Click on images to enlarge or to see full sized drawing.

The air cooled ground type Browning machine guns in caliber .30 are descendants of the Model of 1917 water cooled “Heavy” by way of the M1919 air cooled tank guns. Since they are related it is not unexpected that many of the parts and assemblies are similar if not identical.  In this article we are going to explore the various top cover designs and the associated attached parts that make up the greater portion of the feed system of these iconic air cooled weapons.

While the 18 inch barreled air cooled M1919 tank guns appeared in 1918 at the close of WWI the 24 inch barreled ground mount version really dates from 1936.  The M1919A4 and its variants owe their existence to Cavalry and Infantry desires for an automatic weapon light enough to be man portable and quickly deployed.  Experiments to produce this sort of weapon began in 1931 and continued until 1936 when the Army adopted the design so familiar to the admirers of John Moses Browning’s genius.

The top cover allows access to the feedway, extractor/ejector and even headspacing adjustments, and carries the majority of the parts that advance the cartridges and hold them it in position for extraction from the feed belt. Over the service life of the M1919 family many changes have taken place in parts design, materials, manufacturing methods, and most importantly those changes improved the utility of the weapon for the users.

The M1917 top covers had the rear sight mounted on top of the cover while the production M1919 ground weapons carried the rear sighting mechanism on a bracket mounted on the left side plate.  The M1917 had no provision to hold the cover in an open position once it was unlatched and raised.  This deficiency was remedied in production models of the M1917A1 and the M1919A4, A5 and A6. Some writers believe the changeover in 1938 to a cover that was held open at 45 and 90 degrees had something to do with regard for the operator’s fingers getting pinched by a falling cover. My personal belief is that the change to the cover hold open feature was driven by a need to keep the cover open without someone holding it with one hand during reloading or clearing jams.

There were two prescribed methods of reloading, that is changing belts, in the M1919 the first required the operator to stick the brass starter tab on the end of the cloth belt, or the steel starter tab on the linked belt through the feed way on the left side of the weapon pulling and releasing the bolt handle which placed a cartridge under the extractor, the “half load”, then pulling and releasing the bolt handle again to place a round in the chamber, the “full load”. The other method, which was the one likely most used, involved opening the top cover, raising the extractor placing the first cartridge in the belt against the cartridge stops lowering the extractor over the first cartridge, flipping the belt feed lever to the left so that the lug on the bottom was positioned over the belt feed lever track on the bolt top, closing and latching the top cover and pulling and releasing the bolt handle.  While the second loading method seems more time consuming, my bet is that most operators chose the route of opening the top cover. I have tried both methods, albeit not under fire, and personally prefer opening the top cover.  If someone was shooting at me, well, maybe I might adopt another policy that didn’t require me to expose any more of my body to fire than necessary.

Some cloth belts had starter tabs on both ends, some on only end and the 100 round cloth belts used in vehicles usually had no starter tabs at all, only brass grommets on either end.

One of the advantages of using metallic links is that the belt can be made infinitely long by merely removing the starter tab and a cartridge from the new belt and using the de-linked cartridge to attach the new belt to the old.

M1 steel links were originally developed in 1931. Their use in WWII was confined mostly to aircraft ammunition until very late in the war when they started to appear in ammunition intended for armored units.  The use of cloth belts saved precious steel and manufacturing capacity and continued well into the Korean War.   The Israeli’s used cloth 230 round 7.62X51 belts into the mid 1970’s.  

Top covers for the M1919 air cooled family fall into two basic categories, forged and finish machined and cast and finish machined. All of the production top covers manufactured before 1943 were forged and machined.  

While they were “covers“, they were useless by themselves as they lacked the extractor cam and attaching rivets, top cover extractor spring stud, and the latching “plate” and it‘s attaching rivets.  That’s eight individual parts. Saginaw Steering Gear, Saginaw Malleable Iron and the industrial muscle and know-how of General Motors would change all that, forever.

Ordnance called user replaceable grouping of parts related or attached to each other by either necessity or manufacturing process an “ASSEMBLY”.  That’s how the SNL’s (Standard Nomenclature List) showed the necessary stock numbers, that’s how you ordered them and that’s how you got them. SNL’s listed every part and the quantity of that part need for one complete weapon along with tools and accessories. However, many of the component parts had no stock number listed in the SNL only the drawing/ part number, in other words, while you could see the component parts and their drawing numbers, you couldn’t order the replacement components, only the assemblies or authorized parts. Sometimes the part capable of being ordered and user replaced, like the original all steel barrel, consisted of a single piece, other times the individual parts were combined into a unit of issue like the extractor assembly. The reason for this is that field users of the weapons had limited time and ability to effect repairs. They were trained in tactics, marksmanship, identifying basic operating problems and effecting reasonable remedial action to clear stoppages and return the weapon to operating condition. The users had limited access to tools, and no access to shop facilities like vices, welders, grinders etc.  They were not trained as Armorers; they were trained as operators of the weapon. The average BMG hobbyist likely knows more of the history, and internal workings of the weapons than the combat users.  The combat users had much more important things on their minds, like staying alive.

Forging and machining to final dimensions was the preeminent method of manufacturing firearms parts when the M1917 was designed.  The rough forgings were oversize and the machining operations got the part to usable dimensions.   While this produced high quality parts, it required skilled workman and properly designed and maintained tooling to keep parts within working tolerances. This method of fabricating parts lead to a raft of parts interchangeability concerns which bedeviled Colt, Remington and New England Westinghouse while they were manufacturing M1917’s during WWI.

Click on image to enlarge.

