Type 95 Military Sword Variations Monograph - Available for Download

[Bruce Pennington]

That’s great, Sam!

[drb 1643]

Nice one Sam! I sure love the NCO’s!

Tom

[Scogg]

Here’s a late war maroon ito wrapped handle with unserialized blade.
 

Just sharing because you don’t see these everyday. A bit pricey for my blood.  

-Sam

 

https://ebay.us/m/IOG2fJ

 

 

[The Blacksmith]

Wow, what a labour of love, and a so needed work. Brilliantly executed. Well done Sam and thank you so much both doing it, and for making it available to us. I will greatly enjoy going through it in depth. I really like these swords.

 

Thank you.

 

Russ

[Scogg]

My pleasure @The Blacksmith Russ! Thank you very much for the kind words. I still consider it a "work in progress".

For what it's worth, and for those interested. Currently I am collecting each variation so that I can take more standardized "professional" looking photographs. 

Hope to update the document to the "Second Edition" around its 1 year anniversary. 
-Sam 

Here's a little sneak peak of the "work in progress". Had to shrink file size for this post, and I am still fine tuning stuff 

[The Blacksmith]

@Scogg The sword on the right, is that a field repair, I mean the screw in the mekugi?

[Scogg]

2 hours ago, The Blacksmith said:

@Scogg The sword on the right, is that a field repair, I mean the screw in the mekugi?

Good question. The sword on the right represents what Jim Dawson’s cyclopedia refers to as a “2a pattern”. These swords have thick seppa that have a cutout for the retention clip, and also the tsuba is more thin than other aluminum hilt steel tsuba Variation4 swords. 

The missing mekugi is unrelated, and was probably lost or damaged when somebody tried to disassemble it. I put the bamboo mekugi in there for security. 
 

It is currently my interpretation that these thick seppa and thin tsuba “Dawsons 2a” swords are indeed factory repairs or field repairs. You find this unusual configuration scattered amongst all top-latch variations and on both Tokyo and Nagoya stamped blades.

My photos are an attempt to show those differences next to a “more standard” and what I believe to be a factory original variation 4 sword.

 

Other interpretations may vary…

-Sam

 

 
 

 

[Scogg]

@The Blacksmith If you mean the first photos, where that screw mekugi punctures the ito. That’s a “Variation #2” sword with Suya Shoten ferrule stamp. 
 

Both Suya and Iijima had the mekugi puncture the ito like that for a very narrow serial number range right after the copper hilt Variation 1 swords. Gifu had a slightly different mekugi situation where the Ito kind of bends around the mekugi. 

June 24th 1939 the hilt pattern was changed so the mekugi would go through the diamonds and no longer interfere with the ito. They did this by making the wrap pattern symmetrical whereas before it was asymmetrical. 
 

Copper hilt swords only have the grommet and no mekugi unless they are field repairs. My example up there looks like a copper, but really it’s aluminum #7249

 

Best, 

-Sam 

 

 

[The Blacksmith]

Ok, thank you for that Sam, that is much appreciated.

[Scogg]

Here is an example currently on eBay that has had the serial numbers removed. Presumably someone scratched them out thinking they were a bad thing.. 
 

A shame to see. I can’t even catalog examples like this.

-Sam 

 

 

[Bruce Pennington]

31 minutes ago, Scogg said:

I can’t even catalog examples like this.

Heck, Sam, plenty to catalog - all black paint job, Nagoya with Seki stamp - Ha!  Sorry, that's what I watch for!

[Scogg]

Haha, true Bruce. I do attempt to list them, but they frustratingly have to reside off of the “master list” because I have it in serial number order. I have a few now that have either been obscured like this, or had the numbers polished away 
 

This one will have to be: 1 _ _ 4 _ _  

 

 

[John C]

At least they left their initials carved into the tsuba.

 

John C.

[john42768]

Wonderful resource. Thank you.

 

 

John S.

