The Peenemunde Raid, 1943: Operation Hydra and the War Against the Rocket

The Peenemünde raid, Operation Hydra, was flown on the night of 17/18 August 1943. That night the northern tip of Usedom lay under a waning gibbous moon, the Baltic silver beyond the pine belt, the sand flats and the reed edges of the Peenemünder Haken picked out as clearly as they would be at dusk. Below that moon stood the most secret industrial landscape in Europe: an army research centre, a Luftwaffe test establishment, a power station, an oxygen works, a half-finished assembly hall the size of a small town, a model housing estate at Karlshagen, and, a little to the south, wooden hutted camps holding foreign workers who had no shelters worth the name. Shortly after midnight, red spot fires began to fall. What followed was the first occasion on which a single officer circled a German target for half an hour and talked a whole main force onto it by radio, a method its author had rehearsed over Turin ten days before; it was also a raid whose planners had set out, in writing, to kill the men who were building Germany’s long-range rocket, and whose opening markers went down about a mile and a half past the aiming point, over a labour camp holding foreign and German workers. Forty bombers did not come back. More than eighty years later the argument about what Operation Hydra achieved, and about what it cost the people on the ground, has never settled.

This is a story that begins not with aircraft but with paper: an anonymous seven-page report posted to a British naval attaché in Oslo in November 1939, a scatter of prisoner interrogations, a smuggled sketch map, and a series of vertical photographs in which a handful of intelligent, exhausted people at Medmenham stared at objects a millimetre and a half long and tried to decide whether they were looking at the future. It runs through a Whitehall quarrel between men of formidable ability who could not agree whether a rocket the size of a lorry was a physical possibility. It ends over a strip of Baltic sand with a Lancaster called W for William circling in bright moonlight while its captain, Group Captain John Searby, corrected the aim of hundreds of aircraft that were, in most cases, unaware of what they had been sent to destroy.

Hydra sits awkwardly in the literature because it will not resolve into a simple verdict. It was, by the standards of 1943, a precise operation flown at night by a force whose ordinary business was area bombing. It was also an operation in which the deliberate killing of civilians, specifically of scientists and engineers and their families, was written into the plan, and in which the people who died in the largest numbers were among the least free human beings in Germany. It produced one clear tactical innovation that Bomber Command used for the rest of the war, and one German innovation, deployed against the last wave, that would cost the Command dearly over the following winter. What it did to the rocket programme is disputed to this day, and the range of scholarly opinion runs from “highly successful” to “not effective”. The proper way to tell it is slowly, from the beginning.

The island: what Germany built at Peenemünde

Usedom is a long, low, sandy island lying across the mouth of the Oder lagoon, shared today between Germany and Poland, its Baltic shore a chain of Wilhelmine seaside resorts and its interior a mixture of pine, heath and shallow water. The northern tip, the Peenemünder Haken, is the emptiest part of it: flat, marshy, screened by trees, and, in the middle 1930s, virtually uninhabited. That emptiness was the whole point. Major Walter Dornberger’s small rocket section of the Army Weapons Office had outgrown its proving ground at Kummersdorf, south of Berlin, where the noise and hazard of firing large liquid-fuelled motors could no longer be contained. German reference accounts record that the site search settled on Peenemünde at Christmas 1935, and that the attraction was geometry as much as secrecy: from the Haken a rocket could be fired north east, out along the Pomeranian coast, and tracked from the little island of Ruden and beyond, over a range of some 400 kilometres of open water where nothing that fell would kill anybody who mattered to the Reich.

The acquisition was on the same scale as the ambition. The Reich Aviation Ministry is generally said to have bought the entire northern peninsula from the town of Wolgast on 2 April 1936, for 750,000 Reichsmarks, and construction began that summer; the Heeresversuchsanstalt Peenemünde was formally constituted in 1937 as one of five army proving grounds under the Heereswaffenamt, and the first ninety or so staff came up from Kummersdorf in May of that year. Those figures rest on German encyclopaedic accounts rather than on archival citation, and are best treated as the standard version rather than as documented fact. What is not in doubt is the scale of the investment that followed. Something in the order of 550 million Reichsmarks is the usual estimate for the years between 1937 and 1940, an extraordinary commitment of capital to a weapon that did not yet exist, could not yet be aimed, and had never successfully flown.

The complex divided into three. Werk Ost, the army development works at the very tip, held the test stands and the laboratories. Werk West, formally the Erprobungsstelle Peenemünde-West, was the Luftwaffe’s own establishment with its airfield, a separate empire under a separate service. Werk Süd, to the south, was the production area, dominated by Fertigungshalle 1, an assembly hall of some 248 metres by 120 metres built by Dyckerhoff und Widmann and described as one of the largest column free halls of its day; it was completed in the spring of 1943 and had barely begun work. At the northern end stood Test Stand VII, the principal launch installation for the A-4 rocket, run by Kurt Debus: an elliptical sand wall some 670 yards around, a concrete lined flame pit twenty feet deep with three foot walls, an adjoining assembly hangar, and a blockhouse fitted with periscopes and a countdown clock. It also carried what is described as the first closed circuit television installation used to watch a rocket launch, a detail that says a great deal about how far ahead of its time the place was.

Everything that a rocket range consumes had to be made on site. A coal fired power station, begun in late 1939 by the power station division of Siemens Schuckert and brought progressively into service by 1943, was designed for 33 megawatts, according to the museum that now occupies it; German reference works give 30 megawatts, of which 20 to 25 went to the oxygen works alone. The liquid oxygen plant produced its own supply by the Claude Heylandt process from the summer of 1942, and it needed to: each A-4 launch swallowed four tonnes of liquid oxygen. The power station supplied heat as well as current to the works, ran on into the postwar decades, and survives as the largest industrial monument in Mecklenburg-Vorpommern. Standing beside it now, it is difficult to avoid the thought that this was less a laboratory than a small state, with its own electricity, its own cryogenics, its own railway and its own population.

That population was housed according to a strict and deliberate hierarchy, and the hierarchy is the moral architecture of everything that happened on the night of 17 August. For the German staff there was Karlshagen, a purpose-built works settlement, with housing put up for around 5,000 residents, its own railway connections and its own air raid protection. The museum at Peenemünde records what that protection consisted of: large bunkers on the test site, and, around the settlement, numerous Splitterschutzröhren, splinter protection tubes of about a metre and a half internal diameter fitted with wooden benches and a concrete blast porch, together with Splitterschutzgräben, trenches with reinforced sides and roofs of thick oak beams, many of them dug by the residents themselves. They were proof against splinters and blast but not against a direct hit. Because they stood close to the houses, the museum notes, they saved many residents’ lives in August 1943.

For everybody else the arrangements were different. The workforce included something like 1,400 forced labourers and more than 3,000 Ostarbeiter from Ukraine, alongside Czech, Dutch and Italian contract workers and French civilians. At Trassenheide, south of the works, stood a Deutsche Arbeitsfront communal camp of 42 barracks holding around 4,000 people in 1942 and 1943; English language accounts of the raid usually describe it as 45 huts, of which 23 were destroyed, and the discrepancy in the hut count is one of many small unresolved differences in this story. In 1943 two concentration camp satellite camps were established, each of several hundred inmates, one at the Luftwaffe establishment and one at the army research centre. Prisoners from the army camp were assigned to rocket assembly in Fertigungshalle 1 and housed in the building’s basement level, perhaps 600 of them, sleeping in group rooms of a hundred to a hundred and fifty men in three-tier bunks, working an eleven-hour day, six days a week, on inadequate rations, defenceless against the SS and the kapos. The museum documents 248 deaths of concentration camp prisoners at Peenemünde between 1943 and 1945. Those men had no splinter tubes.

