Just acquired 1898 Krag. with all the research I was able to find online, it appears to be a 1898 carbine. correct 22" barrel, correct unmodified stock but one thing odd. It does not have a serial # stamped on receiver next to Model 1898 US Springfield Armory. Any Ideas?

Additional Pictures can be seen at:

  (You need to Login

---

We all like to see pictures of nice looking Krag carbines. The regulars are here are good at solving mysteries, so please get us some photos.

I had to send an email in the middle of typing the short reply and your photo was there when I finished.
I remember someone else had one with no serial number, forgot the details of why.

'Marksc' - Welcome to the KCA Forum.

I think the logical answer is that someone (God knows why?) removed the serial number from your Krag's receiver. It should be where there is a vague 'cross like' shadow area on your receiver.

Your nice carbine stock is the later type (1899) 'long forearm' stock. These often were used as replacements, during refurbishment or repair of model 1898 carbines, to make them identical to the model 1899 carbine.

Is there a cartouche mark or 'proof P' stamp on your stock?

Is your model 1896 rear-sight a rifle or carbine sight?

Your side-plate screw appears overly long and your magazine cut-off lever is missing. This suggests to me that someone had a Redfield 'no-drill' peep-sight (102-K or 70-K) on this carbine for 'Hunting', at some time.

---

Is that a M1892 sight?

It looks like a model 1896 rifle sight without its 'slide'.

---

The receiver has no distortion from round finish or flat area from any numbers being ground off. A perfect finish. The sight is a 1986 with a C stamped on the side. The cutout I removed because it felt like it was sticking. Appears to be a little bent. Would like to replace. There is no cartouche or P on stock. Just behind the trigger guard is a stamped 50. 

"I think the logical answer is that someone (God knows why?) removed the serial number from your Krag's receiver. It should be where there is a vague 'cross like' shadow area on your receiver." Quoting BR
If someone ground and buffed the number out, there would be a major depression. A strait edge/ruler would show up such a depressed area. Lay the ruler length wise above and below the stamped area and it should show up easily.

Model 1896 carbine rear-sights are quite valuable. (The carbine 'slide' is different than the rifle slide).

Carbine sights have been faked.

There is a good feature (on sights and model '96 fakes) on the KCA Website main page, accessed by using the 'photos' tab.

The shaft of your magazine cut-off lever can be easily straightened. Use a copper, brass or plastic mallet and pound it 'true' on a hardwood surface.

---

Just placed a engineers straight edge on receiver from 8 oclock to 12 oclock position. Dead straight. Never any numbers stamped on this one.

That's it, just missing slide. Thank You for all the input. New to this site.

Welcome Mark!

In "The Krag Rifle Story", 2nd Edition, Chapter 19: Collectors Notes there's a paragraph that states:

"Prototype Model 1898 rifles were marked "Model 1897", and receivers of rifles chambered for experimental cannelured cartridges were marked "Model 1899".  These rifles have no serial numbers".

Maybe just adding to the mystery since we're talking rifles and your carbine is not marked Model 1897 or 1899.

Wow, There's a mind bender. Does a P stamped on barrel 1/2" up from receiver help with anything. There is a P also on the front ring of receiver where barrel connects to receiver.
Great Info.

The "P" on the barrel and the "P" on the front of the receiver ring are the standard Springfield Armory marks, indicating the parts had been tested with Proof Loads.

Receivers and barrels were proof tested before being assembled into arms. Barrels were proofed before being rifled.

Receivers received stamped markings and serial numbers after heat treat.

Numbered/completed receivers went into 'parts' bins', and were pulled out (kind of) randomly for assembly, hence the difficulty in exactly dating Krags by serial number.

Completed Krags underwent a final proof procedure and received the 'circle P' stamp on the stock. (Description attached).

The U.S. government paid a royalty on each Krag-Jorgensen that Springfield Armory manufactured.

I do not know why your receiver appears 'unnumbered'.

