RIA: Japanese ZH-29 Copy by Tokyo Gas & Electric

The Japanese military experimented with self-loading rifle designs through the 1930s, and had 4 major rifles in testing during that period. One was a new design by Kijiro Nambu, one was a Pedersen copy made by the Tokyo Army Arsenal, one was a gas operated toggle locking rifle by the Nippon Special Steel company, and the fourth was this, a ZH-29 copy made by Tokyo Gas & Electric. TG&E was a major industrial concern that made all manner of products, and they chose to copy the ZH-29 for Army rifle trials. However, it appears that while their manufacturing quality was quite good, they lacked the firearms design expertise on staff to fix the problems the rifle was found to have.

Specifically, their ZH-29 suffered from substantially inferior accuracy. It seems that no significant changes were made between the first and second major trials (1932 and 1935), and when the accuracy problems appeared unchanged in 1935, they were dropped from competition. Only a handful of these rifles were ever made, between 10 and 25.

There are a number of differences between the TG&E rifle and the original Czech ZH-29. The most significant of these is a separate non-reciprocating bolt handle on the TG&E rifle, where the Czechs fixed the handle directly to the bolt carrier body. The Japanese rifle was chambered in 6.5x50SR, of course, and used a new magazine not compatible with the Czech type. The trigger group was also redesigned somewhat, although not in a fundamental way.


  1. The lever at the front of the barrel, with no known function is possibly a gas cut-off for rifle grenades? I looked at it a couple of times, and can’t quite see.

    • It is tough to see. The axis pin for the lever seems to go all the way through the gas block, so it could easily act on the gas system. I don’t see anything that locks the adjustable gas plug in place, so that could be this levers purpose?

  2. I asked my wife to translate the writing while I was watching this (Chinese Mandarin but the writing is the same) and the safety is indeed “an” which is “safe”. The writing at the front (near the gas piston) is “kong da” which means “big air” or, I’m guessing (imperfect translation here) high pressure or something along those lines.
    It’s a pretty cool piece of history either way.

    • It is great plus to be able to obtain help in exotic language. I recall from hearing one Chinese man that reading of Japanese was not difficult for him (the truth is that he studied in Japan). What he actually meant was that symbols were rather similar and he thought he would be able to do without this added knowledge.

  3. My idea on the mysterious lever is that it is or was the retainer for the gas regulator plug. There seems to be no other retaining part for this and it obviously needs one. Try wiggling it right out.
    The Bren gun, being derived from a product from the same Brno stable, also has a tilting bolt lock, but the bolt tilts up at the rear. It is renown for its accuracy in single shot mode, in spite of firing from an open bolt. We used to run little informal competitions (7.62 Nato version, the LMG) to see how many shots you could place in the middle of a figure 11 target at 100 metres, firing prone, but there were those that tried it standing up! The middle of a fig. 11 is I think 2″ x 4″. The vibrations would be completely different, the LMG has a much heavier construction, the barrel is quite massive by comparison and the receiver very rigid, but I think the crux of the matter is lateral asymmetry rather than vertical. There is always more vertical tolerance to inaccuracies to take into account variations in propellant dosing. Legend has it the original vz. 26 (or was it 33?) was too accurate so a bit of scatter had to be designed into the .303 Mk1.

    • “Bren gun(…)Legend has it the original vz. 26 (or was it 33?) was too accurate so a bit of scatter had to be designed into the .303 Mk1.”
      I don’t think so, however I heard that a LOT of work was required to make this design work with rimmed .303 cartridge. It is true or not?

      • Original “bren” guns starting from Praga were basically overgrown rifles, that’s how I see them. So I would not be surprised if their accuracy was above par.

        Enfield’s version is total redesign in almost all aspects part of locking concept. There was attempt toward end of the war by J. Inglis in Toronto to offer belt-fed version, which not surprisingly did not take off (still just overgrown rifle, kind like BAR). But, all the credit goes to this Canadian company for lots of improvement, mainly in area pro manufacturing process.

