THE BOLT LATCH AND CALIBER .30 AIR COOLED
BROWNING MACHINE GUNS IN U.S. SERVICE
It’s always best to start from the beginning. In the case of the relationship between the bolt latch and the air cooled Browning in caliber .30 we have to go back to 1918.
Tanks made their first appearance late in WWI they were crude and unreliable vehicles prone to breakdown and the asphyxiation of their crews. Other than the psychological effect on opposing troops they did little to further the allies cause.
Tanks were armed with a variety of weapons including machineguns. When the U.S. became directly involved in the war in April 1917 we had exactly zero tanks, and precious little else including machine guns.
Fortunately, we had one resource that other countries didn’t have, we had John Moses Browning, a self taught mechanical genius who devoted his entire life to firearms design.
Browning had already designed and patented the Model of 1895 ’Potato Digger” gas operated machine gun and Colt had manufactured some quantity of these unusual weapons. Browning had also done much preliminary work on a recoil operated machine gun generally known as the model of 1901 that would become the Model of 1917 water cooled “Heavy” that would serve the U.S. for over 40 years.
When the U.S. entered the war we were dependant on our allies for much of our automatic weapons supply. As the war progressed we began to produce more of our own weaponry including machine guns and tanks. Naturally, there is a tendency for any country to try to use indigenous weapons if possible, and when the subject of arming tanks with machine guns came up, the 1917 was considered.
The Tank Corps was not interested in a water cooled weapon for obvious reasons. It was too heavy and bulky, subject to water jacket damage, and the design did not lend itself to the ball type mounting required to protect tank crews from enemy fire.
Since we were already producing M1917’s John Browning, working at Colt’s factory decided that if the water jacket was eliminated, the barrel shortened and made heavier, and a different sighting system was developed, the M1917 casing and internal parts could be used as a basis for a weapon suitable for tank use. As usual, he was right.
The Army, being the Army, developed a laundry list of requirements for what was to become the Browning Tank Machine Gun, Model of 1919.
One of the biggest problems in the development of an air cooled machine gun was the fact that you lost the efficiency of the water cooling design.
Since the M1917 was originally designed as a water cooled weapon, air cooling meant that the barrel had to be much heavier, some sort of support for the muzzle end of the barrel of the new weapon had to be provided, the weight of the recoiling parts increased substantially, affecting the operation and firing rate of the weapon.
The biggest problem, however, was the dissipation of the heat created during firing.
This problem was not nearly as serious in air cooled weapons used in aircraft because the weapons were only used while flying, and the large amount of air flow dissipated the heat.
The lack of efficient heat dissipation, when combined with the fact that the Browning design fired from a closed bolt, that is, when the weapon reached the end of the firing cycle the bolt was closed and locked on a chambered cartridge, the firing pin was cocked awaiting action of the sear to release it to start the firing cycle over again, made chamber temperatures critical.
Early machineguns were water cooled for good reason. Relatively small amounts of water have the ability to absorb large quantities of heat energy.
This basic law of physics is well known to anyone waiting for a quart of water to boil on the kitchen stove. Heat is the deadly enemy of all mechanical devices, especially firearms.
Cartridges contain all of the elements necessary to function, all they need is heat energy to ignite the propellant creating the high pressure gasses that propels the projectile out of the barrel.
Normally, in small arms, this heat energy is supplied by the primer being struck by the firing pin converting mechanical energy into heat energy.
Unfortunately, the propellant will respond to any source of heat energy including a hot chamber that raises the temperature of the propellant enough to cause the propellant to ignite.
This situation is referred to as a “cook off”.
The trigger is never pulled, the firing pin, never strikes the primer but the weapon suffers an unintentional discharge.
Normally a firearm “safety” locks the internal parts in such a way that prevents the trigger from being inadvertently operated or the weapon firing from being dropped or subject to some other mechanical force.
The Model of 1903 rifle, the famous Springfield's three position Mauser style safety is a good example.
The safety can only be applied when the striker is cocked, and the bolt is closed and locked.
When applied, the safety will not allow the trigger to be operated or the bolt to be opened. The safety has a middle position which will not allow the trigger to be pulled, but will allow the bolt to be opened. This middle position is to allow unloading of the rifle while safeguarding the trigger from unintentional manipulation.
The third position allows the trigger to be pulled and the bolt to be operated.