FWIW - Possibly, the number was skillfully removed (or filled in with weld) and re-contoured by someone?

Maybe, a Springfield Armory worker stole a receiver, prior to numbering, and took it home in his lunch-pail to build his own Krag?

It was a 'Special-Ops Krag', used in an undocumented secret mission? (Joking).

Several S.A. workmen and inspectors screwed up on a Monday morning and stiffed the Norwegians?

(I am skeptical of the prototype thing).

It boggles the imagination and probably makes you a Felon! (Just kidding).

---

I am going to take some better pics and post. It definitely has not been altered or welded and resurfaced. I am 100 percent sure now that it has been broke down and examined it is a carbine that has never had a serial number stamped on it. I am still open to suggestions and appreciate all the knowledge everyone is giving. Thank you everybody. 

Here is a possible thought:

Maybe the OP's receiver was rejected, for being out of specifications or having a defect, before it received a serial number?

Surplus dealers like, Bannerman's and Stokes Kirk, bought up surplus parts, unfinished parts, and Springfield Armory 'junk', in government auctions.

These two companies built a lot of Krag 'parts guns' in the 1920s and 1930s that often perplex collectors.

There is an area of the OP's receiver that has not looked right to me from the start.

Maybe it is just an optical illusion exaggerated by the removed cut-off and missing wood, or maybe, it is a machining mistake that caused rejection?

I have circled the receiver area that looks odd and kind of gouged to me. I have also included a couple of pictures of that area from a model 1898 and a model 1899 action.

---

Marksc - Could you post a clear, up close, picture of your front-sight attachment and muzzle crown?

(This is always a valuable clue on carbine barrels).

Attached is a picture of two Stokes Kirk catalog pages. Some "carbines" have been made privately, outside Springfield Armory, (in the early 1900s and since), from surplus parts.

A carbine stock with no cartouche or 'circle P' stamp was likely put on outside of Springfield Armory.

---

Butlers Rangers, you might be onto something there. Another possibility is that it was not defective but some experiment was started on that particular receiver that did not work out.
The area circled looks like it was milled out on purpose, maybe a sight was planned? It would be interesting to see it with the stock removed. If the "milled" flat area has a mounting hole and the whole area is hard, then it was done before heat treat at the arsenal and it is not some bubba's craziness.

IIRC - The U.S. Krag receiver went through 120 machine operations. To speed up the process, receivers were held in fixtures in pairs, so that machining operations were performed on two receivers simultaneously.

That odd 'angled' cut on the OP"s receiver looks like it could have been a mistake caused by something getting out of whack during a shaping operation???

IMHO - If something was cosmetically or dimensionally wrong with a part, during the course of Krag production, it would have been noticed with all the inspections and gauging that went on. Likely, the part would have been 'tossed into a scrap bin'. A receiver was likely "Right" before giving it a serial number and facing the prospect of paying royalties.

Some experimental Krags, like Cole and Metford rifling tests and Parkhurst Device trials, were documented by serial number.

Serial numbers are a handy thing to have on prototype, test, and trial rifles for identification and comparison purposes.

(I'm not a Machinist, but, my father sure was! I'd like to see more details on that Krag. It's an interesting puzzle).

The picture of receiver shows like there is milling lines. There is no machining on it. when I got it , it was laying on a wash cloth upside down half stuck to it. when I pulled it off it left those lines from rag. not wanting to do anything I just rub it down with Hoppes oil.   

New Picture of receiver showing non touched finish

---

Have someone coming over tonight. He is bringing over a 1898 receiver in excellent condition and compare all measurements with a micrometer and engineers straight edge tools. I am still hold true that it is a receiver that has never been touched and never had serial numbers stamped into it. There is absolutely no flat spots on it. 

