2 ¾ hp Gearbox vs Engine Serial Numbers 1907 to 1926

[Hutch]

Some investigations into 2 3/4hp Gearbox vs Engine Serial Numbers


Firstly, a big thankyou to the LDMCC, Len Boydell and the previous Machine Registrars (since the inception of the Douglas Machine Register in 1971) and all the members who have contributed their machine information. Without their efforts my ability to do this research would have been near impossible!

I gave a short introduction to the following work here;

https://www.douglasmotorcycles.net/index.php?topic=9218.0

Looking at the 2014 Machine Register I could see that there were some obvious trends in the gearbox vs the frame and engine serial numbers for the 2 ¾ hp machines, but sometimes it was not clear. The records seemingly matching well for some years but not so well for others.

What also became evident from looking at the data, is the number of machines that have been mixed and matched over the years and also possibly incorrect data due to typo’s / serial numbers incorrectly read and recorded. Also, there are a number of machines in the Register where the gearbox serial number has not been recorded.

I plotted a chart of frame serial numbers against engine serial numbers and found that they roughly matched on a fairly obvious trendline (or correlation line as “Cardan” describes it), e.g. the frame numbers are generally in the range of the engine numbers year by year over the manufacturing period of the 2 3/4hp machines. This finding appears to correlate reasonably well with what is documented in the immensely useful The Identifying Old Douglas Motorcycles, thread (Part 1:1907-26 Models 2-3/4hp, 3-1/2 and 4hp https://www.douglasmotorcycles.net/index.php?topic=488.0) by Doug Cross and Doug Kephart.

 Unfortunately, there is no gearbox serial number information given for the 2 3/4hp machines in the versions I looked at (Veteran 3.6 and earlier – since about 2012 or so).

I had a thought that there may be enough information in the 2014 Machine Register to be able to extract some clues about gearbox serial numbers using data that fitted obvious correlation lines. I presume that other people may well have gone down this same path, but a quick search has not revealed any data published on the forum – please let me know if you have any knowledge of previous work as I would be interested to hear about it.

The information I used for this work has been used with permission from the Registrar of the LDMCC. I have no control at all over the accuracy of this information in the LDMCC Machine Register and the following is only my interpretation of that data. Therefore, the following may only give a rough indication of the relationship between gearbox and engine serial numbers for the 2 ¾ hp machines and I cannot accept any responsibility for any errors that may occur through the use of this information for identifying or dating motorcycles or parts thereof. This work is definitely “work in progress” and there is much more research to be done.

[Hutch]

If we plot the graph of gearbox vs frame serial numbers (CW not included), we get a similar result except the relationship is not quite so clear during some time periods, for example with the introduction (in 1915) of 3 speed gearboxes and what happened in 1920. Unfortunately, as mentioned before, we also have quite a few less “nicely” correlated data points to play with as there are several machines, although they have their original frame and engine combinations listed, do not have their gearbox details recorded.

[Hutch]

To help clean up the data I manually (using my best judgement, but obviously prone to human error and bias) removed what I thought were outliers, i.e. data points that didn’t fit what I thought was an obvious trendline. I found that there were several “era’s” of 2 3/4hp manufacture where the “filtered” gearbox serial number vs engine numbers data appeared to fit specific “lines of best fit”. These are summarised as follows;

1.   1907 to 1912 – the (non-linear) transition from single speed to early 2 speed gearboxes.
2.   1913 to 1914 – the pre-WW1 2 speeds.
3.   1915 to 1919 – to investigate WW1 production
4.   1920 to 1920 – investigating the “3 speed only” year.
5.   1921 to 1923 – reasonably stable post war production years.
6.   1924 to 1926 – Introduction of CW model and last gasp for the 2 3/4hp.

The last era appears to be non-linear and possibly multiple batches of 2 speed machines were produced towards the end to complicate matters. This is the period I am least confident of giving a reasonable representation. Many gearbox numbers not supplied for this era which does not help with sorting out correlations due to a lack of “valid” data points.