Dornberger and von Braun

Two men ran the rocket. Walter Dornberger, a professional gunner and an engineer by training, headed the rocket division of the Army Weapons Office from July 1935 and was the military director of the programme: the man who found the money, fought the departmental battles, managed the relationship with the Heereswaffenamt and, later, with Speer’s ministry and with the SS. Wernher von Braun, twelve years younger than the century and twenty-three years old when the move to Peenemünde was decided, was the technical director. The partnership between them was unusually durable for a Nazi era project of this size, in part because the division of labour was clean. Dornberger dealt with the Reich; von Braun dealt with the rocket. It was von Braun who held together a research establishment that eventually ran to nine major departments, and whose heads read like a roll of postwar aerospace: Walter Riedel in the design office, Hermann Steuding on aeroballistics, Rudolf Hermann on the wind tunnels, Ernst Steinhoff and Helmut Gröttrup on flight, guidance and telemetry, Arthur Rudolph on development and fabrication, Klaus Riedel in the test laboratory, Ludwig Roth on future projects.

Both men came out of the strange, brilliant, faintly disreputable milieu of the Weimar rocket enthusiasts, the amateur societies whose members fired small liquid-fuelled motors on waste ground outside Berlin in the early 1930s and wrote unembarrassed about voyages to the Moon. The German Army’s interest in them was practical rather than romantic: rockets were not mentioned in the Treaty of Versailles, and a weapon that fell outside the treaty’s schedules of prohibited artillery was, to the Weapons Office, worth a modest research budget. What began as a modest research budget became, within a decade, the most expensive development programme in Germany. It is one of the standing ironies of the story that the Army’s interest was awakened by a legal loophole and sustained thereafter by the enthusiasm of young men who wanted, in the end, to go into space.

The moral question about von Braun cannot be evaded and should not be inflated. His membership of Nazi organisations, the Party and the SS among them, is a matter of record, and so is the fact that from 1943 the rocket programme he directed was worked in significant part by concentration camp prisoners. Michael Neufeld, the leading modern historian of the programme, has put the balance about as carefully as it can be put: that von Braun probably joined those organisations under pressure rather than out of ideological conviction, but that this does not sever his connection to the crimes committed in the programme’s name, and that “to what extent we can hold them responsible for those crimes is debatable, but we simply cannot remember them only as space heroes.” German reference accounts also attribute to Arthur Rudolph a memorandum of April 1943 to the effect that the operation of Fertigungshalle 1 could be conducted with prisoners, since the SS would handle non work tasks and provide greater security for secrecy, and to Dornberger a directive of 4 August 1943 setting a ratio of German workers to concentration camp prisoners of one to fifteen, or at most one to ten. Both quotations are damaging and specific, both trace back through the secondary literature rather than to a published archival citation, and both should be read as the standard German account rather than as settled documentary fact.

The A-4: building a weapon that had never existed

The Aggregat 4, the A-4, which propaganda would later rename Vergeltungswaffe 2, was not an improvement on anything. There was no prior art. It was a single stage ballistic missile intended to be fired vertically, tipped over onto a programmed trajectory, cut off at a calculated velocity and allowed to fall on a target two hundred miles away, arriving faster than sound and therefore without warning. Every element of that description represented a problem nobody had solved. The motor, developed principally by Walter Thiel with Klaus Riedel, burned a fuel and liquid oxygen combination and is credited with something like 56,000 pounds of thrust, fed by turbopumps delivering propellant at more than fifty gallons a second: a rate of flow that on its own defeated the pumping technology of the day and had to be met with a turbine driven by decomposed hydrogen peroxide. Hence the two hydrogen peroxide plants at Peenemünde, electrolytic and electrostatic, which the RAF’s own target key listed as objectives B and D. A weapon of this kind creates its own industrial hinterland, and every part of that hinterland becomes a target.

Guidance was harder than propulsion. A shell is stabilised by spin and by its own inertia; a rocket standing on a column of thrust is a pencil balanced on a fingertip, and it must be balanced by machinery, in real time, against wind shear and asymmetric burn and the shifting mass of its own emptying tanks. Peenemünde solved it with graphite vanes in the exhaust and aerodynamic surfaces on the fins, driven by gyroscopes, with velocity cut off determined either by an integrating accelerometer or by a radio beam. It rarely worked. R. V. Jones, summarising British intelligence in his interim report of 26 June 1943, put his finger on exactly the right difficulty from several hundred miles away: “The principal trouble experienced has been that of controlling the rocket in flight.” Rockets left the stand and rolled; they broke up in the transonic region; they burned through; they went into the sea a few miles offshore. The wind tunnel establishment under Rudolf Hermann, with an aerodynamics institute running to some 200 technical staff by 1943, existed because the behaviour of a finned body at four or five times the speed of sound was not a matter of calculation but of experiment.

The breakthrough came on 3 October 1942. An A-4 powered by Thiel’s motor lifted from Test Stand VII, held its programme, and flew to a height of more than fifty miles before falling into the Baltic close to its aiming point. It was the first ballistic missile to fly successfully, and it reached a height that the American convention of fifty miles counts as space, though it fell short of the hundred kilometre line most of the world uses. A later vertical firing, in June 1944, cleared that line properly. The standard account of that afternoon has Dornberger telling his assembled staff that they had that day inaugurated the age of the spaceship, a remark that has been quoted in a dozen forms and that no source consulted establishes verbatim; the sentiment, however, is entirely in character, and it captures the doubleness of Peenemünde better than any analysis. The men standing on the sand at the Haken understood perfectly well that they had opened a door to the planets. They were being paid to build an artillery weapon.

Success on the stand was not success in production. Walther von Brauchitsch had ordered the construction of an A-4 production plant at Peenemünde as early as November 1938, and Dornberger set up a planning subsection under G. Schubert in January 1939, but the first assembly line trial runs at Werk Süd did not begin until the middle of 1943, four and a half years later. The reason was that a device with tens of thousands of parts, built to tolerances then unusual in German industry, and still undergoing fundamental redesign, cannot be frozen for mass manufacture. Meanwhile the establishment was proliferating: Wasserfall, Schmetterling, Rheintochter, Taifun and Enzian, an entire family of anti-aircraft and guided missiles, were being developed on the same peninsula. By the summer of 1943 Peenemünde was not one programme but a dozen, and the British, peering at it through a lens from 30,000 feet, could not tell which was which.

Peenemünde-West and the flying bomb

There was a second secret weapon on the island, and it belonged to a different service. Peenemünde-West, established in 1938 as one of the Luftwaffe’s four research establishments, occupied the western side of the peninsula with its own airfield and its own chain of command, and from 1942 it became the test centre for the Fieseler Fi 103, the pilotless flying bomb that Britain would come to know as the V-1. In principle the two projects were complementary: a cheap, mass produced, air breathing cruise weapon fired from a ramp, and an expensive, unstoppable ballistic rocket fired from a trailer. In practice they were competitors for the same scarce commodities, which were not steel and fuel so much as Hitler’s attention and Speer’s allocations. The Luftwaffe’s weapon was crude, slow enough to be shot down and far cheaper; the Army’s was a marvel and cost, by any rational accounting, an absurd amount per ton of explosive delivered. The rivalry was institutional as much as technical, and it ran through the whole V-weapon story to the end.

