RIA 1907 US bayonet, SN 81946

[Mariposawrick]

I have an RIA 1907 bayonet from my father. It is in a leather scabbard, stitched on one long side, has an over-fitted leather hanging sleeve with curved hooks that slides over a metal sheath at the top of the scabbard. Serial number on the blade is 81946, clearly stamped below US. on the other side the stampings are RIA, the ordnance stamp and the date 1907.

There is a post in this forum that lists serial numbers for Rock Island bayonets in 1907, but they do not jive with mine.

Can someone help me clarify the markings and significance of this bayonet?

Thanks

[Johnny P]

There is no record of the actual serial numbers for a given year, and any of the lists are just estimates. They were not necessarily assembled in the order they were numbered either. There are large overlaps in the numbers.

[5MadFarmers]

What Johnny said. Except for this part:

and any of the lists are just estimates.

I have the production numbers made. I've also been assembling a table from observed examples. So it's not an estimate but a table of physical specimens observed in the wild. The intent is to eventually see what I can gain from going at it from both sides.

Rock Island had some unbelievable overlap in 1918/1919. They only made under half a million bayonets but I have very clear pictures of a 1919 bayonet fully 100,000 in serial lower than one dated 1918.

The overlap of your bayonet is quite small compared to that.

The 1905 hospital corps knives also overlap. Again, based on physical specimens.

[Johnny P]

Estimate in the sense that no records exist as to the actual beginning and ending serial number for a given year. Would guesstimate also be inappropriate?

[5MadFarmers]

Estimate in the sense that no records exist as to the actual beginning and ending serial number for a given year.

Indeed. Nothing even remotely along those lines was ever generated or kept.

Would guesstimate also be inappropriate?

I don't know. I'll explain what I'm doing and then you guys can apply a name to the process.

Bayonets are dated with the serial on the same blade. This makes it a lot easier than the rifle/barrel serial date line up done on, e.g., M-1903 rifles.

The production reports are FY. This is a problem as the markings are CY. So I'm not going to do a Poyer and just kind of make it all up from simple math:
FY 05-06 production is X so X/2 is half year. Add that to half year from FY06-07 for calendar year production. Start with "1" and then roll forward.
That kind of voodoo doesn't enthuse me. It's not really possible anyway as the first bayonets aren't serialized and, further, they made bayonets in unfinished (but likely serialized) form as spare parts. So lots of voodoo of the type which passes as "gospel" in various books. That's a "estimate" or "guestimate." I prefer "nonsensical voodoo" as the correct term.

Create a spreadsheet. Create a year column. Then a column for "low," "high," "total," and "gap." With the area for the "low" and "high" established, then focus on those. As an example Springfield production for 1918 is 770014 to 1047349 or thereabouts. With that data established focus only on the ends. Something like:
1917: lowest observed/highest observed
1918: lowest observed/highest observed
1919: lowest observed/highest observed.

This will give us three useful things:
1) The range of the 1917/1918 overlap.
2) The range of the 1918/1919 overlap.
3) The total possible number of 1918 serialized bayonets accepting that the number is inflated by the two overlap ranges. So something like "300,000 total with that including a 1,500 overlap range at the bottom and 1,000 overlap range at the top." So production of blades is somewhere in the 297,500-300,000 range.

I have no idea what to call that. It's not guesses based on the production numbers as it's from observed examples. I'm not really guessing the "starting serial" for a year as I'm tracking the "lowest observed" serial.

How about "pointless exercise in serial/year data?"

The biggest value is that it at least gives me enough data to make generally accurate statements like: "1916 and 1921 dated Springfield M-1905 bayonets had roughly the same production quantity (15000-16000) and are thus more scarce than 1911 (30,000) but not as scarce as 1922 (2,000).

How's that?

[Mariposawrick]

Thanks guys. You have certainly put alot of thought into the available numbers. So who has a thought about my 1907 RIA with the 5-digit number, 81946? It is an RIA 1907 scabbard as well. Thanks. Chad