[Scogg]

I'd like to talk about the presumed total amount of originally manufactured, seriliazed on the blade, Type 95 military swords. This data is from my cataloging efforts, and should not be considered all encompassing. (~ = approximate).

Lets break this down simply into just the just the hilt materials. Some overlapping does occur. 

Observed copper hilt swords range:
4 to 6561

Observed aluminum hilt sword range:
6756 to 162718. 
and
200797 to 202802.

Observed wood hilt sword range:

201750 to 218821

and
300060 to 301814

Now lets digest this data into a total amount presumed manufactured.

(~6561 copper) + (~157967 aluminum) + (~18825 wood hilt) = 183,353 swords

It should be noted, that this number is not precise because it's loosely based off of what is observed in the market. Some margin of error is entirely expected due to overlapping, and the simple fact that I havent seen everything.

 

My point in bringing this up, is that I believe based on my limited observations, that these swords are less common than many people believe. 
Hope someone finds this interesting 
-Sam
 

[Bruce Pennington]

It is quite interesting, Sam.  Still trying to wrap my head around it in light of Fuller's numbers.

He stated that according to American records 661,621 swords were collected at the end of the war.  (372,609 were kept and distributed; 289,012 were destroyed)  That number included all the private swords confiscated, too.  It would not account for all the swords lost to the Russians, and lost on the battlefield.  His estimate for total swords made during the war is 2 million.

 

I don't know the officer/NCO ratio, but it's likely that there were 2 to 3 times more NCO's than officer in the Japanese military.  Just using the 661,000 number, and let's presume confiscated swords were 100,000, then of the 500,000 military, you would expect 2/3 to be NCO swords = 330,000 NCO swords.  

 

Now, Fuller pointed out that many officers owned more than 1 sword, so that could account for the larger number of officer swords.  But even if flipped, 300,000 officer; 200,000 NCO, that would require that 100% of all manufactured NCO blades were accounted for at the post-war count, which seems very unlikely.

 

I don't see anything wrong with your math.  But it just seems to me, in light of Fuller's numbers, that there should be more.  

 

Open to hearing other ideas!!!

[Scogg]

Interesting numbers Bruce. Definitely hard to quantify.

Could it be because its wear was restricted to senior NCO's, primarily Sergeant Majors and Warrant Officers. Corporals and Sergeants generally did not carry the Type 95 unless assigned to regimental headquarters or serving in mounted branches such as Cavalry, Transport, or Military Police?

See Nick Komya's comment on post #10, for my source:
https://www.warrelics.eu/forum/Japanese-militaria/short-development-history-type-95-gunto-676112/

This is one reason why I do not like to call these Type 95 swords "NCO swords" anymore.

You're much more privy to the internal Japanese Military and rakings stuff than I am. In your opinion would this restriction about senior NCO's help make a little more sense of Fullers numbers?
Best,
-Sam

[The Blacksmith]

Sam @Scogg  This is a fantastic thread. So much to learn from it, so a big, big thanks to all who have contributed! 

 

Just a thought, the sword with the defaced numbers on the blade, could that have been done by the soldier prior to surrendering it, so that there was no way to trace it back to him? Officers often did this with their mon on their swords prior to surrendering them, so I wondered if this might be something similar?

I do not know if there was ever a list of who had what numbered sword, but it was just a thought.

[Conway]

They made a few 100k Type 32 as well. And obviously not all NCOs were issued swords too depending on their military roles. And then account for additional unknown numbers of enlisted and NCOs who probably had personally-procured swords. I don’t think Sam’s estimates of Type 95 production need to line up with estimated NCO billets. 

[Scogg]

That stands to reason Conway. I'd be curious if your cataloging efforts agree (generally) with my number estimates. 

Russ, anything is possible! Although, I have only encountered a few Type 95 with numbers etched out like that. Coincidentally, that one was actually local to me, and a friend almost purchased it. Your theory could be correct, or someone could have done it more recently because they believed numebrs on Japanese swords to be a bad thing or fake identifier.

