Serial Number Muddle In Early Wheatstone Ledgers

[Robert Gaskins]

Wes Williams has just published an important new article about the "Serial Number Muddle in Early Wheatstone Ledgers" at the Concertina Library.

 

This deals with the strange jumble of serial numbers in the four oldest ledgers, covering sales from the 1830s to early 1854. Wes has transcribed all the serial numbers, and by scatter-plotting serial numbers against dates he has revealed that a major part of the jumble comes from the fact that several different ranges of serial numbers are being sold at the same time. After 1850, particularly, this pattern intensifies, with seven or more different ranges being sold over multiple-year periods each.

 

The demonstration of this unexpected pattern does not yet explain WHY multiple ranges of serial numbers were being sold, but there can't be much doubt that it is a fact, which will focus further work to explain this observation.

 

The article is at http://www.concertina.com/williams/serial-number-muddle.

 

Wes promises an update to the article soon, extending the graphs through the 1850s and 1860s and discussing the different patterns of serial number sales as the Wheatstone company goes through changes in manufacturing and management. But it seemed important enough to publish this "in progress" version now.

 

As a further service, Wes has also published his transcriptions as Serial Number and Date Indexes to Wheatstone Ledgers. So far the indexes corresponding to the time-frame of the article are available: serial number indexes to ledgers C104a, C1046, C1047, and C1048, plus a date index to ledger C104a (the other three are basically in date order anyway).

 

Each of these indexes is a webpage listing all the serial numbers (or dates) in sorted order, with the ledger name and page number and a live link to the photograph of the original ledger page.

 

For anyone who wishes to replicate Wes's results or undertake related work, Wes has futher made his transcribed data available in two formats: CSV (comma-separated value) format, which can be manipulated in Excel and other spreadsheet programs, and XML (extensible markup language) format, a modern data format which is easy to deal with in "bespoke" programs and increasingly in standard applications.

 

It is planned that the indexes to the ledgers will be extended to cover the rest of the volumes, and further indexes (and data sets) will be provided as transcriptions are completed.

 

Speculation is welcome in this thread about how and why the patterns uncovered by Wes might have come about, but the hope is that some additional evidence will gain added significance in light of these observations, and eventually a well-supported explanation will be established.

 

Published now at http://www.concertina.com/williams/serial-number-muddle

[wes williams]

Wes promises an update to the article soon, extending the graphs through the 1850s and 1860s and discussing the different patterns of serial number sales as the Wheatstone company goes through changes in manufacturing and management.

I've now completed transcribing all the Wayne sales/production ledgers by date/serial, up to and including C1054 (to 1891), so if anybody has a desperate need for data for ledgers later than the article provides, they can send me a message. Otherwise, they will be available with the completed article, sometime next year. I will also update the search and plot utilities at the same time.