There were indeed two people in the Olds, but they were not sleeping. The sight of blood sent their would-be rescuer scuttling back to his own car and racing to town, where he alerted city police. They, in turn, called Sheriff Presley, since the Oldsmobile was parked on county turf, roughly a mile outside the city limits.2
Presley must have dreaded what he would find at the scene. Age fifty at the time, nine months from the end of his statutory two-year term as sheriff, Presley’s public service prior to taking on his present job consisted of acting as commissioner of Bowie County Precinct No. 1 from 1940 to 1942.3 Untrained in the formalities of law enforcement, like most rural sheriffs in the mid-twentieth century, Presley still possessed intelligence and hard-nosed common sense, tempered by lifelong exposure to the violence of his rough-and-tumble home environment.
Presley made the drive to Rich Road with Jackson Neely Runnels, two years his junior, chief of police on the Texas side of Texarkana. Like Presley, Runnels—“Jack” to everyone who knew him—was a Bowie County native, but had more time on the job.4 Arriving on the Rich Road scene, they found a man and woman in the bloodstained Oldsmobile’s rear compartment. The man knelt on the floorboard between the front and back seats, his forehead resting on crossed hands, his pockets turned inside-out in a gesture suggesting robbery. His companion lay facedown on the backseat, atop a blanket, with an open purse beside her. Both were fully clothed. Both had been shot in the back of the head.5
Despite the rain, Presley and Runnels noted more bloodstains outside the car, on sandy loam some twenty feet away. Its quantity suggested that the victims had been slain there, then returned to the Olds by their killer or killers. More blood had flowed from one of the sedan’s rear doors, to stain the running board.6
Presley and Runnels soon identified the victims as Richard Griffin and Polly Ann Moore. Griffin was twenty-nine, a war veteran discharged from the U.S. Navy Seabees (Construction Battalion) in November 1945, who presently resided with his mother at Robison Courts, on Texarkana’s Walnut Street. Moore was a seventeen-year-old graduate of Atlanta (Texas) High School, employed as a checker at Texarkana’s Red River Arsenal since July 1945, living with her cousin, Ardella Campbell, at a boardinghouse on Magnolia Street. The couple had been dating for about six weeks, and no one seemed concerned about the twelve-year disparity in their ages. Witnesses had seen them at a café on West Seventh Street, around 2 P.M. on Saturday, and they returned to another café on the same street that evening, dining until 10 P.M. with Griffin’s sister Eleanor and her date, J. A. Proctor. Now, on Sunday morning, they departed from their last date separately, Griffin bound for the Texarkana Funeral Home, while Moore was conveyed to the Hanner Funeral Home in Atlanta, twenty-four miles distant.7
Confusion surrounds forensic evidence recovered from the Griffin-Moore murder scene. Both victims were shot with a .32-caliber pistol, and since no weapon remained at the scene, authorities ruled out a murder-suicide. Early reports said that each was shot once, while later accounts add a second wound for Griffin, and one—published decades later and marred by frequent errors—claims both were shot twice. Accounts of the ballistics evidence include mention of one, two, and “several” cartridge cases found in or around the Oldsmobile, while an FBI memo says only one slug and two shell casings were recovered. Most sources, including an FBI memo dated May 15, 1946, say that no useful fingerprints were retrieved from the car—yet another FBI document, filed seven months later, requests comparison of a suspect’s fingerprints with “unidentified latents” found at the scene.8
The worst—and possibly deliberate—confusion involves the matter of sexual assault against Polly Ann Moore. Her body was found fully clothed, according to newspaper reports, and an FBI memo states that Moore’s corpse was delivered to morticians for embalming without any tests performed for evidence of sexual assault. Contradicting those claims, an early report in the Texarkana Gazette stated that “[e]xamination of the girl’s body by a local physician revealed that she had not been criminally assaulted”—a common euphemism for rape in those more genteel days. Nonetheless, four documents from Texas Ranger files state categorically that Moore was raped, and further injuries are alluded to in later newspaper reports. Articles from May 1946 describe a “sex maniacal killer,” claiming that Moore had been “horribly mutilated.” Eight years later, reporter John Scudder wrote that Moore had been “horribly mistreated for what is believed to have been about two hours” before she was shot. A test performed on blood recovered from the ground near Griffin’s car reportedly told lawmen only that it “may have been” Moore’s.9
Whatever the precise nature of Moore’s injuries, two things are clear: authorities believed she had been raped, and they concealed it from the public. Such discretion was routine in the 1940s, when newspapers commonly lumped rape and other forms of sexual violence under the generic label of “criminal assault,” taking pains to conceal the identity of victims. As Texarkana Gazette editor John Quentin Mahaffey explained, “It wasn’t reported. You guarded the person who was raped.”10 Overt denial of “assault” to spare a dead victim’s feelings, however, was still peculiar. Perhaps investigators held the information back to screen out future false confessions, but since none ever admitted feeding false reports to the Gazette or other newspapers, such speculation is fruitless.