Wish I had a receiver in hand to compare with but something is really different there. My first impression is it is not an accident, could be wrong here. Another photo with some light in those dark areas would be helpful. I am thinking I see half of a counter bore that should not be there.
Multiple cutters were mounted on the milling machines that did that work. To mill that step would have taken a lulu of a mistake if not done on purpose.
Assuming this receiver got the proper heat treatment, it should be safe to fire if everything else is normal. The sharp flat bottomed cut will weaken things but it is in a NON critical part of the receiver.
Hope a better photo will help figure out the puzzle.
Figured out what that half counter bore is, that is the pocket the trigger assembly fits in. The darkness in that area makes it hard to figure out.

Butlers Rangers,
He is a pic of the area you circled. You can see I have not touched any cleaning yet.

---

Thanks for the latest photo 'Marksc'.

It looks to me like Springfield Armory workers missed some machine operations on your receiver, before it was heat-treated and Proofed. This could not be corrected after heat-treat.

I've never seen an unfinished receiver before. Cool!

That probably explains the absence of a serial number. It was probably 'scrapped', but, someone later made a "carbine" out of parts.

Attached - Contrasting photos of Mark's receiver and a model 1899 receiver, showing considerable difference in 'trigger pivot' area. The receivers should look identical.

---

Almost like we are looking at 2 different receivers. The last photo looks like they forgot an operation. Going back and looking at BR's photo the milling operation around the trigger recess was forgotten. That may make what appears to be a step. I am not familiar with all the changes to the receivers between 1892 and 1898. But that "lost operation" would have made fitting a rifle stock a custom operation. Maybe the arsenal did not want to mess with it but a surplus seller might do the work to salvage a receiver.

Now see the difference you are talking about. Another Picture trying to get the angle you used on comparison.

---

Looking at stock it is very clear that part of receiver leaves a very thin wall of wood (approx. 1/8") that possible could be the reason for changes. Does your original stock look like this?

---

Don't know if I should broach this subject or not... but Mark stated that it was definitely a carbine.

Now here are the thought running around in my head:
The receiver is marked 1898.  True 1898 carbines are rare, and command a mighty high dollar figure.  There are many fakes running around, and the serial number is the only true way to discern a REAL 1898 carbine from a fake or simulated one.  Being that this has no serial number makes me suspicious that it is not a true 1898 carbine.  I'm leaning more towards a botch or unfinished receiver that was built into a Bannerman or Stokes Kirk "Parts" gun.  Now it may or may not have a true carbine barrel (let's face it, anyone could buy a carbine barrel and install on a rifle receiver)... but that does not make it a true carbine.

It is... however, one heck of a mystery!!!  Either way, Krags are meant to be enjoyed!  Yours just has a little additional story to tell!

Just out of curiosity, have you ever fired it? 

FWIW, a serial number-less Krag showed up here before, and it turned out that the number WAS machined off.

Springfield missed a spot!

The area outlined in red should have been removed by machining prior to heat-treat.

I wonder how often such mistakes happened?

---

One thing about it guys, you know your stuff. No, I have not fired it. Just got it and wanted to break it down and clean and inspect it good before that. I appreciate all your input and Thank You.

'Marksc' - Someone gouged a lot of wood from your stock in order to 'fit' the unfinished 'left rear-corner' area of your receiver.

BTW - Model 1896 Krag actions were machined pretty much the same as model 1898 actions in that area. In fact a model 1898 action will fit into a model 1896 stock. (However, Springfield did not use this arrangement). A model 1896 Krag will not fit into a model 1898 stock, without alteration, because the receiver is different at the bolt handle 'recess'.

Attached is a photo of a model 1898 stock showing the appearance of the normal receiver inletting.

---

Any opinions as to how this might affect the legality of owning or selling the rifle in question?  If it is unnumbered it is impossible to say whether or not federal law classifies it as an antique.  If it had a number which has been removed, it could mean big trouble.  If the number was left off for some reason, then what? 

