RIA: Walther P38 Development

The Walther P38 was adopted by Germany in 1938 as a replacement for the P08 Luger – not really because the Luger was a bad pistol, but because it was an expensive pistol. Walther began development of its replacement in 1932 with two different development tracks – one was a scaled-up Model PP blowback in 9x19mm and the other was the locked-breech design that would become the P38.

The initial prototypes look externally quite similar to the final P38, although the locking system went through several changes and the controls did as well. Several of the early developmental models used shrouded hammers.

In this video I will take a look at both initial “MP” pistols (the blowback and the locked breech), then the Armee Pistole (aka the AP) in its standard configuration and also a long barreled model with a shoulder stock, then the second Model MP, and finally the HP which was the commercial model of the final P38. In addition, I will check out a sheet metal prototype of the locked breech model form the very beginning of the development program.


  1. “Walther P38”
    http://hungariae.com/Bio.htm (in section Pál Dedai Király) states that:
    In 1929 Király designed an advanced 9x19mm pistol, designated as ‘KD Danuvia’. 20 guns were manufactured for trials. Among others, Germany was very interested in this pistol, but the 1929-33 market crash put a hold on new developments worldwide. Interestingly, the later Walther P.38 displays extreme ‘similarities’ with the KD Danuvia. The two guns internally are the same, outside differences were the P.38’s distinctive grips and a shorter barrel.
    However I’m not able to find blueprints of mentioned KD Danuvia automatic pistol, so I can’t say whatever above cited statement is true or false. Maybe readers more proficient with Hungarian automatic pistols of 1920s will be able to prove (deliver evidence) or deny it.

    • Greetings from Hungary!

      Unfortunately Manowar’s Hungarian Weapons & History is unreliable: it’s only a starting point, nothing else. Unfortunately many documents were lost during WW2 and the communist era.
      Péter Pap, who is a researcher of the hungarian Museum of Military History, investigated that in one of his articles, the Király’s KD and the Walther P38 were developed independent from each other. The english abstract of his article:
      “For executing field trial of the 9 mm pistol designated as KD and designed by Pál Király, the Ministry of Defence ordered from the Danuvia Industrial and Trading Company 20 experimental pieces of this pistol. The Király’s KD pistol was set forth by a study-aid published in 1935. No any other document or data on this experimental pistol are available for the time being, and no further information is known about whereness of them.” (Haditechnika magazin, 2014/2)
      According to Péter Pap and the only known photo of the pistol, I think the Király’s KD and Walther P38 connection is simply a legend. Who knows, why and from whom emerged this legend.

      Best wishes, Poresz

      P.S.: Sorry for my poor english.

      • People often assume common history of similar designs. Somebody makes a claim that some things have to be related, because they are similar and contemporary, and the legend starts to live, even if there is no historical evidence to back it. In a like manner people often think that the 6.5x52mm Carcano cartridge was somehow designed by Mannlicher, because it is very similar to the 6.5×54mm Mannlicher–Schönauer cartridge, and the Italian M91 rifle is sometimes (inaccurately) called a Mannlicher-Carcano. However, there is no historical evidence that Mannlicher had anything to do with the 6.5x52mm cartridge.

      • Poresz:
        “P.S.: Sorry for my poor english.” It’s FAR better than my Hungarian, French, German, Chinese and sometimes English; the content is the most important thing so please NEVER hesitate to post!!!
        PS: Your typing is also far better than mine.

  2. Great video Ian. One niggling little point, since I’ve noticed something from this video and some of your previous ones — Auf Deutsch gibt es keine stillen Buchstaben (there are no silent letters in German). So “Pistole” (nouns are always capitalized) is pronounced “pistoluh” in German.

  3. Ian,
    Not full reason why Germans replace Luger by P38,not because it was expansive in manufacturing but because Germans was seeking pistol with open hammer that could be operated manualy….

    • Oh boy, now you broke it. Everywhere before I read that P38 was introduced because it was “cheaper to make” than P08. Now, I have o rewire myself. 🙂

  4. Actually, I recall reading that the ‘trademark’ issue of this swing lock, being cracking frames was there right from beginning. It took considerable effort to suppress it, just to pop up on Beretta 92 again. Now Beretta eventually recognized it and changed their design approach.

