Notes Relating to the Establishment of a Provisional Serial Number List

 

The following benchmark instruments were used to establish provisional list.

Fluctuations in the production numbers were based on the following:

Anomalies:

Obviously, some adjustments were made to the production numbers in order for them to jibe with the serial numbers of the benchmark instruments. In some cases, it was not possible to create a matrix which satisfied both areas. It is also certain that some of the information provided on the benchmark instruments, while generally quite close, may be off by a year or two.

What was thought would be one of the most important statements about production actually turned out to be one of the single most problematic facts--the 1913 advertisement which touted a total production of 60,000 instruments to that date. In the final analysis, this number simply didn't jibe with either the production figures as presented in earlier articles or the benchmark instruments. Two possible explanations seem plausible.

It was common practice among the instrument manufacturers to produce instruments which were sold to retailers without a signature. The retailiers would then stencil their trademark on the instrument and sell it as their own. It is possible that York did not provide a serial number for instruments which were to be sold in this manner, and this would account for at least part of the disparity.

A second possible explanation is that York produced many instruments either without serial numbers or with a different serial number sequence. Drums, for example, are not likely to have had numbers, and it is possible that the stringed instruments sold under the York name had their own sequence. The Grand Rapids Band Instrument Company, a subsidiary of York, evidently had its own serial number sequence since the details of the GRBI instruments are not consistent with those of York from the same time period. This could also account, at least in part, for the differences. Finally, it is possible that York did not begin to put serial numbers on any instruments until quite late in the 19th century.

Of the two explanations, the first is least likely, in that at least one of the sources, an individual who worked at the York plant for many years, indicated that the serial numbers were stamped on the valves and the valve assemblies were made at a different time in a different part of the plant. If in fact the serial number was stamped on the second valve during the valve manufacturing process and not during the final assembly, then the first explanation seems least likely, in as much as the valve makers would have little idea about which instruments were to receive York signatures and which were to be sent out unstencilled.

As was stated before, the numbers provided for the early years are completely conjectural, based only on what seems to be logical increases in production up to the time of the first verifiable instrument. It is curious, however, that a company with the capacity to produce 3600 instruments per year by 1908 would have such low production figures in the previous decade. The most likely explanation is compatible with the explanation for the inconsistency between the serial numbers of instruments and York's advertised claim to have made 60,000 instruments in thirty years. That is, York did not begin to put serial numbers on their instruments until the last two or three years of the 19th century. If this was the case, then serial numbers would not start until about 1898, and the yearly production figures would be compatible with those estimated for the early years of the 20th century.

Given the foregoing information, here is presented the revised serial number list for York. While I am confident that the suggested list presented here is more accurate that the original document which appeared in my dissertation some 12 years ago, (and which has since been reproduced in a number of other documents) I would caution that it is still far from a definitive list. As with my original document, I would suggest that a buffer of two years on either side of the target date would be a better way to date an instrument. A five year spread will be a much more accurate.

 

1996 by John J. Swain

Return to Table of Contents

Go to Provisional Serial Number List