Research: Epiphone serial numbers and production estimates (1931–1956)
Author: Felix Wiedler (Version: March, 2016)Below a summary of my research findings regarding Epiphone's serial number (SN) systems and estimated production figures – after analyzing data of approximately 3'600 Epiphone instruments and amplifiers (SN/model pairs, documented in c. 38'000 photos). 1
1. Epiphone's SN systems 1931–1956
Epiphone used several different SN systems depending on instrument type and time period:
A. Acoustic instruments 1931–1943: SN starting at 5000 and ending around 20301. (SN on label or stamped inside body.)
B. Acoustic instruments 1943–1956, including electric hollowbody instruments 1950-1956: SN starting at 50000 and ending around 69637. (SN on label inside body.)
C. Electric instruments 1935–1942: SN starting possibly around 1 and ending around 7182. (SN stamped on headstock.)
D. Electric instruments 1946–1949: SN with model-specific prefix and serial suffix on most models. (SN stamped on headstock.)
E. Special electric hollowbody guitars 1949, electric Hawaiian guitars 1949–1956: Some hollowbody models from 1949 and Hawaiian guitars from 1949 onwards use "special" SN systems. (SN stamped on headstock or bridge unit.)
F. Amplifiers 1936-1956: Amps used a number of complex SN systems – partly showing model-specific SN ranges. (SN stamped on logo plate or control plate.)
Chapters 2, 3, and 4 primarily focus on SN systems A + B which are well documented and cover the vast majority of Epiphone's total instrument production. SN systems C, D, E + F are more complex in nature and are discussed separately (see chapter 5).2. Consecutive SN ranges assigned to model batches
Epiphone's main SN systems A + B generally appear to follow this pattern: a range of consecutive, ascending SNs are found on instruments of one single model; a subsequent range of SNs appears on instruments of another single model; and so on (see Fig. 1).
Fig. 1: Sample excerpt from Registry database.
My interpretation of this pattern is that these model-specific SN ranges represent production runs, i.e. each run or batch of a single model was assigned a range of consecutive SNs. I further assume that the subsequent SN range was assigned to the next following production run of a different model, and so on. This means: Epiphone's ascending SNs represent a chronology of production (not a chronology of shipping/sale). 2
I have observed considerable variation in batch sizes – from possibly a single, custom ordered model up to a hundred or more instruments of the same model in a consecutive SN range.3. Revised SN-to-year allocation
The findings stated in (2) in connection with other research (mainly from analyzing/comparing SNed instruments with dated documents) has led to a revised dating approach for Epiphone SNs. See Fig. 2: "W revised" (referring to Wiedler) vs. "F traditional" (referring to the charts published by Fisch/Fred).
revised – Wiedler
(approx. first SN)
|F traditional –
(approx. first SN)
When looking at SN-to-year allocations keep in mind: Instruments were not made in one day – the time span between production start and shipping/selling of an instrument was typically several months (sometimes more), often spanning different calendar years.
The "W” date of an Epiphone instrument/SN always refers to the estimated date when it entered production – and NOT the date it was finished, left the factory, or was sold by a store. It can be assumed that shipping/sale dates of individual instruments from the same production period (or even the same batch) could vary considerably – i.e. some selling quickly, while others remained unsold for months or longer.
This (partly) explains the differences to the traditional dating, which ultimately relied on Tom Wheeler’s research based on inventory lists of one single retail store. 3
Epiphone's SN systems A + B are pretty well documented with more than 2700 entries (SN/model pairs) in our Registry database. Taking into account the assumed lowest and documented highest number of these two SN systems, we can quite precisely estimate the total production of Epiphone instruments with SN inside body (=acoustic instruments 1931-1956 and electric hollowbody instruments 1950-1956): 34'940 units. 4
Based on my findings stated in (2) that ranges of consecutive SNs were assigned to batches of the same model, we can extend our research to another aspect: interpolation methods enable us to "fill in" undocumented SN/model pairs within assumed batches (i.e. ranges of SNs exclusively assigned to one model – see example in Fig. 3).
In many cases the very first and very last SN of a batch cannot be determined as long as there are SN gaps to the adjacent batch of a different model. However all missing SNs within the first and last documented SN of an assumed model batch can be "interpolated" i.e. tentatively identified (with high probability) to also be examples of that same model.
