4,87x,xxx SC rifle question

[IditarodJoe]

As I understand it, in October of 1943 Smith-Corona made an error in numbering upward of 5000 1903A3 rifles. I found this information on Vi Shooter's M1903 Page (http://www.vishooter.net/SC3.txt). On rifles whose serial numbers should have been in the 4780189 - 4784296 range the second and third digits were apparently reversed. An S-C that I bought from the CMP in 2007 is one of those - 487021x. Like most of the ten examples listed in Vi Shooter's receiver/barrel correlation list, mine has an SC 8-43 dated barrel.

My question is, what other references to this numbering error exist? Have other authors written about it? Has any mention of it been found in factory or government records? Just curious to know how well documented this is. Thanks.

[John Beard]

Yes, other authors have referenced it, including Clark Campbell and perhaps Bruce Canfield. But they mistakenly assumed that Smith-Corona, to reach that serial range, had made upwards of 30,000 spare receivers past S/N 4845000 (approximate last serial number) which were unaccounted for in government records. Studies have conclusively shown, however, that the unusually high serial numbers resulted from a serializing error as you describe.

J.B.

[IditarodJoe]

Thank you, John. I appreciate your reply. It would seem logical then that there would not be any properly serialized receivers in the 4780189 - 4784296 range (assuming that there was just one station for stamping the serial numbers).

[John Beard]

Your logic is in error. Smith-Corona caught the serializing error, went back, and resumed serializing where they left off. So the gap you suggest does not exist.

One can reasonably assume that, when Smith-Corona reached the S/N 4870000 range, they would have skipped the previously-used numbers. But alas, production was suspended before they got there.

J.B.

[IditarodJoe]

Interesting choice on their part. (For my "Needle in a Haystack" file: Somewhere out there may still lurk the 478 s/n variant of my 487 rifle!) Do you know if anyone is currently attempting to add to Vi Shooter's "Serial Number/Barrel Date Lists"? As the release of those rifles to the public slips further and further into the past, the opportunity to capture that data is fading. We should certainly be grateful for the information that has been gathered thus far.

As always, John, thanks for sharing your knowledge.

[Rick the Librarian]

I may be mistaken, but I don't think ViShooter has worked on those tables for some time.

[John Beard]

Interesting choice on their part. (For my "Needle in a Haystack" file: Somewhere out there may still lurk the 478 s/n variant of my 487 rifle!) Do you know if anyone is currently attempting to add to Vi Shooter's "Serial Number/Barrel Date Lists"? As the release of those rifles to the public slips further and further into the past, the opportunity to capture that data is fading. We should certainly be grateful for the information that has been gathered thus far.

As always, John, thanks for sharing your knowledge.

A more interesting duplicate serial number variant is the Smith-Corona/Remington serial number overlap beginning at S/N 3708000. The Smith-Corona rifles were subsequently stamped with a C-prefix to the serial number.

J.B.

[m1903rifle]

I more interesting duplicate serial number variant is the Smith-Corona/Remington serial number overlap beginning at S/N 3708000. The Smith-Corona rifles were subsequently stamped with a C-prefix to the serial number.

J.B.
Do you mean like this ?:

[John Beard]

Yes! Now find the Remington with the same serial number.

Thanks for sharing.

J.B.

[Major Tom]

WOW! Sure alot of' Xs' in original post. I always wonder why people do that. Not that I'm being critical tho.

[IditarodJoe]

Major Tom, there's actually only one "x" in the original post. My question referred to a range of up to 5000 serial numbers. The specific start and stop points of that range are apparently not known but they all begin with the digits 487, hence the 4,87x,xxx reference in the title of the thread. Hope that helps.

[IditarodJoe]

I more interesting duplicate serial number variant is the Smith-Corona/Remington serial number overlap beginning at S/N 3708000. The Smith-Corona rifles were subsequently stamped with a C-prefix to the serial number.

J.B.
Thanks for the interesting information, John. Do you know if any detailed production records remain for this time period, either from Smith Corona or the War Department? One would assume that the receivers with the transposed serial numbers and those with the corrected numbers were all included in Smith Corona's production totals for October 1943. Same for the serial numbers that overran S-C's assigned block in July 1943.