As published accounts of ballistics evidence varied, so did descriptions of the weapon used to kill Griffin and Moore. It was a pistol, all sources agree, but what kind? Some reports refer specifically to a revolver, which would not eject shell casings at a shooting scene unless the gunman took time to manually remove and discard them, thus deliberately leaving evidence for the authorities. Semiautomatic pistols, on the other hand, eject a shell for each shot fired, which a distracted killer might not bother to collect. One published source says the pistol was a revolver. Another identifies the suspect weapon as a .32-caliber Colt semiautomatic. Confusion of that firearm’s .32 ACP (Automatic Colt Pistol) cartridge with .32-caliber revolver ammunition is impossible, since they are not interchangeable and do not resemble each other except in caliber (the projectile’s diameter).11
Contradicting the Colt identification, we have another FBI memo stating that the murder weapon was “identified as [a] thirtytwo [sic] automatic Spanish type.”12 If true, a likely candidate would be the Ruby semiautomatic, designed in 1914 and chambered for the .32 ACP cartridge (designated 7.65mm Browning in Europe). This pistol was manufactured by Gabilondo y Urresti (now Llama Firearms) and some fifty other companies during the First World War, with production ceasing in the Great Depression of the 1930s.13
Before any specific weapon could be matched to the attack on Moore and Griffin, though, police required a suspect gun to test against the slug and casings they recovered from the murder scene. Guessing the model was a small help, but Colt had manufactured some 570,000 of its .32-caliber Pocket Hammerless pistols between 1903 and 1945. Aside from civilian sales, many had found their way into military service, including a General Officer’s Model carried in the recent war by Generals Omar Bradley, Dwight Eisenhower, and George C. Marshall. Tank commander George Patton preferred the Model 1908, chambered for .380 ACP rounds.14
Spanish Ruby semiautomatics were equally abundant. Gabilondo alone manufactured at least 250,000 during World War I, and other factories matched that pace. France purchased 710,000 of the guns for its military officers during the war, with 580,000 remaining in service by 1920.15 No one knows how many may have traveled to the States with immigrants from Europe, or with U.S. servicemen returning from the two World Wars.
Once authorities obtained a weapon, they could match the rifling from its barrel to the markings on their death slug, and compare extractor marks found on the casings to the pistol’s inner mechanism. But without a gun to study—nothing.
With two bodies on their hands, Sheriff Presley and Chief Runnels knew they needed help. Texarkana’s two mayors—William V. Brown on the Texas side, S. H. Atkinson in Arkansas—both pr
essed for swift solution to the mystery. Presley and Runnels immediately launched a full-scale investigation, collaborating with Miller County Sheriff W. E. Davis (in office since 1939), Police Chief Richard Marlin Giles on the Arkansas side (appointed in 1938), the Arkansas State Police, and the Texas Department of Public Safety. The Lone Star State’s contribution was Texas Ranger Jimmy Geer, arriving on March 25 to augment the investigation.16
* * *
Deployment of a solitary ranger did not indicate a lack of interest in the Texarkana murders; quite the opposite, in fact. Rangers were relatively scarce in 1946, with only fifty-one available to police 7,197,000 citizens in 254 counties, plus 1,254 miles of the Mexican border. Outnumbered and outgunned, those rangers still felt equal to the task. Since 1896, when Captain Bill McDonald had single-handedly faced down a crowd assembled to watch an illegal Dallas prize-fight, the tiny force had lived up to a boastful unofficial motto: “One riot, one Ranger.”17
The first ten rangers were recruited by Stephen Austin in 1823, to carry out punitive raids against hostile Indian tribes. Twelve years elapsed between that action and formal creation of the Texas Rangers in November 1835, with fifty-six men divided among three companies. All but eight of those rangers were privates, earning $1.25 per day, from which they supplied their own horses, weapons, equipment and rations. Between 1838 and 1845, when the United States annexed Texas, rangers were engaged in near-constant battle with red enemies, including Cherokees, Comanches, and Apaches. During the Mexican War (1846–48), Texas Rangers waged ruthless guerrilla campaigns—including various atrocities against civilians—that earned them a reputation as los diablos Tejanos.18