1907-1912

Next up we have the chart for 1907 to 1912. The first Douglas machines were single speed. I gather Douglas experimented with and gained racing experience with 2 speed gearboxes circa 1910 and introduce the first production 2 3/4hp models with 2 speed gearboxes in 1911.  I performed a little “educated guessing” with the shape of the line of best fit. More data points are required to confirm this relationship between the gearbox and engine numbers. But I think, in general, it shows the gradual uptake of the gearbox models as the buying public gained confidence in the “new” technology of the day. I assumed that the first 2 speed gearbox started with serial number 101.

[Hutch]

1913-1914

During 1912-14 we appear to have steady production of 2 speed gearboxes. Almost no single speed models sold by now so the relationship between gearbox and engine numbers appears reasonably linear. Note the gearbox serial numbers lag the engine numbers by about 3900 at the start of 1913 and about 4200 by the end of 1914 due to the earlier single speed machines and also more engines produced than gearboxes.

[Hutch]

1915-1919

For the 1915-1919 war years we again have a fairly linear “line of best fit” as expected from the linear trend of frame and engine serial numbers. The 3-speed for the 2 3/4hp was introduced in 1915, but to my frustration, no valid data points, that I could see, were to be found in the register (1915 Model W’s are very rare these days and also gearbox serial numbers may not have been submitted to the machine register etc.). Therefore, I only show 2 speed production in the chart.

In 1920 there is a 3-speed gearbox listed in the register with the serial number G3 / 159. This could be a typo’s, but it does sit  on the trend line for that year. If we assume 3 speed gearbox production started with the number “101” then this would indicate that there were not very many 3 speed machines produced from 1915 to 1919 or there is something else going on with the numbers that I cannot deduce due to lack of data? We do know that some 3 speed models were shipped from the factory before they were forced to only produce motorcycles for the war effort as shown by the following advertisement from The Examiner newspaper in Tasmania in May 1916. Could some of the early 1920 three speeds have been purloined from NOS from the war years and there is an overlap with the earlier gearbox serial numbers? I don’t know but it would have been easy enough in 1920 to stamp the prefix “G3” or “G6” etc before an existing serial number? More research is required in this area.

The gearbox serial numbers lag the engine numbers by a bit over 6500 at the end of 1919 again indicating more engines and frames produced than gearboxes. These extra frames and engines could have been made up of spares for the war effort plus some unknown quantity of 3 speeds as indicated above.

 Interestingly in 1919 when Douglas started to sell to the public again the production of frames and engines continued to outstrip gearbox production by about the same amount as in the period from 1915 to 1918. I have read several different accounts of when the Douglas factory was released from its war contracts and must do some more research to find out what was happening during 1919 in the post war Douglas production of the 2 ¾ hp models. Douglas was advertising 2 3/4hp machines for sale to the public in early 1919 but the approximate relationship between engine and gearbox production appears to match the war years – so I can only assume they continued to produce frames, engines and gearboxes in roughly the same ratio as between 1915 and end of 1918.

[Hutch]

From The Examiner 23 May 1916 (The actual picture in the advert appears to be a 1914 2 3/4hp 2-speed Douglas).

[Hutch]

1920

Douglas advertised that the 1920 2 3/4hp models were to be all 3 speeds. It was alleged that Walter Moore received a shilling for every 3-speed produced as a royalty payment for his patent on the 3 speed design (from Clew TBT page 42). Apart for its use in the 4 hp up to this year (Maybe about 10,000 gearboxes?), Moore’s 3 speed design had hardly been used in the 2 ¾ hp up to the start of 1920.
To help show what happened to the 2-speed production during 1920 I have included data from a few years before and after 1920. Were any 2 speeds produced in 1920? Maybe, but the 1920 3 speed data appears to indicate that for a period of time, every 3-speed was pretty well matched to an engine number i.e. same number of engines and gearboxes in that period were manufactured. Some more specific research is required to see if any 2 speed machines were manufactured by the factory in 1920. There is a lack of data for late in 1920, so I extrapolated the line of best fit to indicate approximately where the end of 1920 3 speed production might be. I then started the linear trend line for 1921 to 1923 3-speed serial numbers from this point. More data from this period would help greatly in confirming my assumptions about 1920 3 speed gearbox production.

