There are two public listings of Spitfire histories by serial number. The one in Spitfire by Morgan and Shacklady and the one at
home
A consistent definition of produced versus converted is required, like what version the airframe first flew as. So an airframe sent to the experimental section, modified to a new version before first flight is counted as the new version, an airframe test flown then sent for experimental work is counted as the original version. Ignoring Seafires the online Spitfire list as around 818 airframes undergoing at least 1 conversion, 81 undergoing 2, 18 undergoing 3 and 1 undergoing 4, if I have my counting correct.
Apart from what follows, the RAF Contract Cards and Serial Registers are sources.
British figures do not usually report experimental types, these are mainly the known prototypes but also include other airframes diverted on what were called CRD (Controller of Research and Development) contracts. In addition the treatment of aircraft that crashed during testing was very case specific, sometimes they were counted, at other times they were not, the decision could take quite a while and could even be reversed. If the review decided the aircraft was to be counted the cumulative totals would be adjusted but usually not the relevant monthly report. As a further complication the pre war figures mostly include the non prototype experimental aircraft and also any losses during testing, making such exclusions mainly a wartime change.
The period January 1940 to February 1942 inclusive is where the existing aircraft production documentation is weakest and it requires using a number of reports to build monthly production by version and as expected there are some disagreements between the reports.
National Archives file AIR 8/278 covers production from September 1939 to December 1940 inclusive, with both monthly and weekly figures. Imports and exports are included but not version information. The monthly reports for September 1939 to April 1940 inclusive and November and December 1940 give factory information, as do the weekly figures, except for Tiger Moths. There are three times the end of week is also the end of month, 30 September 1939, 31 August and 30 November 1940 and the weekly figures have been used to give a best indication of production by factory by month May to October 1940 if no official monthly figures are available.
The National Aerospace Library at Farnborough has a copy of the Ministry of Supply report, Rate of attainment of peak output after first delivery, report no 5. 11/313. It gives the first 15 to 18 months of production for a given factory and a given aircraft type, all up 25 factories. Lancaster at A V Roe, Armstrong Whitworth and Vickers Castle Bromwich, Halifax at English Electric, Handley Page and Fairey, Wellington at Blackpool, Hampden at English Electric, Mosquito at deHavilland, Standard and Percival, Beaufighter at Bristol Weston, Barracuda at Boulton-Paul and Fairey, Blenheim at Rootes, Tempest at Hawker, Typhoon at Gloster, Battle at Austin, Oxford at Standard Motors and deHavilland, , Defiant at Boulton-Paul, Vampire at English Electric, Hurricane at Gloster, Spitfire at Castle Bromwich and Supermarine. Most relevant to 1940 are the Wellington, Hampden, Hurricane, Spitfire, Blenheim, Oxford and Battle figures.
National Archives file AVIA 46/146 is another report on time to peak production. Much of it is hand written in pencil and ink and there are plenty of corrections and changes. It has Mosquito production by factory by mark July 1941 to October 1943, it also has the first 6 months planned and actual production for Spitfire I, Vb, IX, XII, XIV all at Supermarine, Whirlwind, Typhoon at Gloster, Beaufighter I and II at Bristol Filton, VI at Bristol Westland, Mosquito II and VI at deHavilland, Welkin, Tempest V and Hurricane I and II at Hawker. In addition there is a table giving production from the beginning to peak production for the parent firm but without mark information. The table reports on Armstrong Whitworth Whitley March 1937 to December 1940, Bristol Blenheim January 1937 to March 1940, Vickers Wellington at Weybridge, October 1938 to March 1942, Bristol Beaufort I October 1939 to March 1942, Short Stirling May to June 1943, Handley Page Halifax October to July 1943, Hawker Hurricane December 1937 to December 1940, Supermarine Spitfire July 1938 to December 1940, Boulton-Paul Defiant September 1939 to December 1941, Bristol Beaufighter June to July 1943, Avro Lancaster October 1941 to December 1943, Gloster built Typhoon June 1941 to June 1943, deHavilland Hatfield Mosquito September 1941 to December 1943, Fairey built Barracuda July 1942 to September 1944. Finally there are the planned and actual production for the first 6 months for Bristol Blenheim I, Rootes built Blenheim V, Vickers Weybridge built Wellington I and II plus Blackpool built mark III and X, Armstrong Whitworth Whitley I, Handley Page Halifax I and II, Short Stirling I, Avro Lancaster I, Vickers Warwick, deHavilland Mosquito B.IX and B.XVI. The report ends with a mainly typed table first 6 months of planned and actual production for Blenheim I and V, Wellington I, II, III and X, Whitley I, Halifax I and II, Stirling I, Lancaster I, Warwick, Buckingham, Mosquito II, VI, B.IX and B.XVI, Spitfire I, Vb, IX, XII, XIV, Whirlwind, Typhoon, Beaufighter I, II and VI, Welkin, Hurricane I and II, and Tempest V.