The Treaty of Guadalupe Hidalgo, signed in February 1848, ended hostilities and left the Texas Rangers with little to do, as federal troops assumed responsibility for guarding America’s new southern border. Over the next decade, the force became—in the words of historian Walter Prescott Webb—“little more than an historical expression.” Commander John “Rip” Ford briefly salvaged the unit’s fighting reputation during 1858–59, battling Comanches and killing their war chief, Iron Jacket, then pursuing border bandit Juan Cortina unsuccessfully, but secession from the Union in March 1861 finished the team. During the Civil War, the Eighth Texas Cavalry would ride as Terry’s Texas Rangers, but founder Benjamin Terry had never worn the ranger star, nor did he focus on recruiting rangers for the unit. Through the bitter years of Reconstruction, Lone Star law enforcement fell to local sheriffs and the State Police, established in July 1870 and disbanded in April 1873.19
With dissolution of that force, thirty-seven of its officers swapped badges to join the resuscitated Texas Rangers, in 1874. Bandits, range wars, and Indian raids abounded over the next quarter-century, suppressed by six ranger companies totaling 450 men. By 1901, they had killed or caged enough desperados that state legislators felt safe in thinning ranger ranks by 90 percent, to eighty men in all. Mayhem increased once more, during the Mexican Revolution and World War I, including reports of ranger crimes against Hispanics on both sides of the border. In January 1919, Representative José Canales pressed for an overhaul of the force, to increase public confidence, and rangers who survived the cut generally distinguished themselves during the 1920s—pursuing smugglers and bootleggers, rounding up rustlers, policing strikes and demonstrations by the reborn Ku Klux Klan.20
The Great Depression put a crimp in ranger operations, with budget cutbacks that trimmed the ranks to a maximum of forty-one officers from 1930 to V-J Day in 1945. The rangers took a fling at politics in 1932, publicly supporting incumbent Governor Ross Sterling against rival Miriam “Ma” Ferguson in the Democratic primary, where victory amounted to election. Ferguson emerged victorious, and fired the rangers en masse upon assuming office, in January 1933. State lawmakers slashed salaries and budgets once again, cutting the force to thirty-two men, while Texas became a happy-hunting ground for outlaws such as Clyde Barrow and Bonnie Parker. When legendary ranger Frank Hamer tracked Bonnie and Clyde to their death in 1934, bringing his personal body-count to fifty-three dead felons, he was working on retainer for the state prison system, rather than the Texas Rangers.21
Ma Ferguson’s administration, like that of her husband James during 1915–17, wallowed in corruption and scandal. Successor James Allred campaigned on a platform of improved law enforcement and established the Texas Department of Public Safety in August 1935, with six divisions: the Texas Rangers, Texas Highway Patrol, Bureau of Communications, Bureau of Intelligence, Bureau of Education, and Bureau of Identification and Records. Five companies of rangers were preserved, with hiring and promotion overhauled to eliminate political patronage. New applicantions were accepted from “crack shots” between the ages of thirty and forty-five who were “perfectly sound” in body and mind, capable of writing “intelligent” weekly reports of their activities. It went without saying that all would be male. Jim Crow politics dictated that none would be African American, while a five-foot-eight height requirement barred most Hispanics.22
No specific educational requirements were established for Texas Rangers, though each new officer received instruction in the latest techniques of ballistics and fingerprinting, communications and record-keeping. A crime lab, located at headquarters in Austin, was equipped to handle evidence that baffled individual rangers, scattered statewide in small towns where they served as a de facto rural constabulary. By the early 1950s, rangers handled more than eight thousand cases per year.23 They had seen everything.
Almost.