Already thought of this. We are located in the middle of GM town. And like others have many friends. It is going to have receiver x-rayed in **** plant tomorrow which will tell the truth. All measurements and test tonight prove to this point it has not been ground or altered.

Oops, maybe GM, Ford or Chryler.

Pic showing bottom of receiver.

---

Lots of U.S. rifles and shotguns were made in the early 1900s without serial numbers. Licensed dealers, FBI, and BATF know that not all firearms were made with a serial number and they have a process for NSN (no serial number) arms.

I would not have legal worries about the lack of a serial number on this model 1898 Krag. It never had one.

Attached picture of a model 1899 receiver bottom for comparison with OP's.

---

I would not worry about the lack of a serial number either. This thing was built by the government and released to be distributed by the DCM to end up in the hands of civilians.

Another disquieting thought, maybe the step was not forgotten but another defect was encountered and the step skipped on purpose. Were the proof marks just done on the barrel? Was the barrel from another gun and this receiver never proofed?
If a dangerous defect was detected and the receiver was not completed on purpose then it should not have been heat treated. Since mark is having tests run checking for heat treat would be in order. My self I would try a small file test in an obscure area of the receiver, at a automotive lab they will have hardness testers but they will have to keep in mind the shallow case with a very soft core is normal for Krags. It will be very different than the modern automotive steels made today.
Right now I am inclined to think the DCM/NRA, Bannerman, or Stokes made a custom stock for the carbine and released a safe gun to the public.
But as Charlie Brown said "It is a mystery."

Where in Automobile Land are you located?  Both Butlersrangers and I live relatively close to the big D.  Perhaps a physical inspection might reveal some additional insight.  Plus... it's always great to get together at a range occasionally!

I believe Mark said there was a "P" on the receiver ring, which makes this whole circumstance more curious.

You would think someone should have noticed things were not quite right prior to heat-treat. Possibly, this is revealing of the speed and volume of the work at SA!

I bet someone got yelled at in the Hill Shops! A lot of effort went into reducing that 6 and 1/2 pound forging into an almost finished receiver.

Hey, 119 out of 120 operations ain't bad.

I doubt that Mark's Krag was completed into a finished arm at Springfield. I doubt it was ever sold by the DCM.

It seems plausible it was salvaged by one of the surplus dealers, who bought the Armory's obsolete parts, finished and unfinished.

'FredC's' suggestions about testing receiver's hardness are good ones.

p.s. This is fun!

Stamping the number after heat treat seems like it would have been hard on the number stamps maybe impossible. I am pretty sure it was done before, if so a spoiled receiver could have the number noted as it was destroyed and the number assigned to a new one later. Any clue as to whether numbers were reused or just skipped?
Thinking about this some I would have stamped the number after a last inspection prior to heat treatment.
There was that irregularity first noted when the receiver was in the stock, wondering if that prevented it from fitting in the fixture where that late operation was done around the trigger pivot point.

Great info Guys, Just got out of work and can add a few things. Receiver was x-rayed in shop. No welding has been done on receiver.  No numbers have ever been stamped into receiver. Original stamping has not been altered.   

Here is the proof on receiver. 

---

Better View

---

Interesting thread! I cannot believe that receiver was ever put on an arm at SA - it HAD to have come out by the back door, in some manner.

If it were mine, I'd abandon all thoughts of building (forcing) it into a rifle, strip it of parts, and mount it to a nice walnut base and use it as a MOST unique paperweight on my desk. Just my $.02.

Is the reveal 100% that it is not some sort of prototype or maybe a early receiver design that was change for improvement? Would hate to erase a piece of history. This is it before I broke it down to take pictures to answer everyone's questions.

---

Does anyone have input on the small number 50 in stock behind trigger guard.

Stock

---

Wow, this has been a great topic and, as usual, educational.
Vic

Stock sub-inspector's mark. Various numbers are known - had nothing to do with the odd receiver.