Two points about the geography matter for what followed. The first is that neither weapon had the range to reach an enemy target from Usedom, so everything fired from the peninsula was a test round, unarmed, aimed at the sea. The second is that when Bomber Command drew up its aiming points for the night of 17 August, Peenemünde-West was not among them. The three aiming points were the living quarters, the production works and the experimental establishment, all of them in the Army’s half of the island. The Luftwaffe establishment, the flying bomb development centre, the one facility on Usedom that would shortly begin killing Londoners at a rate of several hundred a week, was left off the list, and the official history records afterwards that “the west plant did not appear to be hit”. Whether that is entirely true is a question to be taken up later; the intention, at least, is not in doubt. Britain went to Usedom that night for the rocket.

First whispers: the Oslo Report and the agents

British knowledge of Peenemünde began with an act of anonymous conscience. In the first week of November 1939, two months into the war, a package of seven typed pages reached Captain Hector Boyes, the British naval attaché in Oslo. Its author, unknown to London for decades afterwards, was Hans Ferdinand Mayer, a mathematician and physicist who ran Siemens’s telecommunications research laboratory and who had made up his mind about the regime. He had asked, in a preliminary note, for a signal that the material was wanted: the BBC German service duly altered the opening of its bulletin to “Hullo, hier ist London”, and Mayer sent the report, together with a prototype proximity fuse. What became known as the Oslo Report described, in outline and with startling accuracy, a range of German technical projects, and it contained the first references to the German rocket programme at Peenemünde. The Military Intelligence Museum records that it has been called the single most important intelligence report of the entire Second World War, which is a judgement rather than a fact, but not an absurd one.

Most of Whitehall dismissed it. No single source could plausibly know so much across so many fields; therefore it was an Abwehr plant, seeded to waste British effort. The exception was R. V. Jones, newly installed as the Air Ministry’s scientific intelligence officer and possessed of the useful instinct that a document which turns out to be right about the things one can check is probably right about the things one cannot. He believed it, kept it, and spent the next three years watching its predictions come true one at a time. The RAF official history, written by men with no particular stake in Jones’s reputation, endorsed it without reservation: “As early as November, 1939, only two months after the outbreak of war, the British Intelligence Service obtained a document, known subsequently as the Oslo Report, which gave remarkably accurate details of some of the German rocket weapons.”

For three years there was almost nothing further, and then, in the winter of 1942, the fragments began to accumulate. Reports of secret weapon trials at Peenemünde and Swinemünde reached London in December 1942 and again in April 1943, from prisoners of war whose conversations had been recorded and from agents in occupied Europe. The Polish underground contributed a great deal, and it is one of the injustices of the anglophone literature that its contribution is usually reduced to a sentence. Polish resistance workers collected fragments of test rockets from the coastal impact areas before German recovery teams reached them; two Polish janitors from the Trassenheide camp are said to have passed intelligence to the Home Army in early 1943, and individual Polish agents are named in some accounts, including an engineer, Jan Szreder, and a non-commissioned officer, Roman Träger, though these attributions rest on thin sourcing and should be treated with care. The Military Intelligence Museum records a further breakthrough in June 1943, when construction workers from Luxembourg smuggled out a detailed report with a rough sketch map of the facility; that account too is single sourced, and is best attributed rather than asserted.

By late June 1943 Jones was in a position to write the summary that made the case. His Air Scientific Intelligence interim report of 26 June, now TNA DEFE 40/12, is the best single primary document in the whole affair, and it is worth quoting at length because it shows precisely how much and how little was known: “The evidence of the existence of a long-range rocket is reviewed; this evidence comes mainly from three types of source: secret agents, prisoners-of-war, and photography… The evidence shows, beyond doubt of ‘planting’, that the Germans have for some time been developing a long-range rocket at Peenemünde. In its present form this rocket is about 38 feet long by 6-7 feet diameter, with probably three tail fins, and a weight of 40-80 tons; it has been photographed… The principal trouble experienced has been that of controlling the rocket in flight… The only immediate counter measure readily apparent is to bomb the establishment at Peenemünde.” The weight estimate was wrong by a factor of between three and six and a half; the A-4 weighed something on the order of twelve and a half tons. The conclusion, nonetheless, was correct.

Two categories of evidence that are often folded into the pre raid intelligence picture do not belong there, and the distinction matters. The Danish evidence, the crashed pilotless aircraft on Bornholm which gave Britain its first hard physical proof of the flying bomb, belongs to August 1943, that is, to the same month as the raid or later, and cannot have informed the decision to attack. The Swedish evidence belongs to 1944: the rocket that came down at Bäckebo in June of that year, the wreckage of which was passed to Britain, is a landmark in the technical intelligence story but post dates Hydra by ten months. Britain bombed Peenemünde on the strength of agents, prisoners and photographs, and on nothing else.

The argument in Whitehall: Sandys, Jones and Cherwell

The matter came formally before the Chiefs of Staff on 12 April 1943, and the machinery of British government did what it does: it appointed a man. Duncan Sandys, Joint Parliamentary Secretary to the Ministry of Supply, a former anti-aircraft gunner invalided out of active service, was given a committee and a remit under the codename Bodyline. He was to determine whether rockets of more than a hundred kilometres’ range, carrying warheads of between one and five tons, were feasible; how far the Germans had developed them; and what could be done about them. A Bodyline Scientific Committee was assembled around him, its membership including Edward Appleton and John Cockcroft, two of the most formidable physicists in the country. Sandys was energetic, well connected and politically ambitious, and his appointment cut across the existing scientific intelligence machinery, which is to say across R. V. Jones, who was not disposed to be helpful about it. The unedifying feature of the following months is that a good deal of the intellectual firepower of wartime Britain was expended on a question of demarcation.

The substantive opposition, though, came from a different quarter, and from a man with more influence over the Prime Minister than anyone else in the room. Professor Frederick Lindemann, by then Lord Cherwell, Churchill’s personal scientific adviser, did not believe the rocket existed. His objection was not obtuse. He reasoned from the chemistry: a solid fuelled rocket of the required size would be a monster of impossible weight and manufacturing difficulty, and if the Germans were building a liquid-fuelled machine, then the plumbing, the pumps, the cryogenics and the control problem all pointed to an engineering task beyond what he judged the German war economy could deliver at scale. He suspected a deception, a hoax designed to draw British effort away from the bomber offensive. Cherwell held these views with the confidence characteristic of him, and, since he had the Prime Minister’s ear, they mattered a great deal more than their merits alone would have justified.

Against him stood the evidence, and Jones’s insistence on reading it plainly. There is an old and useful distinction in intelligence work between what is possible and what has been observed, and Cherwell was arguing from the first while Jones was reporting the second. The DEFE 40/12 interim report of 26 June is the document that settled the question: it did not claim to know how the Germans had solved the problem, only that they had been working on a long-range rocket at Peenemünde for some time, that the evidence came from three independent classes of source, and that it was beyond doubt of planting. The reason the argument could be settled at all was that, by the end of June, the third of those classes had produced something nobody could dismiss.

