The name attached to the design is misleading. A working prototype existed and was carried into a combat zone as early as late 1943, three to four years before the 1947 date the model number implies.² That figure traces to a single demonstrator Johnson built in late 1946 for a failed Argentine sales pitch, not to when the design was actually conceived.^2,3

Its designer, Melvin Johnson, was a Boston lawyer with no formal gunsmithing training, and the path his design took from a $300 machine-shop prototype to a rejected Army submission to five hand-built carbines runs through one of the most personal ordnance rivalries of the era, covered in full in History below.⁴

Because so few examples exist and none ever entered service, the practical value of the Auto-Carbine today is almost entirely historical and collector-driven rather than military. What follows separates the verified specification record from the folklore that has grown up around the design, including the persistent claim that it could fire automatically, which no wartime or postwar source actually supports.²

A Short-Recoil Action Borrowed Whole from the M1941 Rifle

The Auto-Carbine used the same locking system Johnson had already proven on the M1941 rifle and light machine gun. The barrel recoils a short distance under firing pressure before a rotating, multi-lug bolt unlocks and cycles the action, a mechanism needing only about 20 degrees of bolt rotation to lock and unlock, well under the roughly 45 degrees required by a Garand or the near 90 degrees of a Mauser 98 bolt.¹² That same low-rotation locking geometry is the reason the design attracted attention from an entirely different rifle program a decade later, discussed in History.

Built from Recycled Light Machine Gun Parts

To hold down costs on a project with no committed buyer, Johnson's shop reused 22-inch barrels salvaged from existing M1941 light machine guns rather than machining new ones, while fabricating a new receiver, safety mechanism, and an in-line stock borrowed from the LMG's own layout to cut muzzle rise.^8,9 A folding, Lyman-made rear sight adjustable from 100 to 500 yards sat behind a raised front sight and folded flat for storage, giving the carbine a profile noticeably shorter and slimmer than the standard-length M1941 rifle despite not being a true bullpup.⁹

Feed came from the same 10-round rotary magazine used on the M1941 rifle, refillable in the field from standard M1903 Springfield 5-round stripper clips rather than a dedicated Johnson-only clip, a deliberate nod to existing US Army ordnance logistics.^1,9 At least one source also documents a 25-round single-stack detachable magazine tried on some prototypes, meaning the 10-round rotary was the dominant but not the only feed system across the small run.¹

Despite the automatic-sounding name, every period and modern source describes the Auto-Carbine's fire control as semi-automatic only. A period newspaper account quoted by American Rifleman described contemporaries firing one aimed shot per second for the first 11 shots, a practical sustained rate of roughly 60 rounds per minute rather than a cyclic full-automatic figure.² An adjustable 300 to 900 rounds-per-minute cyclic rate sometimes attributed to this weapon actually belongs to the separate M1941 Johnson machine gun, a genuinely automatic light machine gun that used the same family of parts for a very different role.⁸

A Marine Reservist Who Thought He Could Beat the Garand

Melvin Maynard Johnson Jr. was born into an affluent Boston family in 1909, commissioned a second lieutenant in the Marine Corps Reserve in 1933, and graduated from Harvard Law School in 1934, briefly teaching there afterward.⁴ He was not a career gunsmith. He conceived his recoil-operated rifle action while serving as a Marine Corps observer at Springfield Armory in 1935, watching the Army develop and adopt the M1 Garand from the inside and deciding he could design something better.⁴ By February 1936 he had a working prototype built in a machine shop for $300, using an unprecedented ten-lug rotating bolt.⁴

What followed became, in the words of several firearms historians, one of the more personal ordnance disputes in American military history. Johnson lobbied hard for Army and Marine adoption of his rifle over the Garand, prompting an official Ordnance Department test and a public shoot-off before senior officials in June 1938.⁵ Army Ordnance, having already committed heavily to the Garand's gas system and production tooling, rejected the Johnson rifle outright and had no institutional appetite to reverse course.⁵

Denied an Army contract, Johnson found a different buyer. The Netherlands Purchasing Commission ordered more than 20,000 Johnson rifles in 1940 to arm Dutch colonial forces defending the East Indies against an anticipated Japanese invasion, a large order that never fully delivered before the Indies fell in 1942.⁶ The US Marine Corps, unlike the Army, kept pushing for the design, eventually procuring 750 Johnson rifles, with 242 officially issued to the Marine Corps 1st Parachute Regiment, whose paratroopers needed a rifle that broke down compact enough to jump with, something the Johnson's quick-detach barrel made possible in a way the fixed-barrel Garand could not match.^6,5