The upper illustration in Fig. 1 is from SNL A-6 dated May 28, 1941 and the lower is from the same document dated 6 September 1943 showing drawing numbers for the cover assembly as well as all of the attached feed system parts.

After each of the parts, the cover itself, the extractor cam, the latch plate and their attaching rivets and the stud required to hold the forked end of the extractor top cover spring were assembled they received an assembly drawing number of their own C9801. None of the separate parts of the feed system, the belt feed slide, belt feed lever, belt feed pawl, pins, springs etc. were included, and they were ordered as separate parts. This assembly drawing number, C9801 was associated with, during WWII and until the adoption of Federal Stock Numbers (FSN’s) in the ‘50’s, an Item Stock Number of A006 01 00340.  The A006 prefix indicated that these parts were specific to the group A-6 which was the M1919 family.  Some parts listed in the A-6 SNL carried a prefix of A005 this indicated that these were also used on the M1917/M1917A1 which was a group A-5 weapon. This is how the cover assembly was packaged, inventoried and issued.  If you needed a replacement top cover you ordered A006 01 00340 and what came in the package would work, in spite of changes in materials, design, or even the general appearance of the part. This is not some strange anomaly specific to the M1919’s but a general operational policy of the Ordnance Department’s Field Services Division which was charged with the responsibility of obtaining and furnishing both complete weapons and repair parts. 

The M1919 family of weapons along with nearly everything else Ordnance produced in WWII were in a constant of evolution.  New parts appeared, old parts disappeared, or were listed in the SNL’s as “used on guns of new manufacture” for new parts or “used on guns of early manufacture” or just plain “usable” for old style parts.

The upper Figure 1 illustration shows the original method of attaching the cover to the weapon.  This method of attachment, a holdover from the M1917, was a simple hinge pin, B17506, secured with a cotter pin, BFAX1CC, and had no hold open feature for the cover. Keep in mind that the cover hold open feature was officially adopted for the M1919 in 1938 and this SNL was published in 1941.  This is pretty typical for the SNL’s I have examined.  The SNL’s sometimes lagged far behind new parts development.

Fig. 2
Click on image to enlarge.
This cut from ORD 9 SNL A-6, the successor to the documents shown in Fig. 1, dated April, 1947 shows all three M1919 top cover assemblies.  All of them have the same assembly drawing number and stock number. The first listing is the “standard” cover; the second is “alternate A”. Both are forged and machined assemblies consisting of eight parts. The only difference between the two is the two rivets that hold the extractor cam to the top cover.  The only difference between the rivets is that the in the “standard” cover the rivets are ground flat on the top of the cover while on the “alternate A” assembly the rivets have domed heads, and are alternate parts themselves (note the A suffix on the drawing numbers for the rivets).

Fig 3
Click on image to enlarge.

This assembly drawing C9801 Revision 14 dated 6-12-45 shows the “standard” and the “alternate A”, called out on this drawing as “Alternative A” covers.  Note the domed head alternate rivets holding the extractor cam and cover spring stud to the top cover. While the author has never observed an “alternate A” cover with domed head rivets no doubt there some out there somewhere. And if any reader has one, a picture would be appreciated.

The alternate A cover first appeared as Revision 5 to C9801 dated 9-9-42 which is the same date as the alternate dome rivets original drawing date. The Saginaw ArmaSteel cast cover is illustrated as “Alternative B”.  Many of the Ordnance drawings and SNL’s reviewed have this alternative/alternate semantics duel.  There is also a note on this drawing informing that the parts are only available as an authorized assembly. “Alternate B” is, as you can see, a whole different kettle of fish. Note that the drawing/part number is a 7xxxxxx series number in Fig 2 indicating that this part is a new part for an existing weapon or a new part for a new weapon while the assembly drawing C9801, Fig 3 is common to all three variants because they are totally interchangeable assemblies.

When, in late 1943, Ordnance started on a program to make the drawing number, part number (piecemark), and stock numbers common using 7 digit numbers with no Alpha characters they developed a conversion system for the letter prefix series of drawings and reserved the number range from 7000001 to 9999999 for new parts, even if some of these “new” parts were interchangeable with older designs.

Alternate B is the Saginaw designed and manufactured ArmaSteel cast cover.  It consists of only two parts, the cover and the STUD, which is the cover part that holds the forked end of the extractor top cover spring.  It is plainly evident from a manufacturing standpoint that, all things being equal, an assembly consisting of two parts is superior to one containing eight. Not only that, but when you figure in the time necessary to fabricate to tolerance EACH separate part, and then assemble the parts to tolerance, the cast cover becomes even more attractive. Using innovative manufacturing techniques, like the ArmaSteel top cover, goes a long way to explaining how Saginaw was able to reduce the contract price of a M1919 from $657 in 1940 to less than $60 in 1945. 

Saginaw Steering Gear had been using perlitic malleable iron castings, which is what ArmaSteel is, to produce difficult to machine auto parts prior to WWII, and as early as April, 1942 started suggesting to the Ordnance Department that these same casting technologies could be adapted to machine gun parts production, saving steel, machine time, and machine tools, increasing production, and saving money, all at the same time. Saginaw had never produced a machine gun, or any other weapon for that matter, and approached weapons manufacturing the same way they addressed their primary business which was making interchangeable metal parts for the automotive industry. They had no preconceived notions about anything other than that they knew their business and they knew that they were good at it.  A little self-confidence never hurt anything.

A latter day example would be Gaston Glock, originally an Austrian manufacturer of household hardware, who decided to enter a competition to produce a new sidearm for the Austrian military. Herr Glock had no experience producing firearms whatever.  He also had no investment in tooling, reputation, or patents, and most importantly, he carried no baggage from previously manufacturing firearms. He made curtain rods, and he started with a clean piece of paper. That being said, he produced what is now, arguably, the standard police sidearm in the US. 