Reading the pictures: Medmenham and the rocket on the trailer

The photographic story begins earlier and more casually than the intelligence story deserves. In May 1942 a Photographic Reconnaissance Unit Spitfire, out on a sortie to Swinemünde, speculatively ran its camera over Peenemünde on the way past and brought back images showing three unusual circular embankments. Nobody at the time had any idea what they were. On 22 April 1943, a further sortie produced the first photograph of what would later be recognised as an A-4 standing inside Test Stand VII: an object about a millimetre and a half long on the print, which was noted, catalogued and not identified as a rocket. This is worth dwelling on. Photographic interpretation is not a matter of seeing; it is a matter of knowing what a thing must be, given its shadow, its measured dimensions and its surroundings. A vertical white speck inside an earthwork on the Baltic shore in April 1943 meant nothing, because nothing in the interpreters’ experience had that shape.

On 12 June 1943, Sortie N/853, flown by Flight Lieutenant R. A. Lenton with Sergeant R. S. Haney, brought back photographs of Test Stands I and VII showing what was described as a thick vertical column about forty feet high standing south east of the buttresses. It went to the Industrial Section at Medmenham under Flight Lieutenant André Kenny, whose people enlarged the negatives and, again, could not identify what they were looking at. It should be noted that the dating of this sortie is not entirely secure: Basil Collier’s account places it on 2 June rather than 12 June, and the discrepancy has been reproduced in the literature ever since. Eleven days later, on 23 June, Flight Sergeant E. P. H. Peek of No. 540 Squadron, flying a de Havilland Mosquito PR.IV, landed at Leuchars with the photograph that changed the argument. The Imperial War Museum’s own caption to the frame describes it with admirable flatness: “Clearly seen at bottom centre inside the elliptical earthwork is a V2 rocket on its trailer. Two other trailers can be seen to the right. Also note the light anti-aircraft positions on top of the pre-launch assembly hall at upper right.”

A rocket lying on a low loader is a very different object from a vertical column of unknown purpose. It has a length, a diameter, a nose and fins; it sits on a vehicle of known size, which supplies the scale; and it is unmistakably a thing designed to be moved and fired. After reviewing Peek’s photograph, Kenny’s section went back over the earlier frames and changed the designation of the millimetre and a half specks from “objects” to “torpedoes”, which was as close as the vocabulary of 1943 could come. The RAF official history summarises the resulting picture with care: by July 1943, the sorties of 22 April and of 12 and 23 June had shown “two large objects which appeared to be rockets some forty feet long and seven feet wide”. That is what the War Cabinet was shown. It is also, almost exactly, what Jones had written in DEFE 40/12.

Two corrections to the popular version are needed here, and both matter. The first concerns Jones’s own retrospective analysis. He later obtained a private set of the 12 June photographs from Group Captain P. G. Stewart, the commander at Medmenham, and had the thick vertical columns re measured and confirmed as rockets standing upright on the stand; but that work was done on 4 August 1944, more than a year after the raid, and it therefore forms no part of the case that led to the bombing. The second concerns Constance Babington Smith, who is routinely and wrongly credited with identifying the rocket in the June photographs. Section Officer Babington Smith, who set up the aircraft interpretation section of the Central Interpretation Unit, made an identification at Peenemünde of the very first importance, but it was a different one: she picked out, in her own words, a small cruciform object less than a millimetre in length on the photograph, calculated to have a wingspan of about twenty feet, and it was the Fi 103 flying bomb, not the rocket. Her section then found the launching ramps in northern France. That identification is conventionally placed in the late autumn of 1943, after Hydra. The June 1943 rocket on the trailer was Kenny’s Industrial Section, and the record should say so.

The decision to bomb

On 29 June 1943 the War Cabinet Defence Committee, in its operations form, met to decide what to do. Sandys presented the photographs. Cherwell, as he had throughout, argued that the rocket did not exist and that Britain was being led by the nose. Jones, according to the account that has come down through his own memoir and through Martin Middlebrook’s reconstruction of the raid, answered him point by point and discredited each of his objections in turn; and Churchill, having listened, is recorded in that same account as concluding that the establishment must be bombed by moonlight and on the heaviest possible scale. The Churchill wording rests on the secondary literature rather than on a Cabinet minute in the public record, and is best read as the received account rather than as a verbatim minute. Independent corroboration of the decision itself is not in doubt: the Chief of the Imperial General Staff, General Sir Alan Brooke, wrote in his diary on that same day that they had arrived at the conclusion that a definite threat existed, and that they should bomb Peenemünde.

The strategic reasoning was simple and, in its way, brutal. Nobody in London knew the rocket’s range, its accuracy, its warhead, or how many the Germans could make. Sandys’ own report, written on 26 August, a week after the raid had already been flown, shows how thin the ground remained: “There is little definite information about the range of the projectile… In the great majority of these statements the ranges quoted are over 100 miles… Assuming that the Germans have developed a new fuel with a very high calorific value, a range of 200 miles or over for a projectile of this kind is by no means impossible… In view of the size and weight of the rocket and of the gear required to project it, it seems most probable that the firing sites, whatever form they may take will be rail-served and will be fitted with some form of turn-table.” Against an unknown of that magnitude, aimed at London, and with the concentration of Allied shipping and men for a cross Channel invasion beginning to gather in southern England, the calculus did not require precision. Jones had already supplied the operative sentence in June: the only immediate counter measure readily apparent was to bomb the establishment at Peenemünde. On 15 July the Chiefs of Staff and the Prime Minister approved the plan, to be executed at the earliest opportunity presented by lunar and meteorological conditions.

A word about names is worth inserting here, because the literature muddles it. Hydra is conventionally described as the opening blow of Operation Crossbow, the Anglo American campaign against the German long range weapons programme in all its aspects: research, development, manufacture, transport, launching sites and the interception of the missiles themselves. That description is retrospective. In August 1943 the British effort against the rocket was called Bodyline; Crossbow superseded it as the governing codeword only from 15 November 1943, when responsibility passed to the Air Ministry. Calling Hydra the first Crossbow raid is a convenience of the histories, not the language of the men who ordered it.

Operation Order No. 176

Bomber Command’s Operation Order No. 176 set out three aiming points, to be attacked in sequence by three waves. The first was the scientists’ and workers’ living quarters, meaning the Karlshagen settlement; the second was the production works, the rocket factory in the southern part of the site; the third was the experimental establishment at the northern tip, the development works and the test stands. On the RAF’s own annotated target photograph, now TNA AIR 34/184, these correspond to the lettered key that the interpreters had built up over the preceding weeks: A for the experimental station with its elliptical earthwork and experimental sites, B and D for the hydrogen peroxide plants, C for the power plant area, E for the experimental establishments, F for the sleeping and living quarters, G for the airfield with its test houses and launching track. It is a document of considerable coldness, and it repays the reading, because the ordering of the aiming points was not accidental. The living quarters came first, when the marking would be freshest and the defences least alert.

That ordering followed from the order’s stated purpose, which is the hardest thing about Hydra to set down plainly. The primary objective, as recorded in the operation order and reproduced in the standard accounts of the raid, was to kill as many as possible of the personnel housed in the workers’ quarters; the secondary objectives were to render the establishment unusable and to destroy the documentation of the V-weapon programme. English language readers may reasonably want that claim corroborated from outside the British literature, and it is. The Historisch-Technisches Museum Peenemünde, on the site itself, states the matter without euphemism: Operation Hydra had the aim of killing Peenemünde’s senior engineers and of putting as many development and production facilities as possible out of action. When the German museum standing in the ruins of the target says the same thing as the British operation order, the point may be taken as established. This was, by design, an attempt to decapitate a research programme by killing the researchers in their beds, and it should be described as such rather than softened.