Real Combat for the Parent Rifle, Years Before the Carbine's Official Date

The M1941 Johnson rifle made its combat debut with the 1st Parachute Battalion's assault on Gavutu Island on August 7, 1942, the opening days of the Guadalcanal Campaign.⁵ Despite only 23 rifles present in that fight, the design proved accurate and reliable.⁵ Johnson's companion M1941 light machine gun, nicknamed Emma (Johnson gave every design a pet name, the rifle itself was Betsy), weighed just 12.3 lb without its 20-round magazine against 19-plus lb for the Browning automatic rifle, with an adjustable cyclic rate of 300 to 900 rpm.⁸ It reached Marine Paramarine and Raider battalions and limited numbers of the Army's 1st Special Service Force, and during the fighting at the Battle of Bloody Ridge on Guadalcanal, Marine paratroopers' Johnson LMGs were credited with defensive crossfire that helped hold Henderson Field at a critical moment.^8,9

The Auto-Carbine itself grew directly out of this wartime record rather than appearing from nowhere in 1947. Johnson designed it during the war as a hybrid, combining the rifle's action and rotary magazine with the light machine gun's straight-line stock, on the theory that a shorter, lighter rifle would suit the comparatively small-statured Dutch East Indies colonial troops he still hoped to arm.^1,3 The clearest proof of that wartime origin is Marine Captain Harry L. Torgerson, the Marine liaison officer to Johnson Automatics, who visited the company's Cranston, Rhode Island factory while the carbine was under development and talked Johnson into lending him a prototype, serial S-5.² Torgerson carried it as a personal weapon during the Bougainville campaign in the Solomon Islands in late 1943 and early 1944, serving with the 4th Marine Parachute Battalion, and was photographed with it outside his tent in the field.² The prototype carried the stenciled nickname Daisy Mae, after the buxom Li'l Abner comic-strip character, continuing Johnson's habit of naming his designs, and one surviving example reads Daisy Mae the 3rd on the buttstock, implying at least three numbered prototypes existed by that point.²

Why the Carbine Never Found a Buyer

In late 1946, Argentina expressed interest in the design, prompting Johnson to build one further prototype specifically for evaluation by Argentine officials.³ After review, Argentina declined to purchase, and the most-repeated account holds that this Model 1947 demonstrator was the last Auto-Carbine ever built.^3,2 Separately, the Marine Corps briefly considered ordering 100 carbines for troop trials, but the initiative never became a funded contract.³

The commercial failure had several compounding causes rather than one. The Dutch East Indies, the weapon's original target market, was overrun by Japan in 1942 and then dissolved as a colonial possession entirely, with sovereignty passing to the new Republic of Indonesia in 1949.³ The US Army and Marine Corps already had the Garand in service and showed no appetite to adopt another Johnson product after rejecting his rifle once. Argentina's evaluation, Johnson's best remaining shot at a contract, fell through. And Johnson Automatics itself went bankrupt by 1949, closing off any path from hand-built prototypes to real production even if a buyer had eventually appeared.^3,11

An Unlikely Second Life in the AR-10 and AR-15

Johnson's postwar career did not end with his company's bankruptcy. Winchester acquired Johnson Automatics' shares in 1949, reportedly after Johnson forced the sale, and Johnson went to work for Winchester in the early 1950s as a designer and advisor to chairman John Olin.⁴ He also transferred from the Marine Corps Reserve to the Army Ordnance Corps Reserve in 1949, eventually rising to full colonel and serving as a weapons consultant to the Secretary of Defense in 1951.⁴

In 1955 Johnson was brought on as a paid consultant to ArmaLite to help promote Eugene Stoner's new AR-10 rifle to the Department of Defense, then continued working with ArmaLite and later Colt's Manufacturing as an advocate for the AR-15 as that program developed.¹¹ The engineering connection Small Arms Review traces from Johnson to Stoner is specific rather than a vague inspiration claim. Johnson's cam-controlled, low-rotation bolt (the roughly 20-degree lock described above) was carried forward conceptually into the AR-10's, and later the AR-15 and M16's, rotating multi-lug bolt.¹² Two further shared principles follow it, an upper and lower receiver split rather than a traditional one-piece stock, and an in-line recoil geometry that keeps the reciprocating bolt mass directly on the bore axis to reduce muzzle rise, a layout the AR-15 pushed further by raising the sight line above the barrel.¹²

