passed through rescuers and officers.

The Lake Berryessa archive does not give us one perfect story.

It gives us something more valuable:

• an event remembered under sedation;
• dialogue reconstructed after trauma;
• terrain converted into measurements;
• a car converted into a document;
• a telephone converted into a second crime scene;
• evidence transferred through local, state and federal institutions;
• and repeated attempts to produce identity from traces that remained incomplete.

The reports do not resolve Lake Berryessa.

They reveal how resolution was sought.

The crime became canonical not because every detail was known, but because costume, inscription, voice and institutional linkage made the event impossible to contain within one shoreline.

The offender attempted to author the event through symbol, violence, writing and telephone.

The archive answers with another form of authorship.

It records the witnesses who saw.

The survivor who remembered.

The rescuers who arrived.

The officers who measured.

The laboratories that compared.

And the limits they were obliged—or should have been obliged—to preserve.

The hood created an image.

The reports preserved the event beneath it.

 

Epilogue

Reversing the Scales

This volume should be read beside the blog’s earlier study, Canonical Murder 3: Lake Berryessa, Volume 1 — “Byknife on the Door of Bryan Hartnell Car” [16]. Volume 1 approached the Karmann Ghia inscription as an authored surface: the place where the offender joined dates, locations, symbol, method and claim into a single visual ledger. The present volume has moved in the opposite direction. It has passed beneath that inscription and beneath the hood itself, returning to the reports, voices, uncertainties, measurements and institutional processes from which the received Lake Berryessa narrative was formed.

The relationship between the two volumes may therefore be expressed as a reversal of perspective:

There is also a quieter and more tragic date within the police archive.

On October 2, 1969, Cecelia Shepard’s funeral was held at the Pacific Union College Church. Investigators attended the service, observed those entering and leaving, and photographed the mourners as part of the continuing search for her killer.

That same day was the first birthday to pass after David Arthur Faraday’s murder.

He would have turned eighteen. [17]

The coincidence is calendrical, not causal. Nothing in the surviving reports establishes that the offender knew of it, selected it or arranged the timing around it. Yet once the two victim chronologies are placed beside one another, the date carries an unavoidable human weight:

On the day Cecelia Shepard was mourned, David Faraday’s eighteenth birthday passed without him.

One family buried a daughter.

Another family reached the first birthday their son would never live to see.

October 2 therefore belongs neither to the offender nor to the mythology that later accumulated around him. It belongs first to two young victims, to their families and to the communities forced to measure time through absence.

Within this project’s tropical-zodiac reference framework, Lake Berryessa and the subsequent Presidio Heights murder both occurred during Libra, the calendrical sign represented by the scales. Within the interpretive architecture developed elsewhere in this project, the sequence can be read as a distorted judgment field: Lake Berryessa through costume, binding, inscription and the car-door ledger; Presidio Heights through material possession, the shirt fragment and the attempted certification of authorship. This correspondence is documented as a calendrical and symbolic structure; deliberate astrological selection by the offender remains unproven.

Yet the metaphor of the scales is still appropriate because the offender repeatedly attempted to place himself above the lives he attacked.

He assigned dates.

He counted victims.

He classified methods.

He declared two living people dead.

He joined crimes into a ledger of his own construction.

He attempted to become judge, recorder, executioner and archivist at once.

These were not the scales of justice.

They were scales deliberately weighted by violence, secrecy and self-authorship.

At Lake Berryessa, the imbalance was written upon a car door. At Presidio Heights, it would be carried away in the form of Paul Stine’s bloodstained shirt. In both cases, the offender tried to convert the suffering of another human being into evidence of his own authority.

The responsibility of the archive is to reverse that operation.

To reverse the scales does not mean inventing a counter-mythology. It does not mean assigning intention where only correspondence survives. It means restoring proper evidentiary weight.

The offender’s claims must be weighed against the physical record.

Officer summaries must be weighed against direct testimony.

Canonical quotations must be weighed against the interviews from which they emerged.

Suspect names must be weighed against documented disposition.

Symbolism must be weighed against chronology.

Interpretation must be weighed against proof.

And the offender’s theatrical presence must be weighed against the humanity of those whose lives he attempted to reduce to dates, totals and methods.

When that reversal is made, Lake Berryessa changes.

The hood no longer occupies the centre.

Bryan Hartnell’s survival does.

The car-door inscription no longer speaks alone.

Cecelia Shepard’s fragmented surviving voice speaks beside it.

The telephone confession no longer determines the meaning of the event.

The rescuers, witnesses, officers, photographs, casts, measurements and unresolved impressions answer it.