Fig. 4
Click on image to enlarge.

This remarkable drawing dated 9-22-42, while a little hard to read, originated at Saginaw and was sent to RIA.  RIA, having no really good way to index and catalog drawings produced by suppliers, had someone hand write “D7116476”, the Ordnance drawing number for the cast cover, on the drawing.   It survived 67 years somehow getting microfilmed along with the other “backfile” drawings.  This Saginaw drawing, or one like it, was likely used to prepare the Ordnance’s official drawing which was dated August 10, 1944.

My personal opinion is that Saginaw was producing cast covers well before the 1944 date, using their own drawings which have the note to obtain any dimensions not on the Saginaw drawing from Ordnance drawing D28106, however, there is no documentation that we have uncovered as of yet that offers proof of this. 

Drawing C9801 Revision 10, the cover assembly, has a note on it mentioning drawing 7116476 as an alternate part. This drawing Revision shares the common date of August 10, 1944 with the original 7116476 drawing. The first C9801 assembly drawing that actually shows an illustration of the cast top cover is Revision 12 dated 3-8-45. Saginaw titled this drawing, as shown in Figure 4, “Machining Drawing of ArmaSteel Cover” and it’s the first documentary evidence of a cast cover that our research has uncovered.  The upper left area of this drawing shows the oval cut in the cover necessary to allow a machine tool to cut the recess in the extractor cam to allow the top cover extractor spring to latch in place. On a forged cover with a separate extractor cam this recess was cut into the cam before it was riveted to the cover.

Fig. 5
Click on image to enlarge.

The quickest way to identify a cast cover, it’s the one on the left, is to look for the oval cut in the cover seen near the latching area, in addition, the forged cover on the right has the rectangular latch plate while the cast cover has the tapered lip which is an integral part of the casting.  If you look closely at the forged cover on the right you can see the lighter colored flat rivet heads of the “standard” cover that attach the extractor cam and latching plate to the cover. This difference in appearance is caused by the difference in steel types and often shows up after Parkerizing.

The US furnished many M1919’s to the Israeli Defense Force and the Israeli’s converted them from US Caliber .30 to 7.62X51 NATO.   These IDF converted weapons are the source of most of the “parts kits” being marketed today. Because of the differences in the cartridge length and profile this conversion involved modifying the covers and some of the feed system parts either by altering the parts themselves or by designing and fabricating new components.

Fig 6 
Click on image to enlarge.

This photo, with the USGI forged cover on the left and an Israeli modified cast cover on the right shows how similar the covers are and a casual examination would miss the differences.

Part of the conversion process to the ½ inch shorter 7.62X51 NATO round involved closing the feedway opening by installing spacers and replacing the original rear cartridge stop with one of a different design.  This required changes in the belt feed pawl, cover and the belt feed slide which can be seen in Figure 6.

Fig 7
Click on image to enlarge.

In this photo the Israeli modified cast cover is on the left and the “dog leg” arm, which prevents double feeding if the extractor fails to withdraw a cartridge from feed belt, on the belt feed pawl is plainly visible as are the differences in the belt feed slide and the clearance cut in the cover itself required by this redesigned/modified part.

Most if not all IDF modified covers have a squared off “U” symbol stamped on the top and bottom of the cover in the hinge area.  This is a Hebrew symbol for “Peace through strength” and serves the same purpose as the US “Flaming Bomb” ordnance symbol. Some IDF modified covers have a label affixed to the cover.

Fig 8
Click on image to enlarge.
The modified belt feed slides are similarly IDF marked on the top, visible when the slide is extended to the left.  The mark on the cover itself has been almost obliterated. This IDF modified cast cover is used on a Caliber .30 Ohio Ordnance Works semi-auto M1919A4/A6 with the USGI rear cartridge stop installed and functions perfectly.  Note the oval cut and the extractor top cover spring visible thru the opening.  Also note the “dimple” to the left (rear) of the belt feed lever pivot bushing nut and the old stake marks on the cover to hold the nut in place and the stake marks on the nut to hold the pivot pin retaining cap screw.  Later in this discussion we will revisit the belt feed pivot parts which could make up an article in themselves. Additionally, the IDF modified some of their belt feeds levers and pivot pin assemblies to one of their own design.

Fig 9
Photo courtesy of Rollin Lofdahl
Click on image to enlarge.
The IDF took a standard USGI belt feed lever and modified the pivot pin hole to accept a tapered pin inserted from the bottom which extended up through the bushing and bushing nut and was held in place by a cotter pin resting on top of the modified bushing nut. Obviously the original pivot pin, cap (set) screw and toothed washer were not required with this modification. This arrangement kept the belt feed lever from ‘drooping” due to a worn pivot pin or a worn hole in the belt feed lever itself. The IDF did not remove the US piecemarks imprinted on the belt feed lever. Exercise caution when purchasing belt feed levers.

Fig 10
Photo courtesy of Rollin Lofdahl
Another style of IDF modified top cover note the milled out area to accommodate the folded front sight when the cover is raised to the 90 degree position and the cotter pin pivot  retaining system. This is the forged standard style previously shown in Figure 3. This retaining system using a cotter pin is reminiscent of the pre war US cotter pin style and would be the one used with the modified feed lever and the tapered pivot pin shown in Fig 7A.