FWIW, my paperweight suggestion is not destructive, nor irreversible should other facts emerge over time. Nor would it 'lose' anything, because - in my opinion - there is nothing to be "lost". Look upon it more as righting a wrong. Is it "interesting"? Yes. Should it be "legitimatized" as part of a rifle? No.

One of the key features of military arms is interchangeable parts. That receiver would be useless in the field, since it could not be put into another stock. NO final inspector at SA would ever have passed that as part of a finished rifle.

As has been mentioned, there is also the royalty issue - that is why it was not numbered. As for being an "experimental", the trigger design had been long settled at that point, plus, the work is not "finished" appearance-wise. I believe that it was simply discarded near the end of fabrication, for failing some gauging test, was scrapped, and acquired later by a surplus dealer.

Again, just my opinion - YMMV.

So it is a one of a kind! "Screw up". LOL. Again, appreciate all the great information that's pouring in.

I tend to agree with the idea that the receiver had a error and discarded into the reject pile, however, does the following picture look familiar?:

Taken from "The Krag Rifle Story", 2nd Edition, Subcaliber Attachments, page 96.

You know in U.S. coin collecting there's an "Coin Error" classification.  Wonder if there is one for Springfield Krag errors made by the Springfield Armory?  Not including modifications made post production.

---

Can it be explained a little better to me what a subcaliber attachment is or pictures. Picture is identical to what I have the best I can see.

As a cost saving measure, military units would sometimes use rifle caliber rounds (sub caliber) for artillery practice.  A specialized receiver would be attached to the artillery piece for that purpose.

Wow - GOOD catch!!!! Makes me feel like I have egg on my face.

I'd have to think that is what it was. Of course finding any of the associated linkage is likely to be a long cold trail.

I'm even more inclined now though to suggest leaving it loose - if you want to make a shooter, receivers are not hard to find.

Dick, why you go ahead and wipe that egg off your face because you and other Krag Gurus are a premier source of Krag knowledge and information.  I have truly benefited from you and other's input/insight.

BTW: Did someone say they needed a Krag receiver???

  (You need to Login

'Local Boy' - Great job of spotting that picture of an identical receiver in Mallory's "KRS".

Frank Mallory may have been correct about this 'non-serial numbered' receiver being a component of a 'Sub-caliber Attachment' for Artillery practice. However, I think there are some inconsistencies and errors (which I will explain). 

Both the Army and Navy had sub-caliber devices. The Navy had a device that could be externally attached to deck guns. The Army device was inserted into the bore of artillery pieces. Both systems allowed practice by firing specially prepared .30 caliber cartridges.

Mallory reports Subcaliber Attachments were made at Watervliet Arsenal, using some rifle components supplied by Springfield Armory. Apparently 200 devices were prepared, between 1901 to 1909, for artillery ranging from 3.2" Field Gun to 7" Siege Howitzer.

William Brophy, in the "The Krag Rifle", (pages 206-207), shows detailed pictures of the 'Navy Sub-Caliber Rifle' ("Mark I Mod. 1 Sub-Caliber Rifle"). The device could be solenoid or manually fired, and was attached externally to Deck Guns. The pictures show a regularly machined Krag receiver, with serial number, and a regular trigger which the solenoid plunger strikes. Also, three holes are described as being drilled into the receiver ring to allow the action to be held in the device by lock-screws. The pictured device was made in 1919 by the Boston Scale and Machine Company

Brophy also shows the 'Army Sub-Caliber Rifle', which is clamped into a large bronze collar with interrupted threads, that allow it to be 'screwed' into an open cannon breech. The Krag action is fired with its conventional trigger. The left-side of the Krag action is not shown so I cannot tell if there is a serial number or complete machining.

The picture posted by 'Local Boy' shows a Linkage that I believe to be totally unrelated to the Artillery Sub-Caliber Devices. The Linkage is for a "Cummings Sub-Target Gun Machine". But that's another Story

---

The linkage in the Mallory, "Krag Rifle Story", photo (that 'Local Boy' matched up with the OP's receiver) I believe is an error. It has nothing to do with Firing the mechanism.