    • Uh, Denny, the cracked Beretta pistols came from the users putting overcharged SMG rated cartridges into the magazines. Considering that, the pistols are quite fortunate they did not immediately explode upon firing.

      • Sorry master Chern, I don’t buy it. At time of trials I paid close attention and still recall some of it. Beretta had problem with material and heat treatment. (Whole gun is kind of crappy.)

        But, it was not case with frame; it was slide to be blamed. One time there was involved certain (M9 bayonets making) company from California who had simplest solution possible – remove nonsense cutout. Red tape prevented it, as usual.

        As far as ammo in U.S. service it is standard NATO 9mm Ball; there is no SMG version fielded (no SMGs).

        • I meant the Navy SEALS loaded their M9s with rounds generally used only by Uzi SMGs. Uzis are presumably part of somebody’s arsenal…

          • The SEALs , like other special operations, can get almost anything they want in small batches. Back then they were among the first US users of the Berretta 92. And being hard charging risk takers by nature they loaded them as hot as they could. A Green Beret of my acquaintance stated during the same time period he and his teammates were torturing all their small arms to failure just to see.

      • Actually, no, the Beretta problem was not initially a design problem. The Beretta problems were manufacturing problems, first with heat treating causing slide splits (this is what made the SEALs go from avid supporters to active avoiders of the M92), then, after many years, a running manufacturing change produced locking blocks prone to shearing off their locking “ears.”

        In my last Army unit, we had a lot of that in high round count M9s. The gun would lose one ear, then, if the broken part didn’t lodge in something, seizing the gun, it would get some more rounds off before the surviving ear, too, sheared. When that happened you had a risk of a very energetic blowback and case-head failure.

        The Army and Beretta independently redesigned the locking block — in Beretta’s case, several times — with a view to still making them faster than machining from billet, but accounting for the stress. Supposedly the problem is fixed.

        Beretta also reinforced the slide, this is the Brigadier slide that Wilson Combat uses on their Berettas.

        • Yes, manufacturing/ processing – that was the issue. I recall that organization I was with was also quoting on making M92 slides. They wisely gave up on it.

          • The job of the designer is to come up with something that can be mass produced effectively and efficiently (eg, Messrs Cz, Glock and Ruger and I was going to add Sauer – but thinking of some of their rifle designs I’ll leave them out of the list of manufacturers who have very well process engineered designs*).

            I’ve got to hold my hand up to say that I don’t know what heat treat process gave the problems. The old fashioned quench into a cold media from above the critical temperature is brutal, and is inclined to cause high stresses and even if it doesn’t break or badly distort a part, it is prone to leaving incipient cracks where there are sharp angles or where there are sudden changes in thickness –

            – which is exactly what a P38 and Berretta slide side rails have designed into them.

            The less brutal methods of heat treating, such as “Aust-tempering” and “Mar-tempering” reduce the problems of distortion, cracking and stresses, but do not entirely eliminate them.

            * I’ve a love-hate relationship with several books that I’ve had since my teens or twenties. One of them that’s more on the hate-hate end of that ambivalence is Frank De Haas’ “Bolt action rifles”.
            At the time De Haas was writing, “Investment casting” was about as sexy as putting “leprosy” on a dating profile.
            Poor old De Haas, was very taken with the flat sides of the 2 lug Tikka bolt action’s receiver.
            What he didn’t seem to realise, was the flat sides were put there by the designer, so the investment cast receiver could be finished on a belt sander…

          • Aus-temper or liquid nitride had been common on small gun parts and in some cases “normalizing” is convenient for high alloy steels. It does not obtain top hardness but induces toughness; can be further supplemented with surface pre-stressing as you mentioned earlier.
            As I do not recollect what was a particular process on M92 slide I do remember that normalizing was commonly used on barrels. I have seen barrels swollen in middle like snake-charmer’s whistle, but no rupture occurred.

    • The paradox of a locking system that makes possible an open slide (with its possible benefits of easier ejection and a good view of anything that messes up on feeding).

      It puts stress raising cut outs in the cut down slide rails, exactly where they could least benefit from having cutouts.

      The cures (as opposed to a cheese headed screw to stop the back end of a broken slide coming off the back of the pistol and embedding itself in the firer’s eye socket) seem to consist of an enclosed slide,

      which defeats the reason for the P38 locking system (to allow an open slide), and the increased width it requires when compared to a Browning tilting barrel locking system.