The offender attempted to transform murder into judgment.

The archive must transform judgment back into evidence.

The scales must be lifted from his hands.

They belong to the truth.

They belong to justice.

And, above all, they belong to the dead and the living whose lives he tried to weigh as though they were his to measure.

 

SDS

August 4, 2026

57 years since the “Zodiac Debut” letter

Acknowledgment

This volume gratefully acknowledges Tom Voigt and ZodiacKiller.com for his pioneering role in bringing the Lake Berryessa police reports into broad public online circulation. By collecting and publishing the report sequence in an accessible page-by-page format, ZodiacKiller.com gave independent researchers direct access to primary investigative material that had previously remained difficult to consult outside official repositories. The presentation also protected personal privacy by removing names, addresses and telephone numbers where appropriate.

The documentary classification, Lake Berryessa reference index, comparative analysis and interpretations developed in this volume are independent. They nevertheless remain indebted to that earlier act of archival access. The author therefore records sincere appreciation to Tom Voigt and ZodiacKiller.com for establishing an enduring public gateway into the Lake Berryessa case file. [18]

 

 

References

[1] Napa County Sheriff’s Department, Incident and Initial Crime Reports, case no. 105907, Sept. 27–29, 1969, reporting officers including David Collins and Ray Land.

[2] K. Narlow and R. Lonergan, Napa County Sheriff’s Department, Supplementary Crime Report, thirteen pages, case no. 105907, events Sept. 27–Oct. 3, 1969; typed Oct. 4–5, 1969.

[3] D. Slaight, Napa Police Department, Supplementary Report Concerning Telephone Call, Sept. 27, 1969.

[4] J. Robertson, Napa County Sheriff’s Department, Supplementary Report Concerning Cecelia Shepard’s Family and Acquaintances, Sept. 28–30, 1969.

[5] J. Robertson, Napa County Sheriff’s Department, Report of Taped Interview with Bryan Calvin Hartnell, interview conducted Sept. 28, 1969.

[6] B. C. Hartnell, Dialogue Between an Unknown Assailant and Bryan Hartnell, signed reconstruction, Napa County Sheriff’s Office records.

[7] J. Robertson and B. C. Hartnell, Twelve-Page Transcription of the Sept. 28, 1969 Hospital Interview, transcribed from tape recording by M. Feurle.

[8] Napa County Sheriff’s Department, Individual and Consolidated Reports Concerning the Three Pacific Union College Women, Sept. 27–29, 1969.

[9] D. Land, Napa County Sheriff’s Office, Two-Page Rescue Affidavit, Lake Berryessa response, Sept. 27, 1969.

[10] H. Snook, Napa County Sheriff’s Department, Supplementary Crime and Physical-Evidence Report, four pages, case no. 105907, Sept. 27–29, 1969; typed Oct. 5, 1969.

[11] California Highway Patrol, Vehicle and Impound Report: Hartnell Karmann Ghia, Sept. 27, 1969.

[12] J. Robertson, Napa County Sheriff’s Department, Napa Shoe-Store Canvass Concerning Plaster-Cast Tread Pattern, Sept. 29, 1969.

[13] J. Robertson, Napa County Sheriff’s Department, Report on Pistols and Ammunition Shown to Bryan Hartnell, Oct. 1, 1969.

[14] California Department of Justice, Bureau of Criminal Identification and Investigation, Latent-Print Examination and Questioned-Document Routing Materials, October 6–8, 1969.

[15] Federal Bureau of Investigation, Identification Division, Latent Fingerprint Section, Report Concerning Lake Berryessa Latent Impressions, FBI latent case no. A-12397, Oct. 23, 1969.

[16] “Canonical Murder 3: Lake Berryessa, Volume 1—‘Byknife on the Door of Bryan Hartnell Car,’ Part 1,” Zodiac Killer Research, July 22, 2025, https://zodiacresearch.blogspot.com/2025/07/canonical-murder-3-lb-vol1-byknife-on.html, accessed August 4, 2026.

[17] Verified birth record, official investigative record or other primary biographical source establishing David Arthur Faraday’s birth on October 2, 1951.

[18] Napa County Sheriff’s Department, Lake Berryessa police-report sequence, case no. 105907, reproduced and made publicly accessible by Tom Voigt, ZodiacKiller.com, thirty-five-page online scan sequence, https://zodiackiller.com/LBReport1.html, accessed August 4, 2026.

Appendix A

Documentary Source Forms and Evidentiary Uses

The categories below are classificatory rather than a fixed ranking of reliability. Evidentiary value depends upon the question being asked, proximity to the event, completeness of transmission and the presence or absence of corroboration.