All in all, a huge part of the reason I dived deep into this type of sword, is because I believe there is a lot more to them that the general militaria collecting circles understand. I think they are more rare and interesting than many people give them credit for.

I am obiously a bit biased toward them, so keep that in mind 
Best,
-Sam

[Bruce Pennington]

Good point on the added numbers of 32s, Conway!  That would greatly increase the number of NCO gunto, and we know from photos that many 32s were part of the post-war confiscation.

 

Russ, I also agree that your idea may be true.  Another option to consider is the provision, during shortages, for officers to purchase Type 95s.  I think it's possible an officer could have bought this and filed off the number.

 

Sam,  After reading Nick's post stating that "senior NCOs and Warrant officers" were authorized, it would certainly reduce the number of NCOs that needed swords.  I suspect that the reality was a bit different than the regulatory picture.  Just from photographs alone, my impression is that more guys were carrying these than just Sergeant Majors and Warrant officers.  But your numbers certainly brings that reality back closer to Nick's observation than my idea that all NCOs carried them.

 

Thanks for bringing this up!

[Conway]

Making sense of all these numbers sure does make your head spin! But I don’t think we should assume issuance on a 1:1 ratio regardless of the rank approved to carry them. Also I re-read the posts above, and I assume when Fuller says most officers had more than one sword he is implying they had one for dress (ex. Type 19) and one for the field…

 

Back to the Type 95s - My cataloguing hasn’t gone as in-depth as Sam’s, but I think those are good estimates on total production. 

 

 

[Bruce Pennington]

Sam,

I wish I noted the source of this chart.  I came across it this morning in my files, while looking for production numbers.  Puts your estimate right in the same ballpark:

 

[Scogg]

Thank you Bruce, it is an interesting chart. 

It's my understanding that these numbers were put together as estimations by the United States Bombing Survey United States Strategic Bombing Survey (USSBS).

The figures of total amount of swords manufactured is close to accurate, but the dates of said manufacturing leaves out a lot of nuance. Especially when we account for co-production across multiple subcontractors and arsenals. 
Best,
-Sam

Edited July 4 by Scogg
Correction

[Bruce Pennington]

Yes, I thought it must have been estimates, simply by seeing the numbers so cleanly rounded to 100's.  I would say your estimate is much closer.  It does, though, support your estimate, in that both numbers are below 200,000.

[Bruce Pennington]

An update from Edokko at Gunboards:

"When I did a quick check through era arsenal documents on the fiscal year production plan for 1945, the monthly Type 95 sword planned numbers from April to August (beyond September is meaningless so I did not tally it up) were as follows and significantly lower than the numbers from your reference source, moreover these are production plan numbers which were obviously impacted by the bombing campaign with actual production numbers being much lower than the planned numbers.

 

Suya / Apr : 50, May : 200, Jun : 200, Jul : 300, Aug : 350
Kobe / Apr : 400, May : 400, Jun : 500, Jul : 600, Aug : 600
Iijma / Apr : 0, May : 0, Jun : 50, Jul : 100, Aug : 250
Jinsen / Apr : 100, May : 400, Jun : 1,000, Jul : 1,500, Aug : 1,500
Nagoya contract / Apr : 1,500, May : 2,000, Jun : 2,000, Jul : 2,500, Aug : 2,500

 

Total production plan / Apr : 2,050, May : 3,000, Jun : 3,750, Jul : 5,000, Aug : 5,200 "

[Scogg]

Very interesting Bruce, especially those Jinsen and Nagoya numbers. Hard to make sense of it, and I imagine much has to do with "planned production versus actual production".

My wood handle Variation#8 (Nagoya) range is:  203,306 - 218,821, suggesting around ~15,500 made  

My Variation#9 (Jinsen) range is: 300,060 - 301,814, suggesting around ~2,000 made. 