Applying this interpolation method to my registry data leads to some remarkable results: While my documented SN/model pairs currently represent about 8% of the estimated total production (SN systems A + B, 1931–1956), the addition of interpolated SN/model pairs boosts this ratio to 47% (see Fig. 4 for an excerpt).
The figures in column "Registry+interpolated" can be seen as "minimum" production numbers for the respective models, i.e. how many were "at least" produced. I consider these "minimum" figures as pretty reliable, although they are not to be mistaken as total production figures.
However, this data enables us to also calculate rough estimates of total production numbers for each model and period – by using approximation (extrapolation) methods. 5 Note: these extrapolated totals in the last column of Fig. 4 are ballpark figures and do not claim to be exact in any way. But they give an idea – e.g. if total production of a model was likely in the dozens, in the hundreds, or in the thousands.
Fig. 4: Example of estimated totals: acoustic
archtop guitars per model 1931-1956, plus totals of other instrument
types with internal SN (systems A + B). Model totals include cutaway
and 4-string versions (data as of 26 Mar 2016).
|Total El Hollowb '50s||483||2834||4917|
|Total Ac + El Hollowb '50s||2741||16527||34940|
|Percent Ac + El Hollowb '50s||7.8%||47.3%||100%|
The chart in Fig. 4 summarizes some of my current research findings – for example:
- The estimated total production (1931-1956) of instruments with an internal SN (systems A+B) – i.e. all acoustic instruments plus 1950s electric hollowbody instruments – is 34'940 units.
- Acoustic archtop guitars account for approximately 27'000 units – leaving all other instrument types far behind.
- Triumph and Zenith were the highest production models, with estimated totals of more than 5000 units each over the years.
- Other acoustic instrument types were a relatively small affair – approx. 2000 flattop guitars and 1000 mandolins in total.
- In the 1950-1956 period, electric and acoustic instruments were produced in similar numbers, around 5000 units of each type in total.
- Note: For low-production models (e.g. Tudor, Empire, Windsor, some flattops, etc.) the extrapolated totals are possibly distorted, especially if the (few) documented examples are surrounded by larger SN gaps. This is due to the fact that the applied extrapolation algorithm is the same for all models, i.e. not accounting for any statistical frequency. 5
The chart in Fig. 5 shows an example of model-related data on a timeline – for the pre-war Emperor:
- Documented in our registry database are 64 Emperors with a SN from the pre-war period.
- By adding interpolated SNs (from within assumed batches, as explained in Fig. 3) we can conclude with high probability that at least 213 pre-war Emperors were produced.
- Adding extrapolated SNs 5 (from the gaps before the first and after the last documented SN of assumed batches, as explained in Fig. 9) results in a ballpark estimate of around 373 pre-war Emperors in total.
- The SNs appear to be grouped in 11 batches (plus one batch of Soloist cutaway models); pre-war production of the Emperor seems to have peaked around 1939 – 3 batches with a total of possibly more than a hundred instruments built in a year.
The chart in Fig. 6 shows an example of a comparatively high interpolation ratio due to registry data of excellent quality:
- Acoustic instruments of 1947 (W year) are currently assigned to SN range 55850–57099 = 1250 units.
- According to our data, 5 different archtop models were produced in this SN range – grouped in 10 model batches: 3 batches of Blackstone, 1 batch of Broadway, 2 batches each of DeLuxe, Spartan + Triumph.
- Based on the data of only 95 registered SN/model pairs, interpolation allows for reliable identification of 1024 SN/model pairs = about 82% of the estimated 1250 SN/model pairs in that range/period.
- About 224 SNs are placed outside assumed model batches and can not yet be reliably assigned to models. However estimated model totals can be calculated by extrapolation methods (see column "Reg+int+extrapol"). 5
- Broadway batch SN 56342–56479 is a good example to demonstrate our extrapolation method: Currently there are 15 unidentified SNs in the gaps to the adjacent model batches – 3 SNs before the first and 12 SNs after the last identified SN. The actual number of Broadway models could be any number between 15 (all) or 0 (none). Our approximation (extrapolation) algorithm calculates 8 additional Broadways (50%=7.5, rounded to 8). 5
Although these estimates are still somewhat speculative and approximate at best, they are certainly based on a much wider data sample of higher quality than that on which previous research was based. And the estimates will keep getting more precise as the database of documented SN/model pairs is growing.