* * *
By March 27, authorities in Texarkana had questioned fifty to sixty potential witnesses in the Griffin-Moore case, including patrons and employees of Club Dallas, located near the murder scene on Highway 67 West. No one there remembered the victims, and a $500 reward offered for useful information worked against investigators, generating more than a hundred false leads. Rain and the trampling feet of morbid locals, anxious to view the crime scene, had obliterated any useful tracks or other evidence. Justice of the Peace Oval Cooper convened a brief inquest on Monday, March 25, declaring that Griffin and Moore “came to death at the hands of a party or parties unknown for reasons unknown.”24
One yawning void in the investigation was the failure to perform full autopsies on either victim. Aside from one newspaper reference to a physician’s examination refuting claims that Moore was raped—a statement either misinformed or deliberately fabricated—no evidence exists that either body was seen by a qualified pathologist. There are no documents confirming or refuting claims that Moore was tortured and/or mutilated. More crucial yet, despite recognition that both victims died from gunshot wounds to their heads, no other information is available beyond the fact that the one slug recovered bore “six lands and groves with a left-hand twist.”25
Lands and grooves are the markings left on a bullet fired through a rifled barrel, scored in passage by helical grooves that impart a spin to the projectile around its long axis, improving its aerodynamic stability, range, and accuracy. Marks on the bullet are a mirror image of the barrel’s interior, wherein grooves are depressions and lands are the ridges of metal between grooves. On the bullet, therefore, a raised land conforms to the barrel’s grooves, and vice versa. The lands and grooves of a particular barrel have width and depth conforming to those of other weapons produced in that caliber by the same factory. Rifling is generally described by its twist rate, such as “one turn in six inches,” indicating the distance a bullet must travel to complete one full revolution. Various manufacturers use broaching bits that turn to either left or right, hence the reference to a “left-hand twist.” Standard rifling in a Colt Hammerless Pocket semiautomatic pistol was 1:12-inch with a left-hand twist, employing a barrel 3¾ inches long.26
We know one bullet was recovered from the Griffin-Moore crime scene. The FBI confirms it, without stating where the slug was found. The same memo contradicts statements from Texarkana investigators, reporting that “it is believed that they were sho
t while in their car.”27 If so, the evidence tally—one slug, two cartridge cases—leaves a bullet unaccounted for. How could it disappear inside of Griffin’s Oldsmobile?
That puzzle is resolved if either victim died outside the car, since the missing slug in question may have passed through either victim’s head. Colt manufactured .32 ACP rounds in two varieties, matched by identical 7.65mm Browning ammunition in Europe: full metal jacket projectiles weighing seventy-one grains (4.6 grams), or jacketed hollow-point bullets designed to “mushroom” on impact, weighing sixty-five grains (4.2 grams). FMJ rounds left the pistol’s muzzle at speed of 900 feet per second, while the lighter JHP bullets launched at 925 feet per second. Either fired at close range could pass through a human skull, although the JHP rounds are more likely to flatten and carom around inside, without producing an exit wound. Flattened, mutilated hollow-point rounds are also, obviously, more difficult to match against a given firearm.28
Regrettably, we do not know—and never shall—the kind of ammunition used to murder Griffin and Moore.
Thus far, authorities had drawn no link between the double murder of March 24 and the Hollis-Larey attack on February 22, but locals who recalled the first attack from coverage in the Gazette voiced criticism of the way in which that prior incident had been soft-pedaled. Would Griffin and Moore have ventured onto lover’s lane the night they died, if wider coverage had been given to the first attack? In fact, there was no reason to presume that they were ignorant of the event, which had been covered locally for two days after its occurrence. And, as subsequent events would prove, even a double murder trumpeted in headlines would not keep young lovers from their calling on the lonely roads outside of Texarkana.29
Even with the death of Moore and Griffin, panic had not gripped the city yet.
Sixty years after the fact, author William Rasmussen suggested that Texarkana’s stalker might have killed before targeting Griffin and Moore. On February 24, 1946, the Dallas Morning News reported that a woman named Beatrice Graham Thraser had been found shot to death on a sidewalk in the 3600 block of Wendelkin Street, shortly after midnight on Monday, February 18. Dressed in “a nightgown, a housecoat and a street coat,” she had been shot once behind her right ear. A .32-caliber revolver with one round fired lay at her feet.30 That case remains unsolved today, but we may rule out any link to Texarkana’s later crimes unless the killer had two .32’s—a revolver and a semiautomatic—choosing to leave the former behind for police in Dallas.
The Texarkana Moonlight Murders: The Unsolved Case of the 1946 Phantom Killer Page 3