[Kurt]

Here's another;

34808

[John Beard]

Thanks for the interesting information, John. Do you know if any detailed production records remain for this time period, either from Smith Corona or the War Department? One would assume that the receivers with the transposed serial numbers and those with the corrected numbers were all included in Smith Corona's production totals for October 1943. Same for the serial numbers that overran S-C's assigned block in July 1943.

I suppose production reports from Smith-Corona to Army Ordnance exist somewhere, but I am not aware that anyone has reported finding them. Your assumptions would be correct. The Army was billed for the rifles, not by serial number, but by crates of rifles that passed Ordnance inspection. When the Ordnance inspectors signed off on the inspection reports, the rifles became government property and the contractor could bill the government for them.

J.B.

[IditarodJoe]

After some digging around, I found the information in a table in the back of Clark Campbell's book "The '03 Springfield Rifles' Era". The table is based on "a declassified Industrial Division Small Arms Branch tabulation of monthly Remington and Smith & Corona WWII rifle deliveries (now in the National Archives)". It appears that the highest 03A3 serial numbers used by SC under their assigned first (3,711,999) and second (4,845,831) blocks of numbers are known, as is the total number of rifles delivered by SC through February 1944 (234,580). It is, therefore, possible to calculate an average "scrap rate" for receivers (disregarding the 4000-5000 improperly numbered receivers), and back-calculate the approximate serial number range used each month based on the number of rifles delivered to the government during a that month. From his footnote, it appears that this is more or less what Campbell did to create his serial number chart.

What I find fascinating about this is that (1) if there is strong evidence that a transposition of digits in upwards of 5000 serial numbers did in fact occur, and (2) if any SC rifle between 4,780,189 and 4,784,296 has been found, thereby demonstrating that the company resumed numbering at the point at which the error began, then the scrap rate for receivers jumps by around 2%. This would mean that the serial number/production date estimates for mid-production rifles could be off by several thousand numbers, indicating that the rifles were assembled somewhat earlier than is currently thought.

I realize these are only approximations, but it's interesting nonetheless.

[John Beard]

After some digging around, I found the information in a table in the back of Clark Campbell's book "The '03 Springfield Rifles' Era". The table is based on "a declassified Industrial Division Small Arms Branch tabulation of monthly Remington and Smith & Corona WWII rifle deliveries (now in the National Archives)". It appears that the highest 03A3 serial numbers used by SC under their assigned first (3,711,999) and second (4,845,831) blocks of numbers are known, as is the total number of rifles delivered by SC through February 1944 (234,580). It is, therefore, possible to calculate an average "scrap rate" for receivers (disregarding the 4000-5000 improperly numbered receivers), and back-calculate the approximate serial number range used each month based on the number of rifles delivered to the government during a that month. From his footnote, it appears that this is more or less what Campbell did to create his serial number chart.

What I find fascinating about this is that (1) if there is strong evidence that a transposition of digits in upwards of 5000 serial numbers did in fact occur, and (2) if any SC rifle between 4,780,189 and 4,784,296 has been found, thereby demonstrating that the company resumed numbering at the point at which the error began, then the scrap rate for receivers jumps by around 2%. This would mean that the serial number/production date estimates for mid-production rifles could be off by several thousand numbers, indicating that the rifles were assembled somewhat earlier than is currently thought.

I realize these are only approximations, but it's interesting nonetheless.

The two hypothetical "If's" in your statements are not "If's" at all, but established fact.

Widely-published WWII serial number tables are little more than unverified speculation.

J.B.

[IditarodJoe]

I absolutely take your word for it, John. Just for kicks, I'm inclined to make up my own SC "serial number/production month" table including what little information I have about the range of transposed number receivers (4870189-4874296 based on ViShooters table) and the highest known overrun serial number (3713068 based on the photo posted by m1903rifle on this forum). And before you take me to task, John (he said with a grin!), I'm aware of the speculative nature and inherent inaccuracies of such an exercise. I'm also aware that those with expertise in this area have knowledge of a lot of factors that a novice like myself does not. I'm just curious to see how the results would compare with some of the existing tables.

If I do this, I'll post the results here. Regards, IJ

[John Beard]

As Inspector Harry Callahan once said, "Go ahead..., Make my Day!" :icon_lol:

J.B. :hello:

p.s.,

I was a Scoutmaster for 13 years. I like your signature about campfires.