      There is a certain lack of justice, when someone who’s careful not to use SMG loads, gets issued with a pistol that’s just about to let fly with a broken slide.

      • Yeah, isn’t it something…. and problem was pushed back but never really solved.

        The problem is not the frame; I was not accurate with that. Same as with Beretta.

        Btw, I recall from my past experience a word of wisdom from my senior co-worker. He said to me: ” where is no material, there is no stress “. I shall never forget that – it ought to be cast in stone.

        • I enjoyed studying stresses.

          It’s interesting how the successful gun designers intuitively and empirically understood stresses, long before there were any analytical tools available.

          The first of those analytical methods were the Kirsch equations for calculating the stresses around a circular hole in an infinite plate. They were published in 1898, The year that bolt action rifle design reached it’s highest point.

          Griffith’s understanding of stress raisers and of crack initiation and propagation in the 1920s and 30s was controversial for a long time.

          There are a pair of good introductory books for anyone wanting to get started on stresses, structures and engineering materials (you don’t need them Denny, but you’d probably still enjoy them): “structures”, and “the new science of strong materials”, by Gordon. http://www.bookdepository.com/The-New-Science-of-Strong-Materials-J-E-Gordon/9780140135978?ref=grid-view

          When Gordon was studying naval architecture at Glasgow before WWii, the department there still considered Griffiths’ crack theory to be too controversial to teach to their undergrads. I don’t think tghat Griffiths’ bosses in the RAF were any more sympathetic to his work.

          • U are right Keith – stress study is sometimes mindboggling. There are items beyond microstructure, finishes and size of radii.

            For example – you look at typical airframe: it has only measly 1.5 factor of safety. Yet, it is extremely rare that commercial jet goes down based on mechanical failure. Why? Answer is aero-elasticity and fail-safe mode. Without elasticity everything breaks no matter how thick it is (I recall lecture I received once from a metallurgist). This is something which can be and should be explored (in different form) on guns.

      • There is nothing that beats the FN Browning Hi Power for strength, durability and accuracy. I have had the pleasure of firing the Walther, the Beretta 92 and a classic Belgian Browning made during WWII that had matching serial numbers and Waffen/Nazi Eagle acceptance stamps all over the parts. And until that Browning was stolen by a druggie who probably sold it for a dime bag, it was the best pistol I ever owned or even shot.

        • I tend to concur with that view. Browning HP is simply a slick package, albeit bit challenging to make (have seen process sheets). The lockups added by Walther and Beretta are making same thing wider and bit, eh clunky. (This does not mean I do not like it; in a way I do – its mechanical fanciness.)

          In ultimate sense, the CZ75 is probably culmination of that direction, before Petter’s simplified design was implemented (which now everyone copies). One time colorful world of pistols has become kind of grey.

        • In agreement with you both, the Hi Power is a lovely pistol.

          The SIG P210 is a work of art, but far too good for military issue anywhere other than Switzerland.

          I know that Tokarev’s T33 had a (brilliantly designed) detachable lockwork and mag lips, but I’d be very skeptical of how many armies outside Switzerland could be trusted with that detachable lock, and trusted not to open bottles with it or lose it.

          So to my mind, that leaves the Cz 75 as the only serious competitor to the Hi Power (it will severely punish anyone who tries to dis assemble the firing mechanism), and unfortunately, I don’t think any military ever adopted it.

          I have to admire Glock’s simplifications to manufacture and operation, but I strongly suspect that Glock’s feed reliability is bought at the price of reduced case head support.
          That reduced head support is something that can be duplicated on other auto pistols, but generally isn’t, and for very good reasons.

          • The CZ 75 has naturally been adopted by the Czech Army after the fall of communism and phasing out of the CZ 82 in 9x18mm Makarov. Wikipedia also claims that the Chilean Army uses it (or some CZ made derivative) as their main service pistol, but there is no reference and I’m too lazy to check that info right now…

  5. Armee was on the Bsa prototype, one of them, perhaps then suggesting Kraut usage. Not relevant to this pistol. That tin foil model, could be post war… If it isn’t a Volkspistol, it might, might be a prototype model… Might – $50,000? Five at a push – Dubious provenance brought about by plausible hyperbole, I think the author sold it well.