The Army conducted extensive tests to determine how heat build up would affect the new air cooled Browning.
The tests determined that after firing 400 rounds at a rate of 400 rounds per minute that a chambered cartridge would cook off in about 16 seconds. After firing another 500 rounds the cook off time dropped to 11 seconds.
This, rightfully, caused much concern, on November 2, 1918, 9 days before the Armistice ending WWI was signed, the Army ordered that the bolt latch be fabricated and mounted on the Tank guns. This is the first known mention of the bolt latch.
The Army’s concern was twofold, first, the obvious safety hazard of unintended discharge. Equally important was the fact that due to the Browning design, cook offs could disable or damage the weapon. Keep in mind that the Caliber .30 Model of 1906 cartridge operates relatively high pressure, about the 50,000 PSI level.
In the Browning closed bolt firing design, when a cartridge is chambered, the firing pin is held inside the bolt by the sear.
Under normal firing conditions when the sear releases the firing pin and it strikes the primer, the pin continues to protrude, under spring pressure, from the bolt face, preventing the primer cup material from flowing into the firing pin hole during primer set back that results from the ignition and combustion of the propellant and the subsequent rising pressure inside the case.
In the case of a cook off, the firing pin remains retracted into the bolt, and primer cup material can flow back into the firing pin hole. This effectively locks the fired cartridge case into the bolt face, jamming the action by preventing the ejector from forcing the fired case downward and out of the bolt, or worse, breaking off the ejector completely. This malfunction is made much worse by the design of the bolt where the cartridge is held to the bolt face on two sides by the T slot.
In a weapon like the M1903 rifle the same result would occur from a cook off, however, the case is held to the bolt face only on a portion of one side by the extractor, and a cook off case could be pried off by the camming action of pulling laterally on the case neck, or the bolt could be removed from the weapon, and the extractor removed and the case pulled straight off the bolt.
While a cook off condition could occur in a bolt action rifle, due to it’s rate of fire, it is much less likely. Ordnance directives constantly remind troops to keep ammunition out of direct sunlight which can raise the temperature of the propellant greatly increasing the risk of a cookoff
Cramped conditions inside early tanks made disassembling the weapon inside the tank almost impossible. Use of the bolt latch went a long way to preventing this problem, and by holding the bolt rearward, heat dissipation was much improved by exposing the chamber to more air flow.
Brownings’s M1918 Browning Automatic Rifle solved this problem by firing from an open bolt.
When the tank machine gun was developed, WWI was nearly over, and, after the Armistice, the contracts awarded to New England Westinghouse to manufacture it were reduced from 40,000 units to 10,000 and later to just 500 weapons eventually increasing to about 1300.
There were actually two different types of tank guns, one had a cast bronze tube style rear sight, the other had no sights at all, both were referred to as “Model of 1919”.
This rather profound difference was not deemed of much importance until late spring of 1929, when the Ordnance Department, decided to call the model with the sight brackets “Gun, Machine, Caliber .30, Browning, Tank, 1919A1“. The model with no sights at all was given the same title, except for dropping the “A1” from the nomenclature.
Both of these weapons had the slotted barrel jacket, an 18” barrel, and a bolt latch. This nomenclature lasted for about two years until Ordnance revisited the whole issue, and decided to name the tank guns being developed for Cavalry use 1919A1.
This meant that there were now two weapons with the same designation, one for tanks, with brackets for telescopic sights, and one for cavalry use with different sights. In 1932 it was decided to rename the 1919A1 tank guns converted for use by the cavalry to 1919A2.
Some Einstein in Ordnance, thought this might clear the air, it didn’t.
The Browning Tank Machine Gun, while developed during 1918, actually carried a designation of “Model of 1919”.
The WWI era of rapid
weapons development overwhelmed
the Army’s weapons nomenclature system. By the end of the War,
there were three Model Of 1917 weapons, a rifle, a Colt or Smith &
Wesson Caliber .45 revolver, and a water
cooled machine gun. I
suspect that the Army, not wanting to make a bad situation worse with
the tank gun, decided to call it the “Model of 1919” since they already
had had a “Model of 1918”
the famous, and much revered, Browning Machine Rifle AKA the Browning Automatic Rifle.
AKA the Browning Automatic Rifle.
It would be good to remember that all of these air cooled weapons had the bolt latch installed.