Fig 11

This is drawing 6528106 Revision 22 dated 10-7-48 is the forged top cover.  This drawing has a note that advised “Was D28106”.  The original D28106 drawing produced by Springfield Armory is dated June 1, 1931.  It is a redrawing of the letter prefix and 7 digit renumbering of drawing 51-18-12 for the top cover of a M1919 tank machine gun the weapon that the M1919A4 developed from.  Most M1919A4 production drawings carry a June 30, 1936 original date. This drawing carries a June 31, 1931 date because the tank gun drawing was converted to a letter prefix drawing prior to the development of the M1919A4. Note the rectangular cuts on the pivot pin bushing hole and the dimple to the rear of the hole.  These features date from the original drawings for the M1917 and would be part of the drawings for M1919’s until eliminated on Revision 23 dated 12-6-49, 12 years after they were made obsolete due to changes in the pivot pin assembly design. This 12 year time lag maybe the result of the availability of the original spring type pivot pin assembly at least until January, 1944.

 At the beginning of this article Fig 1 showed the parts that attached the top cover to the casing. In the spring of 1937 the Rock Island Arsenal now in charge of all Caliber .30 ground Browning machine gun production published a document titled “Notes on the Browning Machine Gun Caliber .30 M1919A4” this 22 page illustrated booklet covered the changes necessary to convert the excess M1917 water cooled weapons left over from WWI to the air cooled M1919A4. Before this original “Notes” document was published, it appears that the original spring type belt feed pivot pin used on the M1917 and early M1919A2 and A2E3 was changed to one similar to the Major Item 51-24 Browning Aircraft MG. M18 and Major Item 51-25 /M19 aircraft BMG pivot pin assembly.  This assembly had the headless pivot pin inserted from the top of the cover through bushing nut, bushing and belt feed lever and was held in place by a cotter pin.  We say this because the original A157434 pin lacked a head and only had 1 transverse hole.  There was, apparently, a different style bushing nut that we haven’t as yet seen. Without a cotter pin passing through some sort of retainer, the pin would fall out it the weapon was inverted.

The original “Notes” also changed the belt feed pivot assembly to one where the pivot pin  now with a head, and with the same drawing number A157434 but Revision 2, was inserted from the top of the cover through new style bushing nut, bushing and through the belt feed lever.  This style pivot pin was held in place by a spring steel cap that snapped over the modified bushing nut. These early conversion instructions retained the cover hinge pin with no hold open feature but added a bracket to the left side plate which held the rear sight which was removed from the cover latch.

By spring 1938 RIA published an addendum to the original “Notes” called “Notes… Modified” which included the hold open feature consisting of a spring loaded fixed and a movable plate with bolt and castellated (castle) nut secured with a cotter pin and a slightly altered belt feed lever as shown on the lower half of Fig 1. This same publication also outlined some changes designed to raise the firing rate of the M1919A4 to 600 rounds per minute by using a muzzle (booster) plug with a .617 opening and a redesigned driving spring made with .047 diameter wire. These changes designed to increase the firing rate were not implemented on production weapons.

Fig 12
Photo courtesy of Rollin Lofdahl

This photo shows top and bottom of reproduction spring style belt feed lever pivot pin. I refer to this style pin as “type 1”. This pin was inserted through the top of the cover and the belt feed lever and the rectangular shoulders on the pin passed through the corresponding cuts in the cover.  The pin was rotated clockwise and the spring (arm) was locked in place when the protrusion on the underside of the spring engaged the dimple in the top cover. The 1941 SNL shows this style of pivot pin with an assembly drawing of A135054 as “used on guns of early manufacture M1919A2, M1919A4”. A similar pin design was used by Browning to lock the trigger group and the gas tube assembly to the M1918 BAR.

The M1919A5 design specific drawings were dated May 23, 1942, however, the first illustration that we could find of the A5 appearing in an SNL were from the ORD 9 SNL A-6 dated April of 1947 pictured above in Fig 12.   The lower portion of Fig 12 shows the completely different hold open device officially referred to as a “Detent, cover group assembly (M1919A5)”.  This assembly drawing was B195941 however no stock number was listed in the 1947 SNL. There may be earlier versions of an SNL showing the M1919A5, but I haven’t discovered any as yet.

It appears that during development of the A4 that this pin arrangement was replaced by the BMG aircraft type cotter pin retained pivot assembly.  The cotter pin style did not remain a standard for long.  I call the cotter pin retained style “type 2”. The only mention of this style is in a narrative in SNL A-6 dated May 28, 1941, page 52 which also describes the original M1919A4 as “mainly used as armament for combat vehicles”.

Fig 13
This illustration from the 1937 “Notes” document shows the next advancement in pivot pin assembly design which I refer to as type 3.  The 1941 SNL shows this design with an assembly drawing number of B147304 “ M1919A4 used on guns of new manufacture”.  Most of the parts are recognizable with the exception of 7 which is the spring steel cap that holds the pivot pin in place.  I stared at this for some time trying to figure out what it was and how it fit on the assembly, at the time I was not in possession of the drawing of the cap. 

Details from an early 1919A4 manual reissued in England in 1942.
Source: Jon Moran

Finally I called Rollin Lofdahl who had actually seen one of these and he explained that it was like a bottle cap and snapped over the bushing nut (5 in this drawing) holding the pivot pin in place and in turn the cap was held by the recess shown by 6. The four holes near the bottom of the bushing nut allowed the nut to be tightened using a pin inserted into one of the holes.  

The next figure is the drawing used to produce the cap which makes things much clearer.

Fig 14
This is the cap that holds the pivot pin in place on what I call the type 3 pivot pin assembly.  

