US Test AE's - 900 vs. 100

[Ron Wood]

Boy, you guys know how to bring on the heat! First of all, if you expect me to resolve this controversy, it ainâ??t gonna happen. The documentation just isnâ??t there to establish any concrete answer. To summarize the data I have collected, there are 5 recorded serial numbers â??below the rangeâ?� (this does not consider serial number 6009 listed by Scott Meadows since it is so far below), 235 in the â??accepted rangeâ?� of 6100 to 7100, and 27 â??above the rangeâ?�.

First letâ??s look at the only reference that establishes the accepted range. Mike Reese states in his seminal work â??1900 Luger U.S. Test Trialsâ?� that â??Evidence received from the Bureau of Accounting sets the serial number range of the 1,000 test pieces at 6099-7098â?�. That is the only statement I can find that specifically cites a reference for the serial number range. Oddly enough, with all the documentation reproduced by Reese in his book, this vital piece of evidence is not presented. Without being too judgmental (I think), this reduces this â??evidenceâ?�, valid or not, to hearsay.

Next, consider Scott Meadows equally well researched â??U.S. Military Automatic Pistols 1894-1920â?�. He indicates the serial number range is â??6000-7500 (Approximate)â?�. He lists as recorded test pieces 2 â??below rangeâ?� and 7 â??above rangeâ?� examples. Following Albertâ??s line of reasoning, Meadowâ??s out of range pieces are speculation. I have somewhat subscribed to that thinking by discounting number 6009.

Next, we have the serial numbers acquired by Bannerman that includes 9 â??above rangeâ?� examples. Albertâ??s statement â??â?¦if Bannerman was simply mentioned as a surplus purchaser because he happened to exist in business at the same time or after the US Trails, then the data (and without serial numbers) is not strong enough and can only be considered as a hypothesisâ?� is not germane since his purchase of 780 Model 1900 pistols is documented by serial number, therefore the above range numbers cannot be summarily dismissed.

So far we only have Reeseâ??s unsubstantiated range to go by, and there is no evidence that this range contains contiguous serial numbers. If we are to accept his work as absolute, we will have to rationalize a couple of other inconsistencies in his account. In his section on â??Sale, Receipt and Allocationâ?� he lists â??5 â?? Each commanding officer of the 185 troops of cavalryâ?�, which would amount to 925 weapons. Later he gives a detailed listing of the cavalry troops that received the trial Lugers, and lo and behold, they only account for 180 troops! He discounts the missing 25 Lugers by a footnote that states â??Plus 100 pistols to Springfield for exclusive testsâ?�. This 100 would include the 10 pieces to West Point, 65 pieces to artillery units, and the single weapon issued to LTC Marion P. Mans in California, leaving 24 wandering around Springfield somewhere. This is not to debunk Reeseâ??s entire work, but only serves to illustrate that when it comes to Test Trial Lugers, there is always some ambiguity involved. He resorted to some â??guesstimationâ?� to make his numbers come out right. Is there any other part of his records that might be conjecture and therefore suspect?

OK, now letâ??s look at Kenyonâ??s statement â??The first 900 bear serial numbers in the 6,100 to 7,100 range and the second lot bear numbers in the 7200 rangeâ?�. Based on observed serial numbers, there is nothing to indicate that serial numbers from 7,100 up to 7,200 were skipped! He probably should have stated more correctly that the second lot bears numbers above 7,100. I think that Kenyonâ??s statement is based on conjecture derived from observed serial numbers as I know of no documentation to substantiate it.

Is there anything other than the existence of these â??above rangeâ?� examples to further bolster his theory? Perhaps. Hans Tauscher communicated in August of 1901 that the 1000 test pieces would be delivered in two lots of 500 each before the end of September. Yet the weapons were delivered in lots of 800 and 200 near the end of October. This might be due to delays incurred by the shipping means of those times, or it might mean that production commitments necessitated fabrication of the deliverable pieces in a manner that was not originally anticipated and therefore not necessarily consecutively serial numbered.

If we accept Kenyonâ??s postulation, then the 235 numbers in the â??accepted rangeâ?� that I have recorded represent approximately 26% of the supposed 900 in range examples delivered, and the 27 â??above rangeâ?� serials recorded are 27% of the last 100. A remarkably similar survival rate.

Does any of this resolve the issue? NOPE, just stirs the pot some more. Bottom line is that if you want to be absolutely sure you have a legitimate Test Trials Luger you need to heed Albertâ??s admonitionâ?¦ â??Therefore, if a serious collector wants an undisputable 1900 US Test Trails Luger, find one that falls in the 6100-7100 serial rangeâ?�. If, on the other hand, if you believe the preponderance of physical evidence presented by Lugers with Test Trial attributes that exist â??out of rangeâ?�, you may wish to side with Kenyon and be happy with your odd-ball Test Luger.