The rest of the plan followed from the demand for accuracy. Peenemünde lies about 600 miles from the nearest British bases, at the extreme of a Stirling’s practical radius and a long night for anybody. It required a full moon, or near enough, both to find the target and to identify the individual aiming points visually, and it required the main force to bomb from around 8,000 feet rather than the usual 19,000, an altitude at which light flak becomes a serious proposition and at which the whole stream would be visible to any night fighter that arrived. Practice attacks on ranges in England had shown errors of about a thousand yards, reduced to some three hundred yards by the final rehearsals. Contemporary accounts of the actual bombing heights vary from about 7,000 to 9,000 feet, which is what one would expect from crews finding their own way down to a marked point in bright moonlight. Nobody involved in the planning was under any illusion about the risk. The order provided for it in the bluntest available terms: if the attack failed it would be repeated the next night, and the night after, regardless of casualties.

The crews were not told why. The briefings described Peenemünde as a research establishment concerned with radar, specifically with German countermeasures against RAF attack, which had the merits of being plausible, of explaining the low bombing height and the unusual care, and of revealing nothing about the rocket if a crew came down in Germany. Flight Lieutenant George Dunn of No. 76 Squadron, twenty years old that night, remembered the briefing threat exactly as it was intended to be remembered: if you don’t do the job tonight, you’ll go back tomorrow, the day after that. Nobody who heard that sentence in a Yorkshire briefing room needed the target’s real identity explained; they understood only that somebody a long way above them regarded this particular patch of Baltic sand as worth the whole force.

Whether Sir Arthur Harris shared that view is an open question, and it is one of the places where the popular literature runs ahead of the evidence. Harris’s known and consistent hostility to what he called panacea targets, single objectives whose destruction was said to promise disproportionate results, makes it tempting to write him as a reluctant executor of somebody else’s obsession. The RAF official history says only that Air Chief Marshal Harris despatched the attack once the weather permitted, and no source consulted for this account documents any resistance on his part to Hydra specifically. His own version is in Bomber Offensive, and the fair statement is that the record does not settle it. What is certain is that the Command he ran executed the operation with a degree of tactical inventiveness that suggests considerable institutional commitment.

The Master Bomber: a new way to run a raid

The problem Bomber Command had never solved was what happens after the markers go down. Pathfinder technique by the summer of 1943 was sophisticated: illuminators to light the ground, primary markers to place coloured target indicators, backers up to renew them as they burned out. But once the main force arrived, the raid ran itself. If the marking was wrong, it stayed wrong. If the aiming point crept backwards down the bombing run as later crews aimed at the near edge of the fires, a phenomenon that had wrecked the accuracy of a hundred attacks, nothing could arrest it. Nobody was in charge over the target. For Peenemünde, where three separate aiming points had to be hit in sequence within about three quarters of an hour and where bombing short meant bombing a labour camp, that would not do.

The answer, developed within Air Vice-Marshal Donald Bennett’s No. 8 (Pathfinder Force) Group, was to put an officer over the target for the whole raid with a radio and the authority to use it. Group Captain John Searby, commanding No. 83 Squadron, was chosen. He was an old regular, an aircraft apprentice of 1929 who had qualified as a pilot by 1935 and been commissioned in 1940, a flight and squadron commander with Nos. 106 and 83 Squadrons through 1942 and 1943, and precisely the sort of unexcitable professional the job required. He had rehearsed the technique over Turin on the night of 7/8 August 1943, ten days before Peenemünde, controlling a smaller force under lighter opposition. Peenemünde is universally described as the first occasion on which a Master Bomber controlled a full scale Bomber Command main force attack, and that formulation, rather than a flat claim of first use, is the accurate one. His aircraft on the night was a Lancaster, W for William.

What the Master Bomber actually did was unglamorous and lethally dangerous. He arrived ahead of the force, identified the aiming point, watched the markers go down, judged which were correctly placed and which were not, and then broadcast to several hundred crews, over and over, telling them which colour to bomb and which to ignore. He then stayed. Searby’s Distinguished Service Order citation records that he remained in the defended target area for thirty minutes while coordinating the main force, executing what it called an extremely difficult and exacting task with consummate skill. Thirty minutes of straight and level orbiting over a target that several hundred of one’s own aircraft are bombing from below, in bright moonlight, with the light flak awake and the night fighters on their way, is an ordeal of a particular kind, and it became a permanent feature of the Command’s method for the remainder of the war.

The marking plan itself was the most elaborate yet attempted. The force would make a time and distance run: crews would take a visual fix on the northern tip of the island of Rügen, some way to the west, then fly a precisely calculated course for a precisely calculated number of seconds before releasing, a technique that converts a navigation problem into an arithmetic one and that is exact provided the initial fix is exact. Sixteen illuminator aircraft would open the attack with white parachute flares and red target indicators, with red spot fires to mark the first aiming point at ten minutes past midnight. Thereafter the sequence ran through yellow target indicators to green backers up, the colours changing as the attack moved between the three aiming points. The innovation that carried this was a new category of aircraft, the shifters, tasked with progressively moving the marking on from one target area to the next as the raid developed. Nothing of the kind had been tried before, and the whole scheme depended on crews reading a shifting palette of coloured fires correctly, in smoke, at eight thousand feet, while being shot at.

Searby’s own account of the marking survives in TNA AIR 20/4040, and its opening is worth having in his words: “the first group of yellow markers went down and these were followed almost at once by the red aiming point indicator: this was correctly placed.” That is what he saw over the second aiming point. Over the first, something else had happened. Stratocumulus lay across the approach, and the H2S fix on Rügen, on which the entire time and distance calculation depended, had been taken imperfectly. The opening markers went down about a mile and a half past the point where they were meant to be, further south down the peninsula, over the wooden barracks of Camp Trassenheide. The museum at Peenemünde, describing the same event from the other end of the bombsight, records simply that Trassenheide was bombed because of a mistakenly placed illuminated marker.

Whitebait: eight Mosquitoes over Berlin

Getting nearly six hundred heavy bombers to the Baltic coast and back in a short August night, at eight thousand feet, under a moon a few days past full, required that the German night fighter force be somewhere else. The German controllers had by 1943 a system of considerable sophistication: radar tracking of the incoming stream, running commentary broadcast to airborne fighters, and, since the introduction of Window in July, an increasing willingness to gather the fighters over whichever city the raid appeared to be developing against and let them hunt freely in the illuminated sky above it. That system had one exploitable characteristic. It responded to the appearance of a raid, and appearances could be manufactured.

The feint was the manufacture, flown against the target Bomber Command codenamed Whitebait, which was Berlin itself. Eight Mosquitoes of No. 139 Squadron flew to Berlin ahead of the main force, the first of them over the capital at 22:56 British Double Summer Time, and scattered bundles of aluminium strip, Window to the British, Düppel to the Germans, in a broad pattern designed to imitate the radar signature of a Pathfinder force marking a target. Some carried bombs, and dropped them; the point was not the damage but the impression. The German account of the effect, as the Peenemünde museum sets it out, is unambiguous: eight British Pathfinder Mosquitoes flew to Berlin and dropped bundles of aluminium strips over a wide area, producing a target marking signature on the radar screens; the German air defence assumed that a large bomber formation was heading for Berlin; and as a result no interceptors remained available to defend Peenemünde. Behind the feint, the real stream was crossing the Danish islands and the Baltic, a hundred miles to the north.