Fig 15.
Photo courtesy of Theo Servetas, Aberdeen Proving Ground, June 2009

This photo shows a snap cap type 3 pivot pin retaining system assembled on a forged top cover on a M1919A4 displayed at Aberdeen Proving Ground.  It is likely that the only extant pivot pin assemblies like this are either on museum display or have been recreated by hobbyists building “period” weapons.  The plaque accompanying this weapon described it as “c. 1942” M1919A4 which it is.

Fig 16

This is another Saginaw produced drawing dated 6-5-40 which Jodie Wesemann discovered while copying some B17503 belt feed lever drawings. Why this drawing was hand numbered with the belt feed lever drawing number is unknown. This drawing goes a long way to explaining why the “type 3” snap cap pivot pin retention system went away.  The 6-5-40 date was before Saginaw got its first production contract to produce M1919’s and about the time when the “Educational Contract” was winding down. From the notes on this sketch it appears that problems developed when the top cover was closed with the operating lug on the belt feed lever not over the track in the top of the bolt.

 Apparently, Saginaw noticed, likely during function firing that after the third of fourth time the cover was closed in this manner the pivot pin was pushed upward with enough force to pop off the spring steel retaining cap. Eventually, during firing, the pin could walk its way upward enough the belt feed lever would fall off the pivot pin.  Obviously after the lever fell off the pivot pin the belt would not advance, and depending on how it was positioned when it fell off, the lever lying on top of the bolt could cause serious damage when the weapon fired the last cartridge chambered.  Back to the drawing board.  Literally. 

We, as of yet, have not discovered what transpired when Saginaw’s sketch arrived at RIA but by July 18, 1941 the final pivot pin bushing nut, drawing A196284 that retained the pin with the cap screw and the new pivot pin assembly B110529 containing that nut both appeared.  

SNL A-6 Changes No. 2 dated January 30, 1942 notified the field forces that a new pivot pin assembly, drawing number to B110529, which is the one we are most familiar with was available.  This assembly lasted until the end of the service life of the M1919 ground guns. Changes No. 2 also changed the drawing number of the bushing nut from A157433, the one shown in Fig 10 to A196284, the style that accepted the cap screw to retain the pivot pin, illustrated on the bottom of Fig 1. 

The September 1943 SNL no longer shows this “snap cap” type 3 assembly but continues to list the original spring style pivot pin as an available part along with the assembly B110529 which is shown on the lower part of Fig 1.  The spring pivot pin assembly, type 1, A135054, and the cap screw B110529 style, which I call type 4, are shown with A005 stock number prefixes which indicate that they were also intended for use on the M1917/M1917A1. None of the bushing/nut style pivot pin assemblies require the rectangular cuts in the pivot pin hole shown in the cover in Fig 8.

This pretty much covers the pivot pin assemblies, the only changes made after adopting the B110259 last style assembly was the addition, sometime after 1949, of an additional internal toothed washer between the top of the cover and the bottom of the pivot pin bushing nut.  This can be seen on the forged cover on the right in Figure 5.  My thought is that the toothed washers were substituted for staking the components in place.

TB ORD 366 issued in August 1949 contains rebuild standards for the M1919’s and other weapons; specific mention is made that the B110259 pivot pin assemblies are the only approved parts.  There is no mention of the toothed washer between the bushing nut and the top cover.  This would lead me to believe that there might have been some weapons equipped with older spring style on the loose.  In any event, any weapon undergoing arsenal rebuilds should have had the older style assemblies replaced. This directive also covered M1917A1’s.

Originally, the M1919A4 Fixed BMG was intended use in the M2 and later the M3 Stuart tank; it was nothing more than an A4 Flexible equipped with a horizontal buffer back plate having no pistol grip. In May of 1942 the M1919A5 Fixed BMG, Major Item number 51-114, was introduced into the Ordnance system. As tank development proceeded it became apparent that there was a need for a different fixed type M1919 for coaxial mounting in a tank turret. In May of 1942 the M3A1 version Stuart light tank appeared and the M1919A4 Fixed, previously used coaxially with the M23 37mm main gun was replaced by the M1919A5.  In a coaxial mounting in a tank the machine gun is trained and elevated with the same mechanism as the main gun.

Because of space limitations these mountings required a method of bolt retraction not requiring the operator to reach forward and operate the conventional bolt handle it also needed neither front nor rear sights, or a pistol grip. The same tight quarters also required a different style of hold open feature for the top cover and a different type of latch tensioning mechanism which is shown in Fig 17. Eliminating the fixed/movable hold open plates and the tensioning spring, hinge bolt and replacing the bolt handle with a cocking stud gave additional right side clearance and the bolt retracting handle allowed the bolt to be operated from the rear.

Fig 17

During the 1950-1954 time frames some M1919A4 Fixed and Flexible guns were altered by the Rock Island Arsenal to an A5 style configuration and became known as M1919A4E1. These weapons were modified to satisfy a requirement for fixed type weapons for coaxial mounting in tanks while the M37 was being developed.

The M37, the last of the breed, eventually replaced the A4 Fixed, the A5 and the A4E1’s it could be fed from either right or left and used a completely different top cover design. Many of the M37 parts were not interchangeable with the M1919’s.

One of the problems most people encounter when doing research on Ordnance publications, drawings, and other documents is a lack of understanding how the system worked. Drawing numbers in the letter prefix system, adopted in 1922, which mandated, among other things, that only one part could be depicted on each drawing were assigned in a sequential order for each letter size.  There was no readily observable connection between the drawing number and the weapon or weapons the part was used on. However if you looked at the drawing itself there was a “Drawing Pertains To” block which listed each weapon the part was used on and , if it was part of an “assembly”, what the assembly drawing number was.

