ISMACS Logo

International Sewing Machine Collectors' Society

The purpose of the International Sewing Machine Collectors' Society is to foster the collecting of, and research into, sewing machines.

Why Two Serial Numbers?

John Langdon

Singer stamped its early machines with two numbers. Officially, what we know as the serial number was called the ‘Register Number’ and, until the end of 1899, it kept a running total of all the machines made by both the USA and Scotland factories.

The second number – almost always numerically smaller than the first – was known as the ‘Factory Number’ and usually reflected a running total (plus 1) of that particular class, at that particular factory. At least, that’s how it was in Scotland.

When Singer’s Glasgow factory opened in 1867, the first two runs of just 100 New Family each were commissioned to start from Factory Number 0 to 199 (hence the ‘plus 1’).

In those days, it was common for the company to start production of a new machine with a very small run, in order to iron out technical issues and provide key sales points with samples. The subsequent New Family batch, commissioned early the following year, was for a substantial 10,000 and continued the Factory Number from 200 to 10,199.

Until Glasgow started making other models, (1869) the Register number remained identical to the Factory No.

In America, the numbering was different. Possibly because, initially, Singer himself had been responsible for it - the semi-literate serial adulterer wasn’t the most fastidious of record keepers - and possibly because Singer’s membership of the Sewing Machine Combination required some esoteric accounting. That seems unlikely, however, as no other member of the Combination saw a need for two serial numbers.

Few Singers from the 1850s and early 1860s have come to light but those that have show a Factory Number that is always 4,000 less (OK, fewer) than the Register Number, regardless of the class of machine. The second number is sometimes associated with the Register Number, either stamped next to it, or below, but often it is to be found hidden away, stamped on critical machined parts.

On the ‘Number Two’, for example, it can be found on the top of the needlebar and/or on the shuttle race, under one of the slide cover plate, and often on the shuttle drive wheel under the bed. On an early class 12, the numbers were stamped on several underneath parts and around the main shaft at the wheel end.

Why this second number was used at all is a mystery. Important parts could have been as easily stamped with the Register (serial) Number for identification.

Before moving to Elizabethport in 1873, the production line at Singer was not the model of efficiency it was to become. The factory employed the cumbersome ‘European System’ in which individual castings were machined and had the component parts literally filed to fit by teams of skilled and costly ‘fitters’.

Consequently, the parts from one machine wouldn’t fit on another and had to be kept together with the original casting when the completed machine was disassembled for japanning and decals. It was assembled yet again before dispatch and the identifying Factory Number on the parts ensured a correct match.

Eventually, in the early 1890s, Elizabethport fully converted to the American System which had been pioneered by the domestic arms industry. Precision parts were made to an exact pattern; ‘gauged’ to guarantee fit and interchangeability and put together on an assembly line by a cheaper, semi-skilled workforce.

Inexplicably, the difference between the two numbers in the USA remained a constant 4,000 until 1867, when the company suddenly switched the Factory number to show what appears to be a running total of each model. These ‘totals’ for each class/factory were subsequently kept separate.

The problem is that no one knows at what number these new running totals started. Did Singer re-start at zero for each class? Did it continue from whatever the Factory Number had reached at the time of the change? Or did it - and it’s a lot to expect - start each class with the actual running total? And was the ‘4,000 difference’ factored in?

Another blow to the voracity of Singer’s ‘running totals’ – in particular, the Class 12 – was the mysterious return to zero after the Class 12 Factory Number reached 999,999, during a batch commissioned at the end of 1877. I don’t know what prompted it, unless the numbering machine only had six digits.

One useful aspect of the Register and Factory Numbers is the numerical difference between them. It remains the same for an entire batch. Come the next batch, the Factory No. would pick up from where the last batch had finished while the serial number increased by the number of other machines made between the two batches. Thus the difference between the numbers would increase for every new batch.

Starting with the Improved Family (1879), Singer gave all new models just one number and, in 1889, mid-way through a batch of 12K, abandoned the two number system on its remaining machines.

So, the second number on American-made Singers, I’m afraid, doesn’t mean very much at all. At least, not yet... (JL)