There is no way to tell for sure, but it appears that Ordnance powers that be felt that any air cooled Browning whether a tank gun, a tank gun converted for cavalry use, a 1919A4 converted from a 1917, or a purpose built 1919A4 Fixed or Flexible needed a bolt latch because of the possibility of cook offs.
This line of reasoning continued until about January 1944, when the bolt latch was dropped as an available part. Ordnance document “The Ordinance Catalog, List of All Parts, SNL A6” dated May 28, 1941, and the same document dated 6 September, 1943 lists the bolt latch and the attaching rivet as obtainable parts. Ordnance publication “ORD 9 SNL A6 “, the document that replaced list of all parts, dated January 1944 no longer lists the bolt latch, or the rivet. Sometime between September, 1943 and January 1944, the latch was discontinued as an authorized repair part.
As of this writing, no one seems to have come up with the actual document, or minutes of some Ordinance Committee meeting that first eliminated what Dolf Goldsmith, probably the worlds most knowledgeable Browning authority, called “that excellent accessory”.
However, we have examined drawings 51-83-1 and 51-84-1 which are the full size right side views of the M1919A4 Fixed and Flexible respectively.
Both of these drawings have revisions dated 5-13-43 after which the bolt latch is still shown. The next revision dated 5-31-43 does not show a latch on either weapon, and neither do subsequent revisions to these drawings.
We have, however discovered the Draftsman's Work Order eliminating
the bolt latch with the notation "Mandatory must be applied
We have, however discovered the Draftsman's Work Order eliminating the bolt latch with the notation "Mandatory must be applied immediately".
The way things worked in the Ordnance Department, the supervising facility, in this instance RIA, controlled the engineering changes either ordered by by the RIA engineers or the Ordnance Department.
However, these changes were implemented at the factory by the Ordnance District supervising the actual production of the item. The reason for this is that Ordnance desired to avoid additional charges to the government for the changes.
In the case of the bolt latch Ordnance bit the bullet and required the change to be effective at once probably because it would eliminate two parts, the latch and the attaching rivet and the hole in the right side plate and the labor required to perform the work operations.
It is also likely that any right side plates in the production pipeline with the rivet hole already drilled were simply assembled into weapons, and the hole left unfilled.
D35411 Revision 14
In addition, Drawing D35411 Revision 13, the right side plate, dated 1-15-43 shows the hole for the bolt latch rivet, while the next Revision (14) dated 5-31-43 no longer shows the rivet hole and changes the piece mark to D35411-14.
We believe that this is sufficient evidence to declare that the bolt latch was officially discontinued on new production weapons or about 31 May, 1943.
The actual date when the bolt latch stopped being installed on production weapons is unknown, but due to the mandatory immediate nature of the change document it must have been close to the May date.
One thing is for certain, the bolt latch was considered a required feature for any ground type air-cooled Browning from November 1918 until mid 1943.
At the time of WWI the Ordnance Department used the Class and Division naming convention to identify weapons systems and the drawings used to produce them.
This system assigned a Class number to a type of weapon, automatic weapons were Class 51 and a Division number to identify individual weapons, this combination produced a Major Item number.
The Model of 1917 Machine Gun was assigned the Major Item identifier of 51-10, and the Model of 1919 Tank Machine Gun 51-18 . The drawings used to produce weapons used the Class and Division derived Major Item number combined with a sheet number and in some cases a letter to identify the detail on the sheet to depict a component part.
The first drawing of the bolt latch is 51-18-7C, and it is dated November 20, 1918, 9 days after the Armistice ending WWI was signed. This drawing shows not only the bolt latch as detail “C” assigned a “Piece Mark” of 7A, but the auxiliary trigger as detail “A”.
It also shows an area on the bottom of the bolt latch about 3 inches from the rear, and 2 inches long bent outward 5/16th of an inch.
Apparently, Ordnance desired this, for lack of an official name “finger flange”' to make the latch easier to grasp and apply.
Revision 1 to this drawing made minor dimensional changes to the latch and changed the “Piece Mark” to 7A1.
The “finger flange “continued in use until Revision 2
of the drawing dated October 20, 1920 when it was eliminated, and the
“Piece Part” was changed to 7A2.
Revision 2 also slightly changed the contour of the rear end of the
bolt latch from a squared profile to a round one. This
change was likely instituted to ease manufacture.