The Linkage seems to be the type that was used on Krag and 1903 Springfield rifles adapted for use on U.S. produced "Cummings Sub-Target Gun (or rifle) Machines".

IMHO - It is not a part of the artillery sub-caliber devices. I think the pairing with the 'not quite finished/no serial number' Krag action is a mistake. Sub-Target Gun Machines and Sub-Caliber Devices are different animals. Someone combined parts of unrelated devices.

The "Sub-Target Gun Machine" was a 'rifle shooting simulator'-'training aid' that saw use in the U.S. and British Army & Navy, Military academies, military and public schools, from the early 1900s to around 1920.

It worked on the principle of the pantograph. A rifle attached to the Cumming's machine was aimed at a distant target. A needle mechanism moved in alignment with the rifle's bore, poised to strike a miniature target that was aligned with the distant target.

When the rifle was 'dry fired', a linkage was moved by the rifle's Cocking-Piece. This action pulled a cable that caused the needle to move forward and pierce the miniature target ("sub-target").

If everything was properly adjusted and aligned, the pin hole on the miniature target would correspond to the sight picture through the rifle sights, when the trigger was squeezed. (Yes, we all want one for our Game Room)!

Attached: Said Linkage on a Krag, Linkage on 1903 Springfield, Post Card that is also a miniature target with pin holes.

---

IMHO - It is possible and quite intriguing that 'Marksc' may have a Krag receiver, that was made purposely, for use on one of the Watervliet type sub-caliber artillery training devices.

Attached is a photo of one such device that was excavated on Bannerman's Island (in the Hudson River far north of NYC).

Unfortunately, none of the pictures I can find of the Watervliet sub-caliber artillery devices show the left-side of the receiver.

It seems clear to me that the Krags, used with the artillery sub-caliber devices, had conventional triggers.

That Linkage Mechanism, in the picture supplied by 'Local Boy', is actuated by a small rod joined to the Cocking-Piece. It is not a 'triggering' mechanism. It is designed to pull a small cable a short distance.

---

This all, while fascinating, still leaves the mystery of how this receiver wound up in a finished working rifle, with a legitimate carbine stock and barrel.  Again, me thinks it to be a Bannerman... especially since an Army Artillery example was excavated on the island.  Perhaps ole Bannerman was disassembling the artillery sub-caliber devices and assembling the receivers into full rifles and carbines! 

In the 1920s and 1930s there was probably little market for surplus sub-caliber devices. Parts and labor were cheap and available.
There was a market for low cost Krag rifles adapted for Hunting. Bannerman, Stokes Kirk, and Sedgley are well known for cobbling together salvaged parts to make improvised, sellable guns.

Worth noting:
Krag receiver rings often seem to have other symbols or letters stamped on the front face, in addition to the proof "P" and barrel index-mark. I assume these are inspection stamps.

It is interesting that Mark's receiver has a "W" at the 6 o'clock position. This begs the question if this "W" might have denoted a receiver earmarked for use at Watervliet Arsenal?

Got back kind of late. So I need to take back my suggestion that the Krag receiver is would be too hard to stamp after heat treating. Evidentially the proof stamp was done afterwards.

Mark bought this thing as a Krag carbine, IF the photo shows all to be carbine "correct" could his be more valuable as a carbine, especially if it was assembled by SA? The front sight looks good and not a 1903 modified sight.

'FredC' - As always, you bring up interesting points.

I have been thinking about and trying to understand the process, sequence, and rationale of Springfield stamps and marks for years. (British, German, and French ones, too)!

The marks mean something. They were done with a purpose and communicated information to someone at sometime.

I once assumed gun makers stamped metal components when they were 'soft' and before they were heat-treated and finished.

However, final inspection and proof stamps make little sense if they are done before a component is Finished and Correct.