Conversely, you could go to the Class and Division drawings and find a list of the drawings for each part needed for that item along with illustrations showing where parts went on the item. The letter indicated the size of the paper or other medium the part was drawn or traced on. As you progressed up the alphabet from A through E the drawing size increased. This system allowed a part common with several weapons to have a single drawing number. Each drawing displayed the piecemark (part number) for each part.  Just because each part had a piecemark does not mean that part was so marked.  The drawing of the part would indicate whether or not the part should be marked and where.  Sometimes there was a “drawing” which was not a drawing at all but a table listing which parts of which weapon were required to be marked and how.  For the M1919A4 this “drawing” was B169913.

The term “revision” to a drawing is equally misunderstood.  Revisions are changes to the drawing, not necessarily to the part.  As a matter of fact, the majority of revisions have nothing to do with the design of the part. Equally important to remember is that not every revision to the drawing resulted in a change to the required piecemark.  The drawing shown in Figure 18 had 11 revisions but only revision 2 changed the required piecemark; however, none of the revisions affected the interchangeability of the part. Most revisions are administrative in nature, sometimes adding either weapons or assemblies to the list of what the part pertains to, adding information about the previous history of the drawing, or adding information about what the drawing became.

Fig 18

This is drawing A157434 Revision 11 of the pivot pin.  Each revision has a date and the circles with numbers and the leaders (arrows) pointing to what was changed by each revision.  Revisions 3, 6 and 10 added weapons to the “Dwg Pertains To” block and Revision 5 added the belt feed lever pivot assembly.

Revision 1 actually changed the complete design of the pin and Revision 2 altered some dimensions and added “-2” to the piecemark shown to the right of PIVOT, BELT FEED LEVER with the letter prefix of the drawing number almost enclosed by a large “C”.  The piecemark was not imprinted on the part. At the very bottom of the drawing there is the note that converted this letter prefix drawing to the 7 digit format on 5-10-48 WO/C (without change). The original date of this drawing is September 30, 1936 which is consistent with the majority of drawings specific to the production M1919A4’s. The scale of this drawing is 2/1 which means the part on the drawing was drawn twice the size of the actual part. A size drawings are 8 ½ X14 inches, today’s “legal size” and are the smallest.

Fig 19

This is the original drawing of the pivot pin, type 2, designed to be retained by the cotter pin.  Both drawings have the same drawing number but this one dates from Jan 22, 1936.  The reason that the drawing numbers have not changed is that this pin COULD be used in lieu of the pin in Figure 13 IF you had a cotter pin and the necessary parts to retain the pivot pin. Ordnance was supposed to change the drawing numbers only when the parts lost full interchangeability even though the parts looked nothing alike. It is not likely that this style of pivot pin was used on many if any production weapons. This pin was part of the evolutionary process that developed the standard for the production weapons.

Remember the example of the previously mentioned “Changes No 2” to the 1941 SNL which changed the assembly number for the belt feed pivot assembly AND the drawing number for the nut.  Why?  Because these items were not interchangeable with the previous part/assembly even though the other parts of the assembly remained unchanged.  

Generally parts of weapons where the drawing numbers remained consistent never altered the design enough to compromise their interchangeability even though the piecemark was changed by a revision suffix. SNL’s usually did not include revision numbers when showing parts or assemblies. There were several reasons for this.

Adding revision numbers took more space and would generate confusion about the drawing/part number and issues would arise relating to whether or not a part with a lower revision number was usable. While the revision numbers were vitally important to the manufacturers of the part, as long as the drawing number of the part or assembly provided agreed with SNL listing, the revision number was immaterial to the field services because the part was fully interchangeable. The only persons who presently have an interest in revision numbers are researchers and hobbyists who are attempting to date parts, understand the evolution of the weapon design or to return a weapon to “correct” condition. Herein lays a very large problem. 

Even if you had a brand new weapon in the original, unopened packaging there is no guarantee that all of the parts would be “correct”. When weapons were assembled at the factory any usable parts or assemblies on hand were pressed into service. While parts with revision numbers could not have been used prior to their revision dates, because they didn’t exist, older parts on hand could have been used. 

The belt feed slide, belt feed pawl and spring, and the spring pin used to attach these parts to the cover along with the extractor top cover spring and the M1919A4 cover hold open parts drawing numbers remained constant up to the 7 digit conversion in 1948 and then on to the end of their service life. This does not mean that changes did not occur to the parts, it means that the design of these parts never changed enough that they lost their interchangeability.  

Some revisions merely altered tolerances by .001 or .002 + or -.  Revisions like these would be nearly impossible to detect even if you measured the parts. If a part had lost its interchangeability it would have been given new drawing number, piecemark and stock number. Depending on the particular weapon, when undergoing arsenal rebuild, revision numbers of parts took on great significance.  Arsenals or other rebuild facilities had the resources to identify earlier design parts that required replacement even if they did not have piecemarks to identify them.

A good example of this would be the arsenal rebuilds of M1919’s post WWII. At the end of WWII production of the M1919 the original cover extractor spring stud, drawing number A24604, that held the forked end of the top cover extractor spring in place was redesigned and superseded. This redesign of the stud, shown on drawing A7123315 with an original date of May 25, 1945, increased the clearance between the inside of the top cover and the head of the stud to .160 inches from the original .155 dimension.  The reason for this change is presently unknown. 

This change may have had something to do with the design of the top cover extractor spring, originally .05 inches thick made of ‘forged steel” vs. the last drawing 6517513 showing 18 US gage spring steel (.0478 inches thick) however, this spring’s design never changed enough to effect its interchangeability and warrant a change to its drawing number other than the conversion to the 7 digit system which never altered the design of the part at all. The late design spring was thinner than the earlier design but made of different material. Why a thinner spring would require greater clearance is unknown, perhaps it just made the spring easier to install or possibly altered the force applied to the extractor.