[Hutch]

1921-1923

For 1921 to 1923 we see a fairly linear relationship between the 2-speed and 3-speed gearbox serial numbers and the engine numbers. If the numbers of 2 speed and 3 speed gearboxes manufactured are added together, we find that there are possibly about 3% more engines produced than gearboxes. Note that due to the large injection of 3 speed gearboxes into the production of the 2 3/4hp in 1920, the 2-speed gearbox serial numbers lag the engine number by quite a significant amount.

[Hutch]

1924-1926

With the data for 1924 onwards it becomes a little more complicated with the introduction of the CW. At the start of 1924 there appears to be a boost in the production of the 2 speeds and possibly a small drop off in the production of the 3 speeds. Due to lack of good data I have had a little trouble with the transition between the 2 3/4hp 3 speeds and the CW. Also, the production of the 2 3/4hp appeared to become chaotic towards the end of its life (the data being further exacerbated by the mix and match of machine parts during their life after leaving the factory). Anyway, this is my best stab at it, but there is a question mark on the trend lines of gearboxes vs engine serial numbers. It is possible that Douglas released the final dregs of the 2 3/4hp in batches with machines being made up out what was left in the stores but to date I am not confident in my analysis of the last year or so of production. Most of the production at the end is heavily biased toward the CW and then it appears a few remaining 2 3/4hp machines were manufactured, probably with whatever was left over in spare parts?

[Hutch]

Trendlines 1907-1926

If we collate these “era” charts together, we can now see my best guess at the overall trend of Douglas 2 3/4hp gearbox vs engine serial numbers from 1907 to 1926, including the CW, showing the data points that I used.

[Hutch]

Finally we have a chart of my “best guess” trendlines, based on the data to hand of the 2 speed and 3 speed 2 ¾ hp gearbox serial numbers vs engine numbers and have included the points that are the presently known to me to be accurate. Many parts of this research require more work and more known good data points will go a long way in validating this work. If you are aware of any very original bikes with well documented history please let the LDMCC club registrar know, (or let me know by PM and I will pass the information on). Please do not post your machine details online (unless you want to of course). I will update the charts as any new information arrives.

[Hutch]

Approximate Gearbox Serial Numbers vs Year of Production

Using the lines of best fit from the above charts I have generated the following table. This my best guess of gearbox serial numbers vs year of production and is very approximate only. More research and more accurate data on the relationship between the gearbox / frame and engine from production records and surviving machines with original frame / motor and gearbox intact and reported to the club registrar, are required for me to “tighten” up these results.

Two of the known good data points came from the production records, shown above, that I spotted for sale on ebay (no I didn’t buy them!). If anyone has any more of these records, I would like to hear from you, as even a single extra “known good” data point will go a long way to validating the trendlines.

[Hutch]

Conclusion

Please be careful not to read too much into this information as adding or removing one data point can make a significant difference in the trendlines due to lack of data points for some time periods. This data is very approximate only and it should not be used for machine dating purposes.

The information I used for this work has been used with permission from the Registrar of the LDMCC. I have no control at all over the accuracy of this information in the LDMCC Machine Register and this work is only my “first take” interpretations of that data.

Therefore, the above charts and information may only give a rough indication of the relationship between gearbox and engine serial numbers for the 2 ¾ hp machines and I cannot accept any responsibility for any errors that may occur through the use of this information for identifying or dating motorcycles or parts thereof. This work is definitely “work in progress” and there is much more research to be done.

Please provide constructive criticism and let me know about any faux pas, errors, omissions and clarifications etc and I will endeavour to fix it up where possible!

phew....I'm happy I have finally finished and posted this after starting it about 7 years ago. I have used it to match several  gearboxes to their respective projects - not quite exactly, but near enough for practical purposes! 

Thanks to Doug K. for his help on this project :-)

Cheers

Hutch

(9/09/2022 - various edits to the above have been performed since first posting this information to fix typos and improve layout)

[graeme]

Great stuff Ian, well done on collating this material. I'm sure it will be a great reference for the future interest in the marque

Cheers, Graeme