      • Well, I suppose the other ones are significantly different, to the final version, if the final version was more Beretta 92 unless I am mistaken. But… I don’t like it, looks 50’s proto Nazi “what if” model, why not add a laser.

  6. Just a thought on the HP/PP, and the PP dis assembly compared to the Hamada.

    The dropping trigger guard being used to stop the recoiling slide, has potential to use the bow of the trigger guard as a buffer spring, to ease the stopping of the slide.

    It’s interesting that the PP derivative didn’t meet with sufficeint approval for development to continue, but the blowback Campo Giro derived Astra pistols (in 9mm largo and 9mmP) were adopted as a reserve standard by German forces once the war got started.

  7. I think, if the Walther Staff were smart enough to “Eureka” the buffer of VP70 at that ages, the German service side arm would be a blowback pistol. VP70 may be considered as the most succesfull blowback 9mm pistol of all time and its secret lies in the excellent buffer system hidden in a
    location where Walther MP Staff used to mount the trigger and its bar connection. In fact, nobody notices this amazing buffer in the VP70 except having a glance at its sectional drawing. It does not look outside and in field stripped mode, it moves downward with the take down latch and gives
    no clue about its presence. VP70 recoil spring does not need an Hercules paw and its felt recoil seems also in a level as with locked breech samples. Some connect this happening to the deeper rifling it has, which some said, provided for using SMG rounds which rattles and wreckes even the locked breech samples in some longer usage.

    What Walther Designers should do in the MP configuration was; to supply a buffer group into the upper section of trigger guard used to stop and restrict the rearward travel of slide. In a larger volume of course. Only the spring effect of the trigger guard as Keith cleverly stated above, could not supply sufficient buffering in a reasonale distance I think. A look at VP70 buffer sectional draving would give enough idea what it should to be.

    • Problem with the VP70 is THE worst handgun trigger I have ever encountered. Gritty dragging and so stiff it often took both hands. Maybe if I had milked cows by hand like my Dad I would have had the grip strength. I wound up trading it for a .44 magnum revolver. At least that serves it’s purpose .

      • As for berdysh, I think it shares the same 5mm moving unlocked barrel as the full auto pernach pistol, they have the same designer,
        so its not only blowback with no mods.

    • Looking at the diagram I also discovered today this great buffer, which is overlooked in every mention of this pistol functions,
      do you think this buffer has anything relating to the trigger pull?

      Or the trigger pulls around it, not interacting in any way.

      Because my teory which is probably false that maybe bcos of this buffer is the hard trigger pull,if it relates to trigger pull in any way.

      If its not, then lightening the trigger reset spring could also help, along with striker.

  8. Cracked slides were a problem with the later West German P1 with an aluminum frame, which entered production in 1963. That was later by changing to a reinforced thicker slide. “Fixed” means that you can fire standard velocity 124 grain 9x19mm ammunition without problems at your heart’s content. Shooting +P or SMG ammunition is still not recommended. There was also a steel hex pin added to the frame to reduce wear against the locking cam, but that doesn’t actually make the frame stronger. The purpose of the pix was to extend the life of the frame.

    Cracked frames were only ever a major problem with pistols used with German SMG ammunition, which was considerably hotter than standard; +P+ by modern US parlance.

  9. If a lock system has more than one contact surface, it is quiet possible that only one of them will work. The others will wait their turn until the recoil punch peening the working one pressed back to make the lock reached to the closest one. This may be not important for a lock system like Browning’s tilting barrel. But in a combination like two sided mechanism similar to P38, all stress will be charged only onto the one side and if the given tolerances are not sufficient for the other side to wait the working one being peened and give access for the other side begin to work, being of course in connection of material configuration, the slide will have a high possibility of getting a crack at the stress loaded section.

    Synchronized lock surface working for a P38 can be provided too hard and costy. By EDM machining perhaps. Therefore, if the designers are eager to make a usable service pistol in locked breech version, they should think systems to work all at once. The new RussoItalian “Strike One” with a similar construction, possibly will have such an end.

    • EDM is one of the less desirable methods for forming locking recesses, as it leaves a thin skin of damaged metal which contains the flaws for initiating cracks.
      EDM parts that are used in areas where fatigue is likely to be a problem need to have that damaged layer removed.