While Epiphone's SN systems for acoustic instruments and 1950s electric hollowbodies (A + B) appear to be pretty straightforward, the SN systems of their other electric instruments and amps (C, D, E + F) are much less so. Over the years a number of different SN systems were used. Below an overview of the groups C, D, E + F:C. Pre-war period 1935–1942: SNs of electric stringed instruments cover the range from possibly around 1 to around 7182, their ascending numbering representing a chronology of production (Note: The earliest electric instruments from 1935 don't bear a visible SN).
Similar to SN systems A+B, ranges of consecutive SNs appear to be assigned to groups of single models – suggesting that these represent production runs. The respective SN range suggests that the pre-WW2 production of electric instruments sums up to more than 7000 units in total – if every SN was indeed assigned to a manufactured instrument. During WW2 electric instrument production was virtually halted.
D. Post-war years 1946–1949: When electric model production was relaunched in 1946, new ranges of SNs were applied to most models – with numbers based on a model-specific prefix (2-3 digits) and serial suffix (3 digits, ascending consecutive numbers starting with 000): 15xxx = Century Hawaiian, 25xxx = Zephyr Spanish, 60xxx = Century Spanish, 75xxx = Zephyr DeLuxe; 85xxx = Zephyr DeLuxe Regent, 100xxx = Console. This system allows for pretty precise estimates of production figures in this period: the highest (known) SN suffix of a model indicating the total number produced – summing up to a total of 2248 units.
An exception (and not included in this figure) is the Zephyr Hawaiian model which continued with the pre-war SN system (see C), starting around SN 7307 and ending around SN 7908 (i.e. approx. 600 units in total). The Alkire Eharp, an electric Hawaiian model custom built for and exclusively sold by musician Eddie Alkire, had its own SN system starting around 100.
E. Special electric hollowbody guitars 1949, electric Hawaiian guitars 1949-1956: The year 1949 brought significant changes to Epiphone's model lineup and likewise to their SN systems. Early examples of the newly introduced Kent Spanish model have "special" SNs in the 3000s and 4000s (stamped on headstock), before switching to SN system B by 1950. Documented examples of the new Kent Hawaiian show "special" SNs in the 75 to 171 range (stamped on headstock). In the early 1950s all electric Hawaiian guitars switched to SNs in the 9000s (stamped on bridge unit). Due to the yet unclear logic behind the applied "special" SN systems, production totals for these models are difficult to estimate.
revised (approx first SN)
|1946–1949||7300 Zephyr Hawaiian
15000 Century Hawaiian
25000 Zephyr Spanish
60000 Century Spanish
75000 Zephyr DeLuxe
85000 Zephyr DeLuxe Cutaway
100 Alkire Eharp
|1949||3000, 4000 Kent Spanish
75 Kent Hawaiian, Century Hawaiian
|Ac + El Hollowb '50s||2741
|El Hollowb pre-1950||294||2789||4088|
F. Amplifiers 1936–1956: Amps used a number of different SN systems which are not yet fully understood and make it difficult to estimate production totals (Fig. 9). Therefore I don't present total production estimates for amps at this stage of my research.
The SN system of pre-war amplifiers is based on ascending numbers, however it is questionable if all numbers in the respective ranges were indeed assigned to manufactured units. 6 In some years model-specific SN ranges were used. After 1947 the amp model lines underwent several changes, each reflected in a new SN system.
Fig. 9: SN-to-year allocation: Amplifiers, with model-specific SN ranges in some years (SN system F).
/ models (approx first SN)
3000 Model M
9000 Century, Coronet
10000 Century, Coronet
|1946||8900 Zephyr, Dreadnaught|
|1947||10000 Zephyr, Century|
|1948||30000 Zephyr, Dreadnaught|
General note: My estimated SN-to-year allocations and production figures are approximations that are subject to correction as new evidence material surfaces.