Revision 2 also slightly changed the contour of the rear end of the bolt latch from a squared profile to a round one. This change was likely instituted to ease manufacture.
The final Class and Division drawing showing the latch was Revision 3 dated June 1, 1931 which changed the drawing number to the new letter prefix numbering system . It became B131295.
Photo courtesy of Jon Moran
Photo courtesy of Jon Moran
Here a picture of a M1919A2 equipped with a "finger flange" bolt
latch and the original bottom plate with reinforcing stirrup.
Here a picture of a M1919A2 equipped with a "finger flange" bolt latch and the original bottom plate with reinforcing stirrup.
The 1941 Standard
Nomenclature List for the M1919 series lists the latch and the rivet
priced at $.23 for the latch and $.12 for the rivet and shows that they
are intended for both the M1919A2 and the M1919A4.
The 1941 Standard Nomenclature List for the M1919 series lists the latch and the rivet priced at $.23 for the latch and $.12 for the rivet and shows that they are intended for both the M1919A2 and the M1919A4.
The attaching rivet was first shown on drawing
51-18-8A 1918 and assigned a “Piece Mark” of 8G2 also went
through a series of dimensional/ “Piece Part” number changes until the
final Class and Division drawing
Revision 5, shown above, dated June 1, 1931 changed the drawing number to
The June 1, 1931 date appears on most of the
1917 and M1919 tank gun drawings as the effective date of conversion to
letter prefix drawings.
The June 1, 1931 date appears on most of the 1917 and M1919 tank gun drawings as the effective date of conversion to letter prefix drawings.
You will notice that, in the Class and Division system of identifying drawings of parts, there is no connection between the detail letter where the part is depicted on a particular sheet and the “Piece Part” number assigned. Sometimes a part depicted in detail A of a sheet will have A in the piece part number, and sometimes it will be some other letter, or no letter at all, as in the case of a sheet containing only one part.
In the letter prefix system of identifying drawings, only one part was to be shown on each drawing.
The attaching rivet drawing A13257, starts out with a date of June 1, 1931 and goes through 7 Revisions the last one dated 10-8-42. The “Drawing Pertains To” Block lists 51-18, BTMG (Browning Tank Machine Gun) M19 (Model of 1919), 51-77 .30BMG M19A2 ( the elusive Cavalry version), 51-83 .30BMG M19A4 (FXD) (M1919A4 Fixed), an 51-84 .30 BMG M19A4 (FLEX) (M1919A4 Flexible).
The bolt latch depicted on B131295 starts out with a date of June 1, 1931 goes through 10 Revisions, the last available drawing being Revision 10, dated 5-13-43 which besides shortening the latch overall length likely as a material conservation issue, drops references to the tank gun and the M1919A2 which had been obsolete for several years. Note the bend in the latch intended to place tension on the latch to prevent it from flopping around during firing.
Over the course of production several minor changes in fabrication occurred the two most obvious are the method of forming the rear end of the latch and the material used.
Matt D photo
Top example is the original style where the end appears to have been formed around a rectangular mandrel and forge welded. These were likely made up to 1920 or so when the finger flange was eliminated. Originally the latches for the tank gun were cut from Class A sheet steel but by the time the tank gun Class and Division drawings were updated to letter prefix type in June 1931 the Ordnance Department had started using standard commercial steel identifying information and the preferred material became X1020 steel.
Matt D photo
To speed up production and reduce cost Revision 6 to the latch drawing, dated 3-26-42, allowed the use of mill or band edge stock which resulted in differences in appearance of the finished product, especially the edges. It appears that the method of shearing the stock and rolling the rear portion produced even more variations in appearance that may be tied to individual fabricators. RIA and Saginaw Steering Gear made the majority of parts used in their production while Buffalo Arms used a vast network of sub-contractors to produce parts for their production. It is not known with any certainty who actually produced latches for which companies.
Matt D photo
Manufacturing tolerances for the notch that engaged the bolt handle also changed by Revision 9 in January 1943 this slightly alters the notch profile and produces even more variations as seen in the examples pictured above. The latch shown at the top had the "finger flange" added for a reproduction of a period weapon.
Revision 10 also shortened the length of the latch from 11.375 (11 3/8) inches to 7.750 (7 ¾) inches. This shortening of the latch would not require that the rivet hole be relocated on the right side plate.