The people doing these arsenal rebuilds were assembly line workers, taught to do one or two functions they were not “Armorers” who knew the entire weapon. The studs are not interchangeable because the new part has a new drawing number. The new part also has a new stock number listed in the 1947 SNL.

TB ORD 366 the rebuild standard for small arms issued in 1949 calls for the replacement of the older style stud.  Obviously, it would be impossible to detect the difference of .005 just by looking and the stud was not piecemarked. The arsenal rebuild facilities most likely made a “no-go” gauge to check the head clearance and determine which, likely all of the studs because of the lateness of the new part drawing date, required replacement. Weapons that did not get arsenal rebuilds soldiered on with the earlier stud.

The making of gauges, fixtures, jigs and special purpose tools used in the manufacture, assembly and rebuilding of weapons was a large part of the process.  Some of these special purpose items had letter prefix drawings and some were identified by a special tool room system of numbering drawings.

The last item on the list of top cover feed parts that we are going to discuss is the belt feed lever. The first listed M1919A4 belt feed lever, B144825 did not share the drawing number with the original M1917 belt feed lever  shown on drawing 51-10-18D which had a letter prefix conversion drawing number of C8460 dated June 1, 1931. This is the original date common with the majority of M1917 drawing conversions to the letter prefix system. There are only two C8460 drawings available and they carry a note at the bottom left “For future manufacture use B-17503”.

It would appear that while there was a C8460 drawing it was not used to produce any belt feed levers.  It unknown why Ordnance even bothered to produce the C8460 drawing in 1931, unless it was done in error, since they already had a B17503 drawing dating from 1924. The operating lug on this drawing, 51-10-18D/ C8460, is .446 (-003) long.

The earliest B144825 drawing is dated January 23, 1936 and it appears to be a nearly exact copy of the C8460 but has no authorizing signatures it also has a Revision 1. B144825 next appears on September 30, 1936 this time with authorizing signatures and a date common with most of the M1919A4 drawings it has a lug length of ,446 (-.003). B144825 Revision 1, 3-10-39, changes the lug length to .510 (-005).  

The earliest B17503 drawing available is dated April 4, 1927 and is a Colt’s Patent Firearms drawing endorsed by the signature of Ordnance officers at the Springfield Armory.   This drawing was revised on 1-31-36 to add a note qualifying the tolerances allowed to the part not otherwise shown on the drawing, and to add Major Item 51-77 the M1919A2 to the list of weapons this drawing pertained to.  This drawing showed a lug length of .520  

Other weapons listed in the “Drawing Pertains To” block on the Colt’s drawing were the M1917, M1919 Tank Gun, M1919 Aircraft gun, and the “M18M1” also an aircraft weapon. Plate A in the RIA Notes…Modified document dated March, 1938 shows the B17503 belt feed lever as the type lever used on the modified M1919A4 gun.  All of the dimensions shown on Plate A agree completely with Revision 1 of the same numbered drawing that originated at Colt’s. Revision 2 to B17503 dated 2-1-38 redrew the drawing at RIA and removed the reference to Colt’s. B17503 Revision 3 dated 5-27-39 added the M1919A4 Fixed and Flexible and shortened the operating lug to .468 (-.005)    

During the evolution of the M1919A4 several designs of belt feed levers were developed; the earliest was the B144825 Revision 1 with the .510 lug. Followed by the B17503 which was designed, apparently, for both the M1917 and the M1919 family with the .520 lug and ending up with a .468 (-.005) dimensioned lug shown on B17503 Revision 19, 12-27-45  Even though the B144825 lever design was superseded by the B17503 lever with a shorter lug dating from 5-27-39 it was classed as “useable” through the 1944 SNL. The B144825 lever had a stock number with an A006 prefix indicating that it was for the A4 only while the B17503 had an A005 prefix indicating it would also fit the M1917/M1917A1.

Ordnance, in one of their “Houdini” acts, on 5-27-39 by Revision 2 to B17503 shortened the .520 operating lug to .468 and did away with drawing and belt feed lever B144825 because both parts were functionally identical even though the B144825 had the .510 lug. The last of the belt feed lever drawings for the WWII period is (B) 6017503 Revision 20 dated 5-10-48 which is the drawing that redrew B17503 including all the previous revisions and converting the drawing to the 7 digit format.  This drawing also changed the piecemark, now called a part number, to 6017503 and finalized the lug length at .468. I suspect they kept the B17503 drawing number because this number now appeared in all the publications for the M1917’s.  Switching the drawing around avoided reprinting many documents and revising the M1917 drawings. This is another case much like the drawing number/parts swap that occurred with the M1919 slotted/holes barrel jackets. In any event, the operating lug remained nearly the same .468 length except that its tolerances changed along with the dimension called out on the drawing. 

The end result was that different drawing revisions expressed the lug length in different ways but it was about the same length from about May 1939 to the end of its service life. Belt feed levers were the only part in the top cover and its attached components that was required to be piecemarked starting with B144825 in September of 1936 which was marked B144825 and likely marked with the producers identification. The Colt’s drawing B17503 Revision 1 dated April 4, 1927 changed the piecemark to B17503-1 but there were no instructions to imprint the part. Expected piecemarks on belt feed levers believed to be: 

B144825 (likely very rare) there was a Revision 1 to this drawing, but it did not change the piecemark.
B17503-3, beginning 5-27-39
B17503-4, beginning 3-29-40
B17503-6, beginning 4-18-41
B17503-14, beginning 8-14-43

The B17503-14 and subsequent belt feed levers featured a slightly thicker (.010) lever except for the 1.25 inches on the end of the lever that fit into the belt feed slide.  Regardless of this minor change, the reason for which is unknown, all of the B17503 levers are interchangeable.