        • Well, Denny, I fully agree that the Czechoslovak LMGs were basically overgrown automatic rifles, but so were also most of its contemporaries. It is interesting to read about the attempt by Inglis to produce a belf-fed Bren. I remember reading something about it many years ago, but can’t remember where; possibly in some Czech book or maybe in Dugelby’s The Bren Gun Saga (which, by the way, is out-of-print atm; sadly, I lost my copy of the 1999 reprint during a house move).

    • Agreed. Too bad the manufacturers didn’t hire a real gunsmith to get accuracy up to spec. Just think about it. If Japanese troops gotten themselves Holek action rifles and a better combat doctrine, they might have made the war years miserable for the US Marines in nighttime jungle skirmishes, where planes generally aren’t much help… Or am I wrong?

        • Well, using artillery fire to attack a small enemy party is overkill. And doing such if the enemy is barely 50 meters away or even 50 feet away risks blowing up friendly troops! I didn’t say rifles decided battles but in very small engagements small arms tech makes quite the difference. Did I mess up?

          • Using artillery in fluid jungle fighting requires radios, which the Japanese did not have and even the Allies often got only later in the war. The support weapons of choice were mortars and rifle grenades. 81mm and 60mm mortars can be carried by men when disassembled. Smaller 50mm mortars like the Japanese Type 89 “knee mortar” could also be very useful. The main problem with mortars was carrying the ammunition.

            A self-loading rifle certainly would have helped the Japanese in the jungle fighting, especially if a significant percentage of riflemen would have had one. SMGs would have been even more useful, though, like Daweo wrote.

          • Well, FWIW, LTC John George, a veteran of Guadalcanal and Merrills Marauders, has a line in his book “Shots Fired In Anger” to the effect that in his experience victory always seemed to go to the side with semi-auto rifles.

            Now, as to “Using artillery fire to attack a small enemy party is overkill.”. As a combat infantryman I would have to say; “Yes it is overkill, now SEND MORE!” Trust me, the infantry would be perfectly happy for the artillery (Ours please.) to blast every enemy soldier into little tiny bits leaving us with the job of counting the feet and dividing by two. Yep, great idea. It hardly ever seems to work out that way though.

            Wafa Wafa, Wasara Wasara.

          • I think Daweo got a point: a cheap, easy to manufacture SMG would have been quite useful for jungle fighting at close range (and not only, of course). The Japanese had plenty of LMGs though.
            As for the comment by W. Fleetwood, I heard quite a lot of similar opinions from servicemen of different countries regarding artillery support for the infantry!

          • W. Fleetwood: artillery was of course used to good effect by US forces in Vietnam, although sometimes the NVA/VC forces managed to escape the worst effects by disengaging quickly and moving to another location or by Escaping to tunnels. They still suffered disproportionate losses in most encounters and that was mainly because of the US superiority in support arms.

          • My late father was an artilleryman in New Guinea and Bougainville in the fight against the Japanese, using 25-Pounders and Short 25-Pounders (pack howitzer). He told me that they were subjected to ‘banzai’ charges at close range in jungle terrain, and used their guns by aiming directly at Japanese soldiers as they attacked, with devastating effect. Similar tactics were used by US and Australian/NZ artillery in Viet Nam, especially using ‘beehive’ rounds.
            Also, the effectiveness of the Thompson submachine gun against the Japanese in New Guinea is considered to be a major factor in the eventual success of the Kokoda campaign, in particular.

  4. Indeed those inscription on the gas plug read “Empty”, “Big” and “Small”. I suspect they denote gas volume adjustment.