      I’m guessing that with the clearences that you have got to have for the slide, barrel and locking block to all move freely, there’s probably no way to get them into exactly the same place each time.

      • You may be right. However, we use EDM machining for shaping cutting dies and also for the orientation of the contact surfaces of plastic enjection moulds. In a low current power, those machines provide a smooth, nearly no polishing needed and very hard surfaces. Years ago, a shotgun maker in Japan had been producing doubles with recoil spreading surfaces oriented via EDM machining giving a lifetime warranty against loose.

        • Forgot to mention; At orienting the contact surfaces, including for the possibility of lock areas, The “Electrode” would be the own material of either surfaces. If EDM machining is intented to use in lock surface orienting, blocks and counter recesses should be formed mechanicaly beforehand.

          • the surface damage in EDM can be beneficial, so long as the surface is only being heavily loaded in compression. for example in cutting tools, the EDM surface can result in increased wear resistance.

            The problem with fractures starting and reduced fatigue resistance occurs when the thin skin left by the EDM is subjected to loading in shear or tension. If that type of loading is anticipated, then the skin needs to be removed by another finishing process.

    • Strongarm, you are making right point about P38 with its complex fit/ tolerance requirement. Even if lock piece is done at one tool pass (or EDMed or even extruded) and slide is broached (single tool surface), it is still the lock’s pockets, which have to be machined individually. Result is a stack-up of possible deviations. But then, nothing is completely rigid and some flex takes place and accommodates these factors.

      As you mentioned the “Strike-One”, this is in my mind definitely a culmination of pistol design. Barrel can be held more accurately in place since the swing movement is eliminated.

    • “The new RussoItalian “Strike One” with a similar construction, possibly will have such an end.”
      I assume you mean Strizh automatic pistol – it was one of competitors in Russian 1990s automatic pistol program code-named “Grach” (Rook) which was win by more conventional Yarygin PYa (link-less Browning system)

      • Thanks Daewo, I am aware of all pistols you mentioned. However, surprised as hearing “Stritz ” being a competitor in Russian 1990’s pistol program.

        The importance and uniqueness of VP70, to me, comes from the fact that it being the first in blowback pistols , to control and soften the violent recoil punch of locked breech level rounds through a separate buffer. Others try this through powerfull return springs which rise difficulties in manual slide retracting.

        • “Others try this through powerfull return springs which rise difficulties in manual slide retracting.”
          Dreyse patented …a device … by which the return spring can be disconnected from the breech block, when the breech block is to be opened by hand, so that a return spring can be employed, which would be too powerful to be compressed by hand.
          R.K.Wilson cited in above link said it was faulty design – spring and slide can disengage when they shouldn’t, but I think that on idea level (disconnect slide from spring for manaul repeating) it is worth considering and designed wisely might work as intended.

          • Thanks Daweo. Dreyse’s approach had alsobeen dublicated by Smith Wesson Mod.35 giving nearly the same results.

            The effect of powerfull recoil springs were all intented to carry to the mission of adding potential mass to the breechbolt for postponing its opening when the highest pressure within the barrel was still existing. Whereas, it was understood that their effect through inertia were limited only with their masses. Their initial compression would not give a usable tension to add potential mass to the breechbolt in an event occuring within the time of a few thousands of a second. Their compression would be able in activation at nearly their intended postponing mission ended. But there woulb be usefull in slowing the very fast backward travel of breechbolt after “the bullet in the bore” stage. However, this second stage would not demand a full size buffer filling the whole distance of breechbolt backward travel. A small but powerfull amount being in active at the end of its travel would be sufficient. This was what VP70 designers did. They realised that, nearly every service pistols were able to use the powerfull rounds like 9mm Para. in blowback working system without danger. What important to be done was to dumpen the terrible punch of recoiling breechbolt creating tear and wear to the gun and discomfort to the shooter. They used standart powered recoil spring and breechbolt mass and aso used a very powerfull dumper located at the rearmost travel of breechbolt all creating a very usable service pistol in 9mm Parabellum.

1 Trackback / Pingback

  1. WTW: Eighty Years of Walther P38/P1 Excellence - Breach Bang Clear

Leave a Reply

Your email address will not be published.