The Ministry of Aircraft Production (MAP) monthly Statistical Bulletin includes a table of production by firm, the early versions of the table starting in January 1941, it means for example both Bristol factories are reported as one total. For the period January to June 1941 most entries are total aircraft production by the firm but from July 1941 there is more of a breakdown, for example Austin production of Stirling and Hurricanes is separated.
National Archives file AVIA 10/315 is the MAP Weekly Statistical Bulletin November 1941 to June 1942, the figures are by factory and mark, beginning with the week ending 8 November 1941 but also including cumulative totals from 1 September 1941. The cumulative totals are changed to starting 1 December 1941 during December. The end of the week and month match on 31 January and 28 February 1942, which means the report gives production by mark and factory in 3 groups, 1) September to November 1941, 2) December 1941 and January 1942, 3) February 1942.
As can be seen by the details AVIA 46/146 can be repetitive, it does provide main factory information and at least the initial production of new marks in 1940 and 1941. The total factory information from the MAP figures can help separate out the individual factory figures and AVIA 10/315 gives more detail for September 1941 onwards, including some cases where the monthly figures can be directly calculated. Finally a Vickers document MS412 which gives monthly Spitfire production by version and factory for both Supermarine and Castle Bromwich, though it does not match the official production figures until mid 1943, as well as the RAF contract cards, can be used as a guide. The production figures for 1941 reported here are the results of analysis of all the above reports.
Cumulative production to 31 December 1945, types in production in September 1939 or later. From MAP Statistical Review
1,547 Spitfire I
750 Spitfire IIa
170 Spitfire IIb
100 Spitfire VI
The online listings give
1543 Ia (including 1 for France, no RAF serial)
3 Ib
1 HS (High speed modified mark I)
3 pattern (to Castle Bromwich, never flew)
765 IIa
156 IIb
100 VI
I ended up concluding
1,548 I, including the HS and French aircraft, excluding the pattern ones, this agrees with the RAF Census total.
751 IIa, RAF Census says 750
170 IIb, agrees with RAF census
100 VI including X4942, RAF census has 99 production and 3 conversions
Contract cards for the initial Castle Bromwich order for 1,000 aircraft confirms 921 II and 79 V built.
Mark II a and b, RAF census in June 1944 says 750 a and 170 b. Comparing the individual aircraft histories, the contract cards and the serial registers, the latter two being scanned microfilm and so hard to read at times, the following emerges. Around 179 serials have at least one reference claiming they are a IIb, of these 151 are in all three references and 19 are in two, with another 9 only in 1 reference. Aircraft with only one mention have been discarded, leaving 170. Given the difficulty in reading some entries the list cannot be definitive.
The monthly reports say 89 VI built to end September 1942, then another 8 in October giving a total of 98 (which is 9 more), then 2 in November giving a total of 100. Somewhere along the line a VI was missed or wrongly reported.
R6722, the floatplane, "not yet accepted off production 5 May 1941", counted as mark V.
X4708 mark I, crashed under test 5 November 1940, believed not officially counted.
X4942 counted as V prototype, ordered as mark I
X4942 counted as VI prototype, ordered as mark I