6017503, beginning 5-10-48

All of these should have various manufacturers’ codes imprinted on the lever.  The prescribed location for this piecemark was on the top (cover side) of the lever and on the operating lug end in 1/16 inch characters. I have a RIA B17503-3 lever marked on the bottom side so even they didn’t always follow their own rules. The March 1938 “Notes Modified” publication illustrated belt feed lever B17503 shown below from Plate A.

Fig 20

The text from “Notes… Modified” describes the change as lengthening the operating lug to .520 inches.  The “operating lug” is the stud that travels in the serpentine track in the top of the bolt. The original M1917 belt feed lever had a lug length of .446. Drawing B17503 went through 19 revisions before the 5-10-48 renumbering (Revision 20) when it became (B) 6017503. It was authorized to be used in both the M1917A1 and all of the M1919’s. In the renumbering process B size drawings had 6,000,000 added to the original digits 17503. There is an RIA pamphlet, and some authors including myself, that have described this B size renumbering process as “adding the numbers 60” to the original drawing numbers. This is somewhat misleading.

All drawings were supposed to end up with 7 digits and while adding 60 to the original drawing number in the case of the belt feed lever produces a seven digit number if the original drawing had contained only four digits adding 60 would only produce a 6 digit number. If the original B size drawing contained 6 digits like B224849, a part of M1919A6 butt stock, and you added 60 to that you would end up with an 8 digit drawing number.  

Even though the 1943 Ordnance plan to have a common drawing, part and stock number containing no alpha characters was implemented, and every drawing was redrawn/renumbered, the practice of including the letter size continued because that’s the first step in locating the drawing because that’s how the drawings were filed. All the A size filed in one cabinet all the B size in another, and so on. 

In addition, as we have seen in the previous example, not all B size drawings would start with 60, the A6 butt stock part/drawing/stock number became 6224849. The drawing letter size in 7 digit conversions was usually placed in a box adjacent to but separate from the drawing number. The belt feed lever is another example of a large number of revisions to a part drawing which didn’t have any effect on interchangeability, even though the belt feed lever was required to be imprinted with the current piecemark.  

In addition the B144825 belt feed lever continued to be “useable” until the supply was eventually exhausted even though the B17503 lever was the only lever to appear on the B169913 drawing dated 4-7-42 listing all of the parts of the M1919’s required to be marked. It was “useable” because it was for all practical purposes the same part with a different nomenclature.

Fig. 21
This Saginaw drawing, another of Jodie’s finds, from May of 1942 which they numbered B-17503 in their title block and someone at RIA hand wrote B17503 shows an experimental belt feed lever featuring a slight offset bend on the end of the lever that operates the belt feed slide. It also slightly lowered the height of the upper pivot bearing area and lengthened by the same amount, approximately .020 inches, the operating lug. It does not appear that this design ever went anywhere. This and the previous Saginaw drawings, most likely, have never been published before.

Fig 22

This drawing which appears to have originated at Springfield Armory shows a “modification” to the B17503 design that features a roller on the operating lug of the belt feed lever. Section AA shows the lug end of the belt feed lever with a shaft for the roller and a hole through the lug for the rivet needed to secure the roller onto the lever.  The roller is shown right of center and the entire assembly to the left of center.  Because of the poor quality of the microfilm we have been unable determine a date when this was proposed.

The drawing is a nonstandard size and was microfilmed in two pieces.  It is likely that this drawing was transferred to the Rock Island Arsenal from the Springfield Armory along with all of the other Caliber .30 BMG documents when RIA assumed responsibility for production of these weapons. Like the Saginaw proposed belt feed lever shown in Figure 16 there is no evidence that this design ever went beyond the drawing stage.

These last two drawings give us a unique look at what was going on behind the scenes. This concludes our discussion of the top covers and their associated feed system parts. Additional articles describing the M191A4 design history will be forthcoming, stay tuned. I hope the reader has found our efforts both entertaining and informative.  

As always, comments are welcome just click on the “send e-mail” tab on the “Research” page. 

As usual, I have had a lot of help with the research; however I am solely responsible for the content, if there are errors or misstatements I bear the responsibility.


Rock Island Arsenal Museum, Jodie Creen Wesemann who supplied all of the ordnance drawings and publications featured or quoted in this article. Jo is one of those rare persons with an instinct for the important.  The Saginaw Steering Gear produced drawings and the “roller lug” Springfield Armory drawings were not requested by the author but Jo included them because she sensed that they were unique, and as usual she was correct.  Without the Museum’s cooperation none of the things displayed here would be possible.  

Thanks to my good friend and an important contributor Rollin Lofdahl who is always there to help in deciphering all of this information.

Thanks to all those members of the and forums who have helped with pictures and comments, and for sharing the really odd finds that crop up now and again.  

And a special thanks to Nosegunner, without his Web site and the time he spends making my scribbling presentable we couldn’t share this unique information.

Photos are individually credited; those lacking a credit are the author’s.

The Browning Machine Gun Vol. 1, Dolf L. Goldsmith, Collector Grade Publications, INC

Hard Rain, History of the Browning Machine Guns, Frank Iannamico, Moose Lake Publishing, LLC

U.S. Infantry Weapons of WWII, Bruce N. Canfield, Andrew Mowbray Publishers