The deception was reinforced by a considerable supporting programme, the details of which vary between reconstructions and is given as reported rather than as exact. As usually given, some twenty-eight Mosquitoes and ten Beaufighters of Nos. 25, 141, 410, 418 and 605 Squadrons flew intruder patrols against the night fighter airfields at Ardorf, Stade, Jagel, Westerland and Grove, eight Halifaxes ran supply sorties for Special Operations Executive, and a small Channel patrol of Typhoons, Hurricanes, a Mustang and a Whirlwind covered the western approaches. The cumulative effect was of a night in which a great deal was happening in a great many places, none of it obviously pointing at a stretch of Baltic sand a hundred miles north of the Reich capital.

It worked completely, for a while. When the first wave ran in over the coast shortly after midnight, there were no fighters over Peenemünde. The opposition was light flak, a scatter of searchlights and naval gunfire from vessels offshore, and a smoke screen that was started too late and never became complete. The debriefing of a No. 97 Squadron crew that night captures the eerie quality of it: “Defences were very meagre; a few searchlights straggled round the sky.” For the first time in the experience of most of the men over the target, a major German installation was being bombed from eight thousand feet in full moonlight and almost nothing was coming up. It could not last. Somewhere south of them, the controllers were working out that the fires over Berlin were nothing much, that the aluminium was falling on empty air, and that the real weight of Bomber Command was two hundred miles away and would shortly have to fly home. The fighters began to turn north.

The first wave: the marker that fell on Trassenheide

The clocks are the first thing that goes wrong in any account of Hydra, and it is worth saying so before the bombs start falling. The reconstructions differ by up to twenty minutes. The most widely used sequence has the red spot fires opening the marking at ten minutes past midnight, the sixteen illuminators laying their white parachute flares from a minute later, and the first wave turning for home at twenty-seven minutes past. The Aircrew Association account of a navigator who was there gives the three waves as quarter past, half past and quarter to, in round numbers, which is how a man remembers a night rather than how a staff records it. The RAF Memorial Flight Club states that the first bombs fell at thirty-five minutes past midnight. Searby’s own report, written within days and lodged in the file that is now TNA AIR 20/4040, uses times two hours later again: heavy bombing commencing at 0200, the attack going well by 0210, all aircraft told at 0245 that they were going home. That is almost certainly a different time reference rather than a different memory, and no purpose is served by forcing the figures to agree. What follows uses the conventional British Double Summer Time sequence, with the caveat standing.

The first wave belonged to Nos. 3 and 4 Groups, Stirlings and Halifaxes, something in the order of 227 aircraft attacking. Its aiming point was F on the RAF’s own lettered key: the sleeping and living quarters, which meant the Karlshagen estate, the neat works settlement built for around five thousand people, with its gardens and its splinter tubes and the families of the men who were building the rocket. The approach depended on the time and distance run from the northern tip of Rügen, and the whole method rested on one assumption, which was that the initial fix would be exact. It was not. Stratocumulus lay across the run in, the H2S picture of the Rügen coastline was misread, and the calculation that followed was accurate arithmetic applied to a false starting point. The opening markers went down about a mile and a half past the point where they belonged, further south down the peninsula.

A mile and a half, on that peninsula, was the difference between one kind of target and another. What lay under the misplaced fires was Trassenheide: the Deutsche Arbeitsfront communal camp, forty-two barracks by the German count, forty-five huts by the English one, holding perhaps four thousand people, foreign labourers and German construction workers together. The museum on the site describes what happened in a single clause, without adjectives, from the receiving end: Trassenheide was bombed because of a mistakenly placed illuminated marker. About a third of the 227 attacking aircraft of the first wave bombed on that marking before it could be corrected. Roughly seventy heavy bombers, at eight thousand feet, in moonlight, with a mixed load of high explosive and incendiary, put their bombs into a camp of wooden huts.

Searby saw it and worked. This is precisely what the Master Bomber had been invented for, and the fact that it took some minutes to have effect is less remarkable than the fact that it had any effect at all. He noted one yellow marker as very well placed, called the correction, and went on calling it. His own account records that several broadcasts were made to all aircraft from time to time assuring them of the accuracy of the marking and exhorting them to keep up this very high standard, which is the language of a report rather than of a man in a cockpit, and which conceals a great deal. He was orbiting a target being bombed from below him by hundreds of his own aircraft, in bright moonlight, for half an hour, talking. The shifting of the marking on to the correct point, and thereafter on to the second and third aiming points, is the reason the raid worked at all.

The German smoke screen is another matter on which the sources will not agree, and both versions come from the same archive. One Pathfinder account calls it an extremely effective smoke screen. Another, quoting the debriefing of Pilot Officer J. F. Munro’s crew of No. 97 Squadron, records that cover was by no means complete and that one could see quite a lot through the lines of smoke. The squadron summary logged simply smoke haze or screen; the crew bombed on yellow and green target indicators and saw no results because of it. The likeliest resolution is that both are true at different moments. The screen went up incompletely, thickened as the attack developed, and by the time the third wave arrived the smoke had married with the fires and made individual buildings impossible to distinguish. Searby’s report says exactly that: one final run over the target area, and it was now quite impossible to make out individual buildings owing to the tremendous fires sweeping the area.

In the Karlshagen settlement, meanwhile, the shelters did what they had been built to do. The Splitterschutzröhren were concrete tubes about a metre and a half across, fitted with wooden benches and a blast porch, sunk close to the houses so that a family could reach one in seconds. They stopped splinters and blast and nothing else. Horst Urban, who spent that night inside one, left an account of what it was like to be in a concrete pipe under a main force attack: they began to count the impacts, one, two, three, coming closer, and asked themselves whether the next would hit; no, the fourth fell on the other side of the tube, then the fifth, and so on, always the same game. The museum’s judgement is that because the tubes stood so close to the houses they saved many residents’ lives that night. That is the reason the German civilian death toll at Peenemünde, for a raid deliberately aimed at German civilians, was as low as it was.

At Trassenheide there was nothing of the kind. The huts were timber, set in rows, inside wire. Foreign workers had no access to shelters of any consequence, a restriction that was as much a matter of security policy as of construction cost, and the camp gate was a long way from most of the barracks. English accounts record that twenty-three of the huts were destroyed. German accounts of the deaths, taken up in detail below, record a firestorm effect from phosphorus incendiaries among the timber buildings alongside the high explosive. Some of those inside are said to have died trying to climb the wire because the gate was too distant, a detail that rests on thinner sourcing than the rest and is best left as a report rather than a certainty. What is certain is structural and needs no eyewitness at all. A camp of wooden huts, unsheltered, wired in, was hit at close to the density of a city area attack, by aircraft that had been told the marking was correct.

The second wave: the production works

At about half past midnight the marking shifted north, on to the second aiming point, and No. 1 Group came in. This was the smallest of the three waves and the one that worked: something in the order of 131 aircraft, of which 113 were the Lancasters of the main force, with six Pathfinder shifters and twelve backers up to hold the marking in place. Their objective was the production area of Werk Süd, the rocket factory, and above all the enormous new assembly hall, Fertigungshalle 1, some 248 metres by 120, one of the largest column free structures of its day and barely in use. It was the physical expression of the decision to turn a research programme into a manufacturing one, and it was the single most valuable building on the peninsula.