Now, I don't know for sure and freely admit my puzzlement.

Some U.S. Krag markings look like they were done in the back room of a 'Khyber Pass' gun-shop. There are lots of funny hieroglyphics around the magazine opening and on the back of the bolt-handle root, odd 'steel lot' symbols, periodic changes in serial number size, style, and font. Some numbers and letters are very difficult to read 'correctly'. Some stamps appear very primitive given that they were done in a relatively modern 'machine age'. The system appears like it was always changing. Such is the charm of Krags!

I agree with Dick Hosmer. The OP's Krag did not leave Springfield Armory as it is now assembled. SA would never have compromised a Gun Stock by gouging out all that extra wood. It is not a model 1898 carbine. The most interesting aspect of this Krag is the novelty of the receiver!

We have not seen a close picture of the front-sight. So I am not sure if it is a carbine barrel. (This is actually irrelevant. All types of Krag parts were in the hands of surplus dealers). This is a 'parts gun' assembled outside SA.

One thing that impresses me about SA 'markings' is that they don't seem to disrupt or displace the surrounding metal. Oddly, sometimes a SA marking will be inconsistent in its depth. Clarity of letters and numbers is variable.

IMO-Fake marks are noticeable because they disrupt surrounding metal and finish.

---

BR,
Good point on the no raised metal around the "US Springfield Amory and serial number.

"One thing that impresses me about SA 'markings' is that they don't seem to disrupt or displace the surrounding metal. Oddly, sometimes a SA marking will be inconsistent in its depth. Clarity of letters and numbers is variable. Fake marks are often noticeable because they disrupt surrounding metal and finish."

The only way to have done that is to buff them before heat treat.
The proof "P" on the front to of the receiver, does it have raised metal around it? If so it and any other marks like it would have been done after heat treatment.

I just want Mark's carbine to be a one of the kind original worth a million bucks. Then we could all wish we had one also.

He are some more numbers that is on front of receiver. Nothing stamped on the gun has any raised material around any stampings.

---

Here is a better picture I found of the receiver.

---

IMHO - "P" and other letters on front of receiver ring are clean and free of deforming. Sometimes the "P" is not deep or clear.

Somehow, SA seems to have cleaned up around stamped numbers and letters, even if the steel was hardened??? It makes me ponder.

I posted a picture of a receiver that shows SPRINGFIELD heavily struck at both ends, but, lightly struck in the middle. All I can envision is that the Die had begun to bend?

Sure wish someone had asked all these questions before all the guys doing the work had died.
When hand stamping stuff you sometimes do not hold the stamp square and get a very heavy impression on one side and not anything on the other. You try to fix it by setting the letter punch in the deep side and rocking it over then striking it again. About half the time it works real good. The other half not so much.
But on these long words like Springfield Amory, they would have had a single punch with all the letters or a frame with replaceable elements, probably applied with a lever press or the like. Today they would have a roller die that would travel the length applying one letter at a time. But not sure if they had that capability back then or not.
Several possible causes for that one that was deeper on the ends, the receiver could be low in the middle or if misapplied on the first impression lined up by hand with too much correction on the second application.
All those marks by the bolt recess and others visible on the assemble rifle I would expect to be filed or buffed before hardening. If it was possible to do after heat treat the only ones with raised metal I would expect would be final proof marks and such. If filed or buffed it would have take the final case hardening colors off. 
One thing I know for sure I was not there to see how it was done!
If someone wants to buy that pile of receivers and send them to me, I will do a high magnification analysis and give you all a more educated guess.   (You need to Login

Went back and looked at Mark's proof mark and it does look like a file was applied to it after stamping, maybe they did a second tempering to get color and slight corrosion resistance after it was stamped? What do you guys see on other proof marks?
A second tempering would be done at a slightly lower temperature. For those that do not know tempering is usually a second process done after the hardening. Double tempering is not common.