Revision 10 just shortened the distance from the rolled end of the latch to the bolt handle retaining notch. This latch appeared only about two weeks prior to the order eliminating the latch altogether. It is unlikely that any were actually fabricated.
Rollin Lofdahl and DRH of the 1919a4.com forum photo
Rollin Lofdahl and DRH of the 1919a4.com forum photo
So far none of the Revision 10 latches have surfaced.
We have not observed any drawings that directed the that the
“Piece Mark’ be placed on the latch, however, we have observed some
imprinted B131295-5. It does
not have a manufacturers identification.
This is probably a Buffalo Arms produced latch as they had a
tendency to mark all kinds of parts that Ordnance did not require to be
This is probably a Buffalo Arms produced latch as they had a tendency to mark all kinds of parts that Ordnance did not require to be marked.
Russ Brindisi photo
Russ Brindisi photo
There are also have been a few seen
marked B131295 RIA.
There are also have been a few seen marked B131295 RIA.
Cut from 6 September1943 SNL showing m1919A6 with no bolt latch.
Cut from 6 September 1943 SNL showing M1919A4 casing assembly with bolt latch.
The next development in the M1919 series was the M1919A6, this weapon was a modified M1919A4. It featured a different barrel, barrel jacket, folding bi-pod mounted between a newly designed front barrel bearing and after the spring of 1944 a booster, a carrying handle, and a sheet metal butt stock.
This was a stopgap measure to produce a weapon that was intended to be more portable and easier to deploy than a standard tripod mounted M1919A4.
This development was not an unqualified success. The attempt to develop a more easily deployed light machine gun began at least as early as November 20, 1942, and it involved much wrangling between the Ordnance Department and the Infantry.
Ordnance wanted to develop a new weapon better suited to its prospective role, while the Infantry wanted something to place in the field yesterday.
At first Ordnance agreed to fabricate a parts kit to be distributed as an accessory to the M1919A4, after more wrangling it was decided to purpose build the new weapon as the M1919A6 a Substitute Standard.
A camel was once described as a horse designed by a committee, in this case, the horse/camel was the M1919A6.
Everybody seemed to get something, Ordnance continued to work on a new weapon, which turned out to be the M60GP machine gun introduced in 1957.
Good thing the infantry didn’t decide to wait.
The Infantry got to field something quickly, even if it wasn’t exactly what they wanted.
Everybody won, or lost, depending on your point of view.
By the time all of the A6 wrangling was completed it was February of 1943, and Saginaw Steering Gear was selected to be the builder.
Besides being the premier M1919A4 builder, Saginaw also experimented with casting certain M1919A4 parts from an alloy they had developed called ArmaSteel.
The ArmaSteel castings were designed to replace parts that had formerly been forged and machined.
Casting parts speeded up production and cut scrap losses, saving steel, and lowering the price of the weapon. In the fall of 1942 Saginaw produced cast casings, and other parts including a pistol grip back plate and forwarded them to the Aberdeen Proving Grounds for testing.
As a result of these tests, the final report recommended, besides approving using some of the castings, but not the casing, that the bolt latch be removed.
Apparently, the bolt latch, installed on the cast receivers caused mounting problems on tripods. What this had to do with conventionally fabricated casings (receivers) is debatable. My guess is that someone in authority who had wanted to remove the bolt latch all along used this as a reason to do so.
Apparently, the D35411 Revision 14 right side plate drawings eliminating the bolt latch rivet hole arrived before Saginaw received the order for production of the purpose built M1919A6 in August of 1943.
The Ordnance Catalog SNL A6, List of All Parts, dated September, 1943 lists parts for M1919A4 and for the first time mentions the M1919A6.
The bolt latch is listed as an available part, but only for the M1919A4. Page 49 of this document shows an A6 it does not have a bolt latch.
Page 50 shows the Casing Group Parts of a M1919A4 with the bolt latch.
This would lead one to believe that
purpose built M1919A6 weapons were never equipped with bolt latches, but
purpose built M1919A4’s including A4’s rebuilt from M1917’s
M1919A2’s, and anything found lying around, were equipped with
bolt latches at least until some time shortly after 31 May, 1943 when
all references to the bolt latch disappeared from Ordnance drawings .
