Andrew Clark wrote:
> You are quite right that 27,000 Territorial troops were evacuated from
> Dunkirk between 19/20 May and 25/26 May. But evacuation had begun
> before that date, not under Gort's order but that of the LOC
> commanders and the Admiralty, who from 14 May onward had been using
> the cross-Channel shuttle of routine supply ships to remove "useless
> mouths", such as railway specialists, pioneer battalions, engineers,
> base depot staff and so on. Most of the Admiralty preparations
> for Dunkirk date from 14 May.
>
> The figure of 90,000 thus relates to the period 14 May to 25/26 May
> and is from PRO ADM199/82, quoted in "Dunkirk: The Necessary Myth" (N
> Harman, Hodder & Stoughton, London 1980),
Thank you for the reference. I haven't yet located that book, but from
an Internet search I gather that it is primarily about how the British
turned Dunkirk into a "victory" (accounts of which are available
elsewhere, and I have no problem with that), so perhaps you might
explain the discrepancy between other accounts ?
Here's what I have, and as you will see the figures don't match.
1. "By the end of April the strength of the British Army in France had
increased to 394,165" (Ellis, "The War in France and in Flanders, p.19)
of which 264,035 were "real army" troops (BEF GHQ, corps or division
troops, territorial divisions, drafts, unallocated troops and
reinforcements held at bases) and 130,130 rear-area personnel (troops
LOC duties, not allocated, in various HQs, hospitals etc, in territorial
divisions, or with the AASF).
2. "In the evacuation of non-fighting troops already in progress the
Navy had begun by using Boulogne, Calais, Dunkirk and Ostend, and by
midnight on the 26th of May they had brought to England 27,936 men who
were no longer needed in France" (Ellis p.181, which sources to "A.L.
58, Gort's Despatches, para.49, 9.5926; B.E.F./G.H.Q. Files/Box 2/3,
p.102).
3. So according to Ellis (p.246-248), 198,315 British troops were
evacuated as part of Dynamo, plus another 26,402 before then, for a
total of 224,717, including 13,053 casualties (plus 141,445 non-British
troops which don't necessarily concern us here).
4. In addition to these, other British troops were evacuated after
Dunkirk. Ellis (p.305) provides a figure of 144,171. Robin Brodhurst
("The Battle for France & Flanders 60 years on", p.137) quotes the same
figure (both use Roskill as their source) and adds 30-40,000 "other
British subjects", mainly civilians evacuated from southern French
ports, and 22,656 from the Channel Islands between 19 and 23 June.
5. So between them, operations Dynamo and Ariel brought the total of
British troops evacuated to 368,491. Compared to the initial figure of
394,165, this means that 25,674 are unaccounted for, not counting
British troops despatched to France in May-June (more on this below).
According to Ellis (p.263), "140,000 British troops were still" in
France after Dunkirk so we are short some 30,000 troops compared to the
figures in point #1, not counting later reinforcements: same situation.
British troops arriving in France were the 1st Armoured and parts of
52nd infantry (Beauman) division, let's assume they were at full TO&E
strength to account for additional non-divisional assets so that's some
25,000 additional troops. Casualties were 68,111 of which most were
POWs. The Germans had 43,000 British POWs in May 1941 (source: OKW
archives quoted by Dr Arzalier in "La campagne de 1940" by Levisse-Touze
(ed) p.429), so given a quick look at RAF air activity between June 1940
and May 1941 let's say 1,000 of the POWs were RAF crews captured after
the fall of France (this looks generous) so that's 42,000 POWs and
according to the usual ratio a third of the rest were KIA's, probably
more but let's be safe.
6. So to sum up, the nice thing about "there were no 90k troops
subrepticely evacuated before May 26" scenario is that the numbers
match: some 419,165 British troops were shipped to France, and we have
an estimate of 419,194.66 (probably a cripple !) who were either
evacuated, in a POW camp or dead.
By contrast, the Harman theresis means that Harman, and perhaps the
Admiralty, disagree with point #2 above which means that unless I'm
mistaken, one of the following is true (or a combination of both):
a/ either 60,000 RAF/RN (in any case, non-Army) personnel were evacuated
prior to 26 May,
b/ or some 30,000 personnel (all arms) was evacuated between 14 and 19
May despite Ellis claiming that army evacuation began on 20 May.
The first hypothesis seems the most likely, although I find the numbers
very high no matter what. We're talking of a daily evacuation rate of
some 4,850 men in addition to the effort recorded in Ellis which amounts
to 4,000 men daily. In other words, the army gave the word to evacuate
non-essential troops on 20 May but only managed 84% of an unofficial
RN/RAF effort which had began a week earlier. For further comparison, a
table in Churchill's "Their Finest Hour" (chapter "the deliverance of
Dunkirk", p.102 of my edition), the Dunkirk operation lifted 7,700
troops on 27 May, 17,800 on 28 May, 47-68,000 29 May - 1 June, and some
26,000 daily for 2-4 June, these showing what the full British capacity
was, not counting the fact a fair bit of the RN was tied up in Norway
and it took time to bring it back to Dunkirk.
The second hypothesis is that Ellis is correct about the 20-26 May
figures, but an additional 62,000 personnel had been evacuated between
14-20 may, an daily average of 8,900 men.
Both hypotheses suggest a constant British sealift capability of some
8,900 troops daily starting on 14 May. Even counting the - very low -
number of Dutch and Belgian troops evacuated from the ports, plus the
negligible number of French troops evacuated on French vessels prior to
25 May (the number remaining very small until Dynamo started), I think
that the figure is really high. Additionally, I've seen mention of
contingency planning as early as May 14-15, in the form of gathering
ships in the Dover area, but we're talking about a dedicated sealift
effort here.
All of this to say that this sort of thing takes a ridiculously long
amount of time to look up, mostly because a lot of the sources reference
each other (which is why I'm using Ellis, but although I could have
sworn that my copy of Roskill had been returned I can't find it so I
didn't look it up), and I'd appreciate some clarification if you can
find the time.
LC