Here is a peek at "P" mark of two more model 1898 Krag receivers.

1st two pics are of a '98 receiver made around November, 1899. The "P" is almost impossible to identify as a P. There are various scratches from stone or file work on receiver face. A lot of the case-color survives.

The last two pics are a '98 receiver made around April, 1903. Markings are clear, but, the finish is probably Armory refurbish prior to WW1. Also has a "DD" stamped on front face of receiver.

---

BR,
I could not see any raised metal on those stamps. It looks like each was filed or buffed after stamping. Some time ago you mentioned the processes done to each part as far as finishing. If I remember right the barrel and receiver had different types of finishing. Do you think they pulled them apart one last time before finishing after proofing?
Was refinishing part of the arsenal refurbs?

Proofing was one of the first things done, before ANY finishing operations. Note that the last "proofmark" (actually more of an inspection for functioning, since the barrel was deemed good at that time) was the (P) which was struck in the wood - no worries about buffing over. Yes, the receiver and barrel had different finishes - were they disassembled after once being joined? I was not there, but I'd have to say "NO". You can get a fairly decent picture of Krag fabrication from reading OM22. VERY little changed between 1878 and 1892.

As I understand it, Krag barrels were tested with special proof cartridges (developing 70,000 lbs. per square inch) after the bore was drilled and an interim chamber reamed.

This happened prior to the barrels being rifled. The un-rifled barrels must have been screwed on to some type of 'master action' for proof firing. (They likely were stamped with a "P" at this point).

After rifling, the barrels underwent all of the finishing operations: multiple straightening's, final reaming, 'timing of threads, drilling & threading sight-holes, brazing front sight block in dovetail, and 'browning' (bluing).

Krag actions went through their intricate manufacturing process and were proofed sometime after heat-treat. I suspect, the actions were screwed onto some type of special 'master barrels' for testing.

I don't think a Krag action met its ultimate barrel until final assembly. Close tolerances and gauging must have allowed the index marks to 'line up' as the finished barrel was screwed into the finished receiver.

'Timed Threads' & parts finished before assembly - It blows my mind!  (Not how it would likely be done today).

Once Krag rifles or carbines were assembled, I believe a percentage were subjected to firing another proof cartridge, all were tested for function, and received the "Circled P" on the stock wrist.

I could have this process wrong in part or whole. I base my understanding on the "Scientific America" description of Springfield Armory manufacture of the Krag.

I've also attached a later description of the final proof of 1903 Springfield rifles. By that time, all assembled rifles were proof fired before receiving "P" stamp on stock.

---

It may be that there was an intervening process between the TD and the '03 which called for a proof cartridge AFTER assembly - just don't know.

Distinction needs to be made between "proofing" and "guaging". While they needed to weed out any bad apples, they HAD to have pretty well satisified themselves that the barrel was sound before expending a lot of labor on it. No one expected a rifle to fail structurally after assembly.

As to timed threads, how else would you do it? The extractor cut has to line up - it isn't made later after they thought the barrel was 'tight enough'. They were clearly able to index the threading and machining of the parts such as that when the witness marks aligned, all was good.

Just a thought swimming around in my head about the stamping of numbers and lettering on the receivers.  Is it possible that the numbers were pressed rather than stamped... as in a slow hydraulic press operation as opposed to striking with a hammer?  It seems to me that slowly pressing the stampings rather than a quick sharp strike would be less likely to produce raised material around the number/letter. 

Just a thought anyway...

The metal has to be displaced whatever speed it is done at. I would guess that a fixture was used for the US SA stamp with a hand lever toggle set up to get enough pressure. I have used and built air toggle presses and you probably use hand operated toggle press to reload. The serial number could have had a stamp holder with replaceable elements or rotating wheels. It may have been semi automatic as it would be difficult to number receivers all day long without using the same number multiple times and skipping others. When you have done production machining for 40+ years it is easy to visualize all the things that can and do go wrong.