Having said all that, we have to say this. While no purpose built M1919A6 was supposed to have a bolt latch, some M1919A6’s were fabricated from M1919A4 Fixed weapons that were not being used for tank armament.
This rebuilding process may have required removal of previously installed bolt latches, we don’t know. We also don’t know if when M1919A4’s were refurbished after 1943 but during WWII if directives required removal of the bolt latch.
If the weapon was intact without the bolt latch, that is if the right side plate were present, it would be an simple matter to just look for the rivet hole, and compare the serial number to the list of known serial numbers to determine when the weapon was produced. It is also possible that even after the directive eliminating the bolt latch arrived that the rivet hole was still being drilled even though it was not required.
Just check out the top cover of a .30 BMG. See that little dimple near the pivot pin hole? It’s still there from early M1917 production that used a spring retainer for the pivot pin similar to the pin used for locking the trigger housing and gas tube assemblies on a BAR.
We have a Base Shop Data book for the M1919A4 with a Table of Contents dated 11 October, 1943, still showing the method of tightening the bolt latch river and the tools required to perform this overhaul operation. Armorers rebuilding weapons using this document would most likely either repair or replace the latch.
We went through all of this to get to this point.
A summation of the bolt latch issue is:
There are three types of latches, the pre-1920 version with the “finger flange”, the common 11 3/8 inch latch without the “finger flange”, and the as yet unseen B131295 Revision 10, 7 ¾ inch short latch.
All air cooled .30 caliber ground Browning Machine Guns, regardless of whether they were made by converting M1917, M1919A2, M1919 Tank guns, or any other variant, or were purpose built as M1919A4’s, WITH THE EXCEPTION of purpose built M1919A5 and M1919A6‘s, from November 1918 until about May 31,1943 were supposed to have a bolt latch.
Does every one of the previously mentioned weapons have a bolt latch?
The answer is a qualified no, there were somewhere around 400,000 of these weapons produced doubtless many of them had unauthorized “field modifications” preformed on them.
In addition, after January, 1944 the bolt latch was no longer listed as an available part.
If it were damaged, and no spare was handy, it was merely eliminated, probably by drilling out the rivet and discarding the both parts, or the field expedient of just breaking it off.
The final official word on bolt latches is dated 9 August 1949 and is contained in Ordnance Technical Bulletin TB ORD 366
Overhaul and Rebuild Standards for Small Arms Material,
Section IV, Special Rebuild Standards-Cal.30 Machine Guns and Mounts,
Receiver Group Assembly,
(2) Cal.30 ground guns
(c) Remove all bolt latches B131295 from guns M1919A4.
We hope that this article will put to rest some of the controversy surrounding the bolt latch. Given the nature of 1919 community’s propensity to wrangle over things, this hope is likely in vain.
Dolf, please forgive me, but I never liked the bolt latch, I could just see it scratching up the side of my Saginaw/Ohio Ordnance/Izzy M1919a4/A6 semi-auto mutt.
I always use a method of preventing cook offs or other unintentional discharges similar to the one described in FM 23-45 dated 1940 which describes placing a ’clearing block“, a simple piece of wood between the retracted bolt and the empty chamber. Instead of the wooden block, I use a piece of scrap yellow PVC gas tubing. about 2 feet long, it sticks up above the open top cover and out the bottom of the casing showing me and everyone else that the weapon is cleared.
CREDITS AND ACKNOWLEDGEMENTS
Well, it’s the usual suspects:
Jodie Creen Wesemann, Rock Island Arsenal Museum for all of the
Ordnance drawings, and Ordnance publications quoted.
Jodie Creen Wesemann, Rock Island Arsenal Museum for all of the Ordnance drawings, and Ordnance publications quoted.
The Browning Machine Gun, Vol. 1, Dolf L. Goldsmith, Collector Grade Publications, Inc.
Members of the 1919a4 Forum, and the director/owner of the forum “SHOTS” for their assistance.
My personal thanks to my “Editor” and an important contributor Rollin Lofdahl not only a very knowledgeable guy, but the rudder of the ship. He keeps me pointed in the right direction, not an easy task.
A special thanks to Matt Danker, who's sharp eyes caught the
difference in the roll contour of the latch end.
A special thanks to Matt Danker, who's sharp eyes caught the difference in the roll contour of the latch end.