The wave did not begin cleanly. Crews arriving over a target already burning from the first attack had the natural and almost irresistible instinct to bomb what was already alight, and a number of them initially ignored the correctly shifted markers in favour of the fires burning behind them to the south. Searby corrected them, and the 113 Lancasters went on to the production works. It is over this aiming point that his account produces the sentence quoted earlier in this narrative, and it deserves its second hearing because of where in the night it falls: the first group of yellow markers went down and these were followed almost at once by the red aiming point indicator, and this was correctly placed. Everything about the marking plan that had failed over Trassenheide worked here. One considered modern judgement, that of the Defense Media Network’s survey of the Crossbow operations, is blunter still: only the second wave successfully struck its targets.

The damage was real and it was concentrated. Fertigungshalle 1 was heavily hit. The administration building and the drawing office of the main workshops were completely destroyed, along with a housing shed and the hutted camp, according to the RAF official history’s summary of the post-raid interpretation. Production drawings for the A-4 were burned, although the Germans had duplicated their records elsewhere, which blunted the loss almost at once. And in the basement level of the assembly hall, where the army’s concentration camp prisoners were housed so that they could be walked upstairs to the production line without ever leaving the building, men died. The museum states the fact plainly: the hall was heavily damaged and the prisoners quartered there for rocket assembly were among the dead. Perhaps six hundred of them lived in that basement. The most accurately bombed part of the entire operation killed some of the least free people on the island.

The third wave, and the fighters arrive

The last wave was the largest and it had the worst of everything. Nos. 5 and 6 Groups provided it, something in the order of 178 aircraft, the bulk of them Lancasters with a substantial component of Halifaxes, and their objective was the experimental establishment at the northern tip: the development works, the laboratories, the test stands, the elliptical earthwork of Test Stand VII that had been photographed from 30,000 feet in June. A green marker went down on the development works at about twelve minutes to one, and the attack was to run from then until five minutes to the hour. The whole force was to be clear of the target by then, some six hundred miles from home, with the shortest August night in prospect and the German fighter force by now certainly awake.

This was also the first operation on which No. 6 (Royal Canadian Air Force) Group flew Lancasters, a milestone that its squadrons would probably have preferred to mark on a different night. The third wave’s problem was the same time and distance run that had misled the first, compounded by exhaustion and by the state of the target. Crews overran the timing point by twenty or even thirty seconds. At the speed of a loaded Lancaster that translated into bombs falling between 2,000 and 3,000 yards past the aiming point, which is to say back down the peninsula and into the camp and settlement areas that had already been hit an hour earlier. Some accounts render the same error in metric as two to three kilometres from the development works. Either way the effect was to add a third helping of bombs to ground that did not need it and to leave the experimental works less damaged than the plan required.

And then, at the worst possible moment, the timing of the whole operation came apart at the back. At five minutes to one, with the attack notionally complete, some thirty-five stragglers were still over the target area waiting for their release point. These were aircraft that had lost time somewhere over the North Sea or the Danish islands and were now, individually and in the open, running in over a burning target under a moon that was ninety-six per cent full. The whole architecture of a bomber stream is concentration: the force protects itself by being in one place for the shortest possible time. Thirty-five aircraft strung out behind the stream had no such protection.

By then the deception had been read. Whitebait had done its work for the better part of two hours, and the eight Mosquitoes of No. 139 Squadron had held the German controllers’ attention over Berlin long enough for the first two waves to bomb almost unopposed. But aluminium strip does not drop bombs, and once it became clear that the weight of the raid was not falling on the capital, the fighters were released northwards. Roughly thirty Focke-Wulf Fw 190 single engined machines flying the Wilde Sau method arrived first, Hajo Herrmann’s freelance system in which a day fighter pilot hunts visually over an illuminated target with no radar of his own and no ground controller holding his hand. Over Peenemünde that night there was no need for radar at all. There was a burning target, a moon a few days past full, a layer of smoke below to silhouette against, and a stream of heavy bombers at eight thousand feet.

The twin-engined night fighters came in behind them, and the argument about how they were controlled is a live one in the specialist literature. One account holds that Peenemünde also marked the first successful use of the Zahme Sau method since the abandonment of rigid Himmelbett box control, with running commentary steering the fighters at the concentrations of Window and crews then hunting independently on their own airborne radar. What is not in doubt is the outcome. The German night fighter arm claimed thirty-eight victories that night, thirty-three of them over the target area itself. Almost all of that fell on the last wave and on the stragglers behind it, in the twenty minutes after the bombing was supposed to have finished.

Schräge Musik

The weapon that did much of the killing was a pair of cannon bolted into the roof of a night fighter’s fuselage, pointing upwards and slightly forward. Schräge Musik, slanting or jazz music, was the Luftwaffe’s name for it, and its logic was a matter of geometry rather than of technology. An RAF heavy bomber of 1943 carried a nose turret, a mid upper turret and a rear turret, and it carried nothing at all underneath. The ventral position had been deleted from the Lancaster in the interests of weight and the H2S scanner; the Halifax’s arrangements were little better. A fighter that could get beneath a bomber and stay there was invisible to every gunner in the aircraft, and the standard German attack, a curving pass from astern and slightly below, had always suffered from the problem that pulling up to shoot exposed the fighter to the rear turret at the worst moment.

Upward firing guns removed the problem entirely. The fighter formates below and behind, matches speed, holds station in the bomber’s blind cone, and fires vertically into the wing between the engines, where the fuel is. Aiming was through a simple reflector sight in the cockpit roof. The ammunition was frequently loaded without tracer, so that there was no luminous line back to the attacker and, very often, no indication to any surviving crew of what had happened at all. Aircraft that had been hit reported an explosion in a wing, or a fire that started for no reason. Bomber Command spent months attributing losses of this kind to flak, to fuel system failures and to collisions, and the tactical response, when it came, was the corkscrew and the ventral observation blister rather than any specific counter to a weapon the Command did not know existed.

Peenemünde is usually given as the first operational use of Schräge Musik, and it is certainly the first night on which the weapon made a significant difference to a major action. The unqualified claim of first use, however, cannot survive contact with the development history. Oberleutnant Rudolf Schoenert, who had pushed the concept through against considerable institutional indifference, is credited with a confirmed Schräge Musik victory as early as May 1943; Oberfeldwebel Mahle devised the practical mounting; and Luftwaffe trials had been running at the Tarnewitz test establishment through 1942. The defensible formulation is that 17/18 August 1943 was the first occasion on which Schräge Musik equipped aircraft made a decisive contribution to a major night action, and that even the standard reference works, when they assert more than this, hang a cautionary footnote on the claim.

Two crews are usually named as having flown the first operational Schräge Musik sorties that night, in two Bf 110s: Leutnant Peter Erhardt, a Staffelkapitän, and Unteroffizier Walter Höker. The night’s most striking individual score went to Leutnant Dieter Musset, with his radio operator Obergefreiter Helmut Häfner of 5./NJG5, shooting down five bombers with the upward guns, the only victories either man scored in the whole war, before the cannon jammed on a sixth attack, their aircraft was hit and both baled out injured near Güstrow; Höker with two four-engined bombers; Erhardt with four in thirty minutes; and Peter Spoden of the same unit opening his account with Lancaster III JA897 of No. 44 Squadron, flown by Flight Sergeant Johnnie Drew. Musset at least has independent German corroboration: the museum at Peenemünde records that Lancaster DV 202 was probably shot down by his Me 110. Beyond that, the names and the scores vary between sources, and anyone who states them as settled is overreaching.