American Rifleman later profiled both men jointly as Marines Turned Arms Inventors, noting that Stoner had also served in the Marine Corps in World War II, as an aviation ordnanceman rather than a designer at the time, giving the two men a shared service background distinct from their later, more famous succession.⁶ Johnson died of a sudden heart attack on January 9, 1965, in New York City on a business trip, living just long enough to see the AR-15 and M16, the rifle his consulting work had helped push toward adoption, become the new US service rifle in the mid-1960s.⁴

The Auto-Carbine's only documented appearance under fire is unofficial and belongs to a single prototype rather than any fielded unit. Captain Harry Torgerson's personally borrowed serial S-5 accompanied him through the Bougainville Campaign in late 1943 and early 1944, part of the wider Pacific fighting of World War II, but it was never issued, tested in a formal trial, or used by any unit as a whole.² No other Auto-Carbine is documented as having reached a combat zone, and the design has no service record of its own beyond that one prototype in one officer's personal kit.

Surviving Auto-Carbine prototypes carry hand-stamped serial numbers with an S prefix, for Sample, apparently running from S-1 through at least S-5 or S-6 across the design's short life.¹ Torgerson's Bougainville-carried example was S-5. A separate prototype stenciled Daisy Mae the 3rd is now held by the Cody Firearms Museum at the Buffalo Bill Center of the West in Cody, Wyoming, and was the subject of a detailed photo showcase by TheFirearmBlog in 2017.¹⁰

Beyond the standard 10-round rotary-magazine configuration, at least one prototype used a 25-round single-stack detachable box magazine instead, a feed system tried but never standardized across the run.¹ The final prototype, built in late 1946 specifically for Argentina's evaluation, is the example that gave the design its commonly used M1947 name, even though it was mechanically the same weapon Torgerson had carried three years earlier rather than a distinct redesign.^3,2

Total production is usually given as five, though TheFirearmBlog notes that sources disagree between five and six, likely reflecting how the hand-built Sample-series prototypes were individually modified and renumbered over the design's short life rather than a genuinely unresolved count.¹⁰

The clearest comparison for the Auto-Carbine is its own parent, the M1941 Johnson rifle, since the two are separated by contracts rather than mechanics. The M1941 rifle at least reached real, if modest, military hands, with a 20,000-unit Dutch order and 750 Marine Corps rifles, while the Auto-Carbine derived from it never secured a single confirmed order and stopped at five hand-built units.^6,1 Both share the same short-recoil rotating-bolt action and .30-06 chambering, so the gap between them is a story about timing and a shrinking customer base, not about which design worked better.

A useful contemporary rival outside the Johnson family is the M1 carbine, which chased a similar idea of a shorter, lighter rifle-class weapon for troops who did not need a full-length infantry rifle. Where Johnson's carbine stalled at five units, the M1 carbine became one of the most mass-produced firearms in American history, with more than 6.1 million built across multiple manufacturers during World War II.⁷ The difference was not really about mechanics. The M1 carbine had an Army-defined requirement and a dedicated intermediate cartridge behind it from the start, while Johnson was chasing an export customer that disappeared and a domestic market that had already committed to other designs.

How many M1947 Johnson auto carbines were ever built? Five are confirmed, with one specialist source noting collectors disagree between five and six.^1,10

What cartridge does the M1947 Johnson auto carbine use? The standard American .30-06 Springfield, the same cartridge as its parent M1941 Johnson rifle.¹

Is the M1947 Johnson auto carbine fully automatic? No. It is semi-automatic only. No source documents a select-fire capability on the carbine itself, and figures suggesting an automatic cyclic rate actually describe the separate M1941 Johnson machine gun.^2,8

Why is it called the M1947 if the design is older? The name comes from a single demonstrator prototype Johnson built in late 1946 for a failed Argentine sales evaluation. A working prototype had already existed and been carried in combat conditions on Bougainville three to four years earlier.^2,3

How is the M1947 Johnson auto carbine connected to the AR-15? Its designer, Melvin Johnson, worked as a paid consultant for ArmaLite starting in 1955, promoting Eugene Stoner's AR-10 to the Department of Defense. Johnson's rotating multi-lug bolt, split upper and lower receiver, and in-line recoil geometry are documented design influences carried into the AR-10 and, from there, the AR-15 and M16.^11,12

Where can you see a Johnson Auto-Carbine today? At least one surviving example is on public display at the Cody Firearms Museum in Cody, Wyoming. Another, Torgerson's wartime-carried S-5, remained in his family's hands after his death in 1958 and is understood to be in a private collection today.^10,2