  5. Daweo, I don’t think you are right there. The original Bren was very accurate, the high centre of gravity would preclude muzzle climb. There are plenty of written references to the accuracy of the Bren’s parent design, and the need to design some scatter into it. It is a slow firing weapon, so spread of shot would not come as “naturally” as with the GPMG or the MG3 (in fact slow enough to allow “single taps” quite easily, which would enfuriate the range officer no end!).
    Regarding difficulties with the rimmed ammunition, I think they are grossly overstated. Vickers made the Vickers-Berthier work fine with .303, and so did the Bren. You had to know what you were doing when loading a magazine that’s all, if you were exhausted, with cold wet hands doing this at night the potential for a crossed rim was certainly there, it boils down to training. I have never come across a report of this problem. Can you provide any references?
    The business with the 28-29 capacity instead of 30; I’ve talked about this before, I think the magazine was designed for 30, using a Swedish steel spring, like the VGO magazine was designed for 100. Come the war, they had to make do with inferior steel and to avoid springs fading out on you the last 2 rounds were not loaded (4 in the VGO). British ores need a lot of cleaning up in Bessemer converters to make them good as a basis for quality steels, on the other hand they often have high manganese contents such as the Hadfield steel used for Mk 1 steel helmets.

    • “You had to know what you were doing when loading a magazine that’s all, if you were exhausted, with cold wet hands doing this at night the potential for a crossed rim was certainly there, it boils down to training.”
      Is misunderstanding, I think about work needed to redesign weapon from rimless cartridge to rimmed cartridge, not work from point-of-view of loader.

  6. With due respect to creators of this interpretation of Holek’s rifle it impresses me on two counts.

    First its incredibly dainty and impractical controls (safety and sight), fragility and general awkwardness. Second is finesse of machining and detail work. Generally, this does not look to me like useable military rifle and in that sense was probably viewed original Holek’s creation as well.

    This side tilt action is really quirky idea, but on other weapon, like Goryunov machine gun it worked well.

      • This spawn question: who is responsible for this solution? It was done from factory initiative or it was requested by army?

        • Answer we will get is by anticipation – we do not know.
          But from looks of it, it was the factory; responsible project leader to be precise. But, military was free to throw it back on their head which is what they did.

  7. I don’t think it’s that difficult to redesign a weapon from rimless to rimmed ammunition, as long as it’s magazine fed and the bolt pushes the round from the back into the chamber. You have to make sure it clears the lips of the magazine after a good part of the round is in the chamber, otherwise you get a bump as the rim jumps out of the magazine, but much the same applies to rimless ammo. Extraction and ejection are not an issue, and headspacing is much easier on a rim than on a bottleneck. Belt feeding is another problem altogether, you are essentially denied the much simpler and neater method of pushing the round out of the belt into the chamber, and have to resort to a claw affair like in the Vickers gun to pull the round out of the belt backwards and then load it. Funnily The Browning 30-06 had a similar mechanism because of the cloth belts, which made it much more amenable to adaptation to .303.
    Accommodating the rounds in the magazine requires some careful geometry to stagger the rims keeping the front ends in pressure contact, hence the curved mag.

  8. I have just remembered, the vz.26 etc were designed to run on 8 mm IS, and .303 was cordite filled. This must have been the real headache as the combustion characteristics are completely different, the gas port was brought back substantially, amongs other changes. When the BREN was re-engineered in Canada for 8 mm IS, for the Nationalist Chinese, the gas port location was maintained, but other changes were made.
    The Japanese 7.7 mm was a copy of .303 but filled with something else (anyone know what?) and the rimmed variety seemed to work well with the Lewis gun copy they used on aircraft.
    Obviously a gun designer’s life is not easy, unless he gets a blank sheet of paper, very wide budget, lots of time and told to “get on with it”!

  9. Denny, the video you suggest is really good, and nostalgic. Very down-to-earth comparisons, the small aperture sight on the Bren (II, III and IV) was useful when you were engaging at long ranges, but up to 200 m or so there should be a battle sight. I cannot remember if the sights were any better on the LMG, only that the front post (we called it a blade sight) was quite narrow. Ajustment is by a tool with 2 screws so that you can push precisely what you want to zero the barrel.
    Please note that at no time is the author of that video griping about rimmed ammunition, guess he knows how to load a magazine…
    Also see how superbly stable both weapons are when fired from the hip, it’s that heavy magazine on top, and their 9.5 Kg or so weight!

Leave a Reply

Your email address will not be published.