Geoffrey Sinclair wrote in message ...

Reprise,

"Francis Dean in America's Hundred Thousand has a table of the 8
main WWII fighter types, by numbers produced and I am wondering
about the differences he has with the USAAF Statistical digest."

I contacted his publisher and was informed Mr Dean sadly died last
year, so that ends the chance of consulting with him over the production
figures he presents. His book is clearly on the technical aspects of the
various fighter designs and his production figures have a few errors in
them, mainly minor.

I posted to the newsgroup about the P-39 and P-40, here is a follow up.

When in doubt reach for other references to try and resolve the
differences, so I have,

USAAF Statistical Digest, or SD
America's Hundred Thousand by Dean
The American Fighter by Angelucci and Bowers, or TAF
The Complete book of Fighters by Green and Swanborough or TCBOF

TAF is good because it includes the USAAF serial numbers by
version produced of each fighter type presented.

P-39

The raw total production figures given are, in order,

9,528 Dean, list by versions of the P-39, excluding prototype
9,547 Dean, production numbers table, excluding the prototype
9,558 TCBOF, including prototype
9,585 TAF, probably excluding the prototype, the total given in the text
9,588 SD, excluding the prototype
10,016 TAF, adding up the serial numbers, prototype excluded.

The number of P-39s produced has the complication of the P-400
or Airacobra I production that did not receive USAAF serials, most
of which ended up in the USAAF as the P-400 or in the USSR,
some 675 aircraft. There were also significant losses at sea of
P-400/Airacobra I and P-39

The Dean list by versions has 863 D (37mm cannon) D-1 (20 mm
cannon for USAAF) and D-2 (20 mm cannon for RAF) in total. TAF
notes there were 60 D, 705 D-1 and 158 D-2, so 863 D-1 and
D-2, with 60 more D models than Dean, the difference between
Dean and the SD is 60.

As for the serial number list the difference is the result of serials
being allocated for one version but built as another, and so the
serials are reported twice. Mainly some 369 D models, when
the D was reordered as the D-1 also 35 N-3 models that are
also reported under the N-1 list.

The rest of the difference appears to be changes when the second
order for 394 D models serials 41-6722 to 41-7115 was changed
to D-1s. It appears of the original order 41-7053 to 7056 and
41-7059 to 7079 were not built as D, D-1 or D-2s.

Instead 41-7053 to 7076 were built as J models, with 41-7057
and 41-7058 being reported as both D-1 and Js. Most probably
the serials for the D-1s should be the two "missing" ones, 41-7077
and 41-7078, that is the serials from the original D order TAF does
not list against another version. If this is correct then all the serials
in the original D order are used by various P-39s.

Take away the duplicates, 369 D models, 24 J models and 35 N-3
models and the serial numbers total becomes 9,588 excluding the
prototype or the same as the SD.

So the Dean production table figures for the July to December 1942
period for the P-39 are incorrect as they give a figure of about 40
less than were built, also the list by model type is missing 60 D models.

P-40

The raw totals are, in order,

13,143 Dean production table, excluding the prototype
13,738 SD, excluding the prototype (same total for TAF in the text)
13,755 TAF adding up the serial numbers allocated, prototype excluded
14,047 Dean list by versions of the P-40, excluding the prototype

The P-40 list traces are complicated by the fact not all production was
given USAAF serials, the pre lend lease orders by France and Britain,
that eventually had some 1,740 aircraft built to fill. Also the initial French
order for 185 was cut to 140 by the British when they took it over. Then
add the diversions of aircraft from these orders and Lend Lease
allocations to the USSR, the AVG, Egypt, Turkey, Free French and the
aircraft lost at sea. Even some of the non lend lease British aircraft ending
up with the USAAF.

The difference between the SD and TAF serial numbers is simply the
17 serials TAF gives for the P-40G, all P-40Gs were conversions and
the serials also appear in the P-40F list. Take 17 away from 13,755
and you are at 13,738, then add in the prototype.

Something similar happens in the Dean list by versions, the P-40Gs
are counted as production not conversion and some of the lend
lease allocations are double counted, so subtract 44 P-40G and
250 Kittihawk IIs and you end up at 13,753. Most of the rest of the
difference is Dean stating there were 4 less P-40N-40s, the last
version, than the serial number lists.

So the Dean production table has a major error in the June, July and
August 1942 P-40 production figures, resulting in a total of nearly 600
less than the correct figure. Deans list by model type goes the other
way, double counting some production. The USAAF statistical digest
would appear to be the correct source for monthly production figures
and this also applies to the production figures presented for the other
types in the book.

Geoffrey Sinclair
Remove the nb for email.