-----------

As far as Tokyo First Arsenal and surrounding subcontractors, by April 1945 all production had ended due to bombing. See Nick Komiya's write up here, including some really interesting wartime documents:

https://www.warrelics.eu/forum/Japanese-militaria/short-development-history-type-95-gunto-676112-9/?__cf_chl_f_tk=sIiuCzCONG3oxtas4qMmPBaCxNrota_w2wMtEEdUEAg-1783175461-1.0.1.1-Aw9oLoAARry4NkDDMTNZHJOY8QxdRMwKiUWxZPweuG4

...

"2. 1945 April Production Report (Remarks column)
"Recent bombing raids have destroyed production facilities at all suppliers (Suya, Kobe and Iida). Though progress is being made in relocating their production elsewhere, resumption of production is anticipated only from July onwards. Officer gunto production also lost all blade workshops, both within the arsenal and at private companies. Through desperate means, prospects for a supply of approximately 700 blades have been secured, but as facilities for exterior components also had been decimated, none can be assembled into complete swords. Regarding these exterior parts, we are currently calling on Association members such as Iida, Hinodeya and Sugito to work together in establishing new production facilities in Kofu City in Yamanashi Prefecture."

3. 1945 June Production Report Remarks
"All suppliers are still in the process of evacuation and have not been able yet to resume production. Efforts are being made to merge and reorganize suppliers to secure a steady supply.""

... 
"As of April 1945, all Type 95 suppliers as well as the full size production line planned at the Tokyo Arsenal were bombed out of business, losing most of their production equipment. Bombing was particularly heavy in April 1945 as shown below.
November 1944: 4 Raids
December 1944: 12 Raids
January 1945: 7 Raids
February 1945: 10 Raids
March 1945: 8 Raids
April 1945: 13 Raids
May-August 1945: 52 Raids
TOTAL: 106 bombing Raids targeted Tokyo”

[Bruce Pennington]

42 minutes ago, Scogg said:

My wood handle Variation#8 (Nagoya) range is:  203,306 - 218,821, suggesting around ~15,500 made  

My Variation#9 (Jinsen) range is: 300,060 - 301,814, suggesting around ~2,000 made. 

Plan vs Actual - yes, always different

 

Just looking at Jinsen + Nagoya, the plan was for 15,000 vs you actual of 17,500 -- so, believable!  Also, I don't know how many of your numbers were strictly '45.  Could they have been spread over '44 & '45?  If so, you actual for just '45 would be less.

As for Tokyo, I think their plan for 9,000 swords got interfered with by the Allies.

[Scogg]

1 hour ago, Bruce Pennington said:

Also, I don't know how many of your numbers were strictly '45.  Could they have been spread over '44 & '45?  If so, you actual for just '45 would be less.

As for Tokyo, I think their plan for 9,000 swords got interfered with by the Allies.

Yes, it's a good point. It's my beleif and current understanding that the Nagoya wood handle with fuller (Variation#7) would date to around May 1944, and the wood handle without fuller (Variation#8) would have began around June or July 1944. How far into 1945 production continued in or around Nagoya army arsenal is not clear to me. See Kiipu's comment from Nov 27th here:
 https://www.militaria.co.za/nmb/topic/54210-type-95-military-sword-variations/#comment-567821

The serial number block of 300k+ was allocated to Jinsen in May 1945, so these sword would have been made on or after that date. See post #126 here:
https://www.warrelics.eu/forum/Japanese-militaria/short-development-history-type-95-gunto-676112-13/

One thing that has cause me great confusion when interpreting documents and dates, not that it applies here necessarily; is the usage of "fiscal year" instead of calendar year. Fiscal year being April to March.  

All the best,
-Sam

 

[Kiipu]

On 7/3/2026 at 7:13 AM, Scogg said:

It's my understanding that these numbers were put together as estimations by the United States Bombing Survey.

 

United States Strategic Bombing Survey (USSBS).