Table A. Documentary Source Forms: Uses, Strengths, and Limitations

Category	Source form	Example	Principal strength	Principal limitation
1	Original physical evidence	Door, casts, lifts, clothing	Material continuity	Many original objects unavailable
2	Original recording	Hartnell interview tape	Preserves voice and phrasing	Present survival unknown
3	Typed transcription	Twelve-page Hartnell interview	Preserves questions, uncertainty and corrections	Dependent upon transcription accuracy
4	Direct written statement	Land affidavit, witness statements	Identifiable witness and account	Memory and wording still retrospective
5	Officer summary	Robertson report	Clear investigative organization	Compression removes hesitation and prompts
6	Relayed testimony	Cecelia through rescuers	Preserves otherwise lost statements	Multiple transmission stages
7	Signed reconstruction	Hartnell dialogue	Preserves remembered conversational sequence	Not stenographic
8	Investigative synthesis	Narlow–Lonergan report	Integrates chronology and multiple records	Retrospective and source-mixed
9	Laboratory result	DOJ and FBI reports	Technical comparison and documented limits	Attribution depends upon elimination and exemplars
10	Later canonical retelling	Public histories	Accessibility and narrative continuity	Frequently removes uncertainty and variation

 

Appendix B

Principal Documentary Variations

Table B. Principal Documentary Variations and Controlled Conclusions

Question	Surviving forms	Controlled conclusion
What prison was named?	Deer Lodge; Montana compound name; Colorado in one relayed source	Montana/Deer Lodge is best supported, but the exact wording was not uniform
How many guards were allegedly killed?	One; two; “a couple”	Number unresolved
Who introduced nervousness?	Offender volunteered it; Hartnell asked	Hartnell observed shaking; exact dialogue varies
Did the offender wear gloves?	Relayed statement says yes; direct interview uncertain	Possible but not confidently observed
What calibre was the gun?	Possible .45; no pistol identified	Firearm unidentified; .45 resemblance tentative
How many times was Hartnell stabbed?	Approximately six; approximately ten	Final medical count required
Was the three-woman subject the offender?	Possible investigative lead	Not established
Did fingerprints exclude named persons?	No identification; incomplete palm/peripheral comparisons	No match, but not always conclusive exclusion
Did Lake Berryessa prints match another Zodiac file?	No identification with Times-Herald latents	No established latent-print linkage
Did the offender leave by vehicle at the payphone?	No high-speed tire trace	Departure method unknown

 

 

 

Appendix C

Master Chronology

Table C. Master Chronology of the Lake Berryessa Event and Investigation

Date and approximate time	Event
Sept. 27, 1:00 p.m.	Hartnell and Shepard leave Pacific Union College
Early afternoon	Visit St. Helena rummage sale
About 2:00 p.m.	Hartnell collects Shepard
About 2:45 p.m.	Three women enter Lake Berryessa parking area; blue Chevrolet arrives
About 3:00 p.m.	Women reach shoreline
About 3:30–3:50 p.m.	Man observed watching women
About 4:30 p.m.	Women return; Chevrolet absent
About 5:15 p.m.	Hartnell and Shepard reportedly seen arriving
About 6:30 p.m.	Attack
About 6:55–7:00 p.m.	Ranger and sheriff rescue notifications
Shortly afterward	Land finds Hartnell near road; Cecelia located near shoreline
7:40 p.m.	Slaight receives Napa Police call
About 8:40 p.m.	Snook notified
About 8:50 p.m.	Payphone secured
8:50–9:37 p.m.	Victims at Queen of the Valley Hospital; preliminary interviews
11:49 p.m.	Payphone processing completed
Sept. 28, 12:20 a.m.	Snook reaches Lake Berryessa
2:45 a.m.	Karmann Ghia moved to secured building
7:00–8:00 a.m.	Daylight footprint and scene processing resumes
12:30–1:00 p.m.	Robertson interviews Hartnell
About 5:00 p.m.	Hospital property transferred into evidence
Sept. 29	Door and vehicle transferred; passenger door removed
Sept. 29, afternoon	Cecelia Shepard dies
Sept. 30	Autopsy; preliminary cause recorded as shock and blood loss from principal stab wounds
Oct. 1	Firearm/ammunition comparison; Vallejo and Travis AFB inquiries
Oct. 2	Funeral surveillance
Oct. 3	Preliminary suspect interviewed and cleared
Oct. 6–8	California DOJ examination
Oct. 23	FBI latent-impression report issued