1) Acoustic instruments generally refer to guitar and mandolin family instruments. Epiphone banjos and bass viols (which had their own SN systems each) are out of the scope of my research.
2) Epiphone's SN systems A + B appear to be similar in
concept to the SNs of C.F. Martin & Co: Each model batch was
assigned a consecutive SN range – with ascending SNs representing a
chronology of production. Most Gibson SNs (pre-WW2) on the other hand
appear to relate rather to the completion date (shortly before
shipping) of an individual instrument; for their batch-oriented
production Gibson used a second numbering system – the Factory Order
Numbers (FONs) –, as pointed out by Joe
Spann's invaluable research published in: Spann, J. E.
(2011), Spann's Guide To Gibson 1902-1941.
Note that in this context the term "batch" is used for a series of instruments within a SN range exclusively assigned to one single model, although technically speaking a larger series of one model may actually consist of several consecutive production batches of the same model.
3) Wheeler, T. (1982), American Guitars: An Illustrated History, p.40. Tom Wheeler based his Epiphone SN dating on records of a single music store, Pettey Music Co in Pittsburgh, PA. This fact suggests that Wheeler's SN-to-year allocation related to the date when an Epiphone instruments was present at that store – not when it went into production. Furthermore, Wheeler's figures seem to suggest that the initial digits of a SN coincide with certain calendar years. No evidence has been found to support such a theory.
4) Generally, our estimates are based on the assumption that all numbers in the respective SN range were actually assigned to instruments. Theoretically there is a possibility that certain numbers were omitted and not used for reasons unknown. Certainty in this matter will grow as the gaps of undocumented SNs in our Registry continue to be filled with data.
5) Extrapolation methods tentatively attribute undocumented SN/model pairs in gaps between two (assumed) adjacent model batches. In the example in Fig. 10, the 6 undocumented SNs 55236–55241 are likely to include models of the previous batch (=Triumph) and/or subsequent batch (=Blackstone), although their relative distribution is not known. Our applied extrapolation algorithm equally assigns 50% of the missing SNs to the previous model and 50% to the subsequent model – in this example 3 Triumphs and 3 Blackstones. Note: Theoretically, such SN gaps could include a small batch of a third model. In our database there are still quite a few larger gaps (>50) with no documented SN so far; we don't know today for sure if such gaps consist of models of the adjacent batches, or include undocumented batches of third models – or even in some cases were omitted (see note 4).
In an earlier stage of our research we considered the theory of one shared SN system between electric stringed instruments and amplifiers (as opposed to two separate systems of the same number range): This because no case of an identical SN on units of each type was documented – until 2016. Today we know 2 cases of an identical SN on units of each type (SN 1342: Model M Spanish + Model M Amplifier; SN 5220: Model M Hawaiian + Zephyr Amplifier). Based on these cases of double SNs we revised our theory: We assume that electric stringed instruments and amps used separate SN systems.
Carter, Walter, and Jimi Stratton. Epiphone: The Complete History. Hal Leonard Corporation, 1995.
Carter, Walter. The Epiphone Guitar Book: A Complete History. Hal Leonard Corporation, 2012.
Fisch, Jim, and L. B. Fred. Epiphone: The House of Stathopoulo. Amsco Publications, 1996.
General or other makers:
Gruhn, George, and Walter Carter. Acoustic guitars and other fretted instruments: a photographic history. GPI Books, 1993.
Gruhn, George, and Walter Carter. Gruhn's guide to vintage guitars: an identification guide for American fretted instruments. GPI Books, 1991.
Gruhn, George, and Walter Carter. Gruhn's guide to vintage guitars: an identification guide for American fretted instruments. Updated and revised 3rd edition. Hal Leonard Corporation, 2010.
Johnston, Richard, and Dick Boak. Martin guitars: A history.
Vol. 1. Hal Leonard Corporation, 2008.
Johnston, Richard, and Dick Boak. Martin Guitars: A Technical Reference. Vol. 2. Hal Leonard Corporation, 2009.
Spann, Joseph E. Spann's Guide to Gibson 1902-1941.
Centerstream Publications, 2011.
Wheeler, Tom. American guitars: an illustrated history. Harper & Row, 1982.