The cost: forty bombers

Forty heavy bombers failed to return. The Bomber Command War Diaries figure, which is near universal in the literature, breaks down as twenty-three Lancasters, fifteen Halifaxes and two Stirlings, a loss rate of 6.7 per cent of the force despatched. Some accounts add one of the Mosquitoes and record a further thirty-two aircraft damaged. Searby’s own report gave the loss as forty-one Lancasters, which is wrong on both the count and the type, and it is worth pausing on that error rather than passing over it, because the same report is the source of two other numbers that have echoed through the popular literature for eighty years. A senior and conscientious officer, writing days afterwards with the resources of a group headquarters behind him, got the losses of his own force wrong. The lesson is not that Searby was careless. It is that contemporary assessment, in every direction, was worse than we tend to assume.

The distribution of the losses is the whole story. Twenty-nine of the forty came from the third wave, that is to say from Nos. 5 and 6 Groups, which between them had put up something in the order of 166 to 178 aircraft. Expressed as a rate for those two groups alone that is comfortably above the ten per cent that the Command regarded as unsustainable, and it was suffered in about twenty minutes, over and just after the target, by crews who had already flown six hours and had six more to go. The first wave, which had bombed in near silence with the fighters still orbiting Berlin, lost both of the Stirlings and little else. The difference between flying in the first wave and flying in the last, on the night of 17 August 1943, was the difference between an unusually quiet operation and one of the worst of the war. Flight Lieutenant George Dunn of No. 76 Squadron, twenty years old, bombed from seven thousand feet at twenty past midnight in the first wave, and knew for the rest of his life that a last minute squadron reallocation had moved him out of the final wave and probably saved him.

Against the bombers’ loss the German night fighter force is generally reckoned to have lost around a dozen aircraft, one account itemising four Bf 110s, a Dornier Do 217, two Fw 190s, a Bf 109 and a Junkers Ju 88, a tally that does not reach its own total, with five German fighters claimed destroyed by the bomber crews. A further and much more dramatic claim, that around a hundred fighters attempting emergency landings at Brandenburg-Briest produced a mass crash destroying thirty-four of them, rests on a single low tier source and should not be repeated as fact until somebody has found the German records.

How many men died is a question with four incompatible answers, and the honest form is a range. Between 215 and 280 aircrew were killed, depending on whether the count includes those who died of wounds, those killed in crashes on return to England, and those who came down in Germany. The commonly cited figure is 215. The RAF Memorial Flight Club gives 243; the RAF Benevolent Fund says almost 280; and the arithmetic offered in one set of teaching notes, around 287, is derived by multiplying Searby’s erroneous forty-one Lancasters by a crew of seven and should be set aside for that reason alone. No Victoria Cross was awarded for Operation Hydra. The gallantry of the night is recorded in Searby’s Distinguished Service Order, and in a great many names on the Runnymede memorial.

One of the lost aircraft is better documented than almost any other Bomber Command loss of 1943, because it is still there. Lancaster III DV 202, coded KM-Z, of No. 44 (Rhodesia) Squadron, lifted from Dunholme Lodge at twenty to ten and came down in the Kölpinsee, the shallow lake about a kilometre north of Peenemünde village, apparently in an attempt to ditch. All seven aboard were killed: Pilot Officer R. C. Harding of the RCAF, Pilot Officer L. Prendergast, Sergeant T. N. Weston, Sergeant W. H. Quance, Sergeant L. F. McDermott, Flying Officer P. Pynisky of the RCAF and Flight Sergeant S. Shaw, although the ranks are recorded differently in different registers and one account lists the aircraft as lost without trace with the crew commemorated at Runnymede. Six of the seven bodies are believed to have been recovered by concentration camp prisoners and buried on the southern shore of the lake. The same squadron lost JA897 to Spoden. At Fiskerton, another 5 Group station, twenty-three aircrew did not come back.

For everybody else there was the flight home: six hundred miles of Baltic and North Sea in the dark, over the Danish islands, with the fighters following as far as they could and the fuel calculations tightening, and then the long descent into a Lincolnshire or Yorkshire dawn. No. 97 Squadron had detailed twenty aircraft, of which sixteen attacked, and every one of them came back; its crews had seen a single bomber coned and blown up over Flensburg on the way out and nothing else all night. Other squadrons counted differently. The Aircrew Association account records a landing time of twenty-seven minutes past four in the morning, which meant a little under seven hours of daylight before the battle order for the following night went up.

The dead on the ground

Somewhere between about 600 and 735 people were killed on the ground at Peenemünde that night. That is as precise as the evidence allows, and the spread is not a failure of research but a consequence of who the dead were. The Historisch-Technisches Museum on the site, which is the institutional German authority and which has no interest in minimising anything, gives approximately 700 on one of its pages and over 700 on another. German reference accounts give at least 703. Two English language sources give 732 and one gives 735. Dornberger himself, in British captivity at Trent Park after the war, said 720. The RAF official history offered a bracket of between 600 and 800. A works settlement, a labour camp and two concentration camp satellites were hit on the same night, and only one of those four populations was on a register that anybody troubled to keep accurately.

The composition of that total is the point at which this story is most often told wrongly, in both directions, and it needs stating with care. British language accounts, following the Bomber Command campaign diary, usually say that between 500 and 600 of the dead were foreigners, mostly Poles, at Trassenheide, against around 180 Germans. The museum at Peenemünde says something materially different: that over 700 died, of whom over 200 were forced labourers, killed by high explosive and by firestorms started by phosphorus incendiaries in a bombardment that fell chiefly on the settlement. German figures for Trassenheide itself are more granular than anything on the British side and give 294 dead in the camp, of whom 97 were German, 131 foreign and 66 of unknown nationality. It is not possible, on this evidence, to assert as fact that most of the dead were slave labourers. It is possible, and necessary, to say that a great many of them were, that the exact proportion is disputed between reputable authorities, and that the disagreement itself tells us something about how the Reich counted people.

A separate museum figure, drawn from its work on forced labour rather than on the raid, records at least 232 concentration camp prisoners, 197 foreign forced labourers and eleven Soviet prisoners of war killed in Allied air attack, a total of 440. That breakdown may aggregate the raid of August 1943 with the American attacks of 1944, and it should be handled accordingly. The museum separately documents 248 deaths of concentration camp prisoners at Peenemünde across the whole period between 1943 and 1945. Among the named individuals its exhibition preserves is Leon Dropek, a Polish civilian worker photographed in 1940, one of the very few of the peninsula’s foreign workforce whose face and name have both survived.

Two Germans of consequence died, and the museum is exact about both of them, which the English literature is not. Dr Walter Thiel, born in Breslau in March 1910, who had joined the rocket work at Kummersdorf in 1936 and come to Peenemünde in 1940, was the head of the propulsion department, the man principally responsible for the motor that had lifted an A-4 more than fifty miles above the Baltic the previous October. He was killed with his family by a direct hit on a splinter protection trench, the kind with reinforced sides and a roof of oak beams, proof against everything but the thing that killed him. There is a crater on the Moon named after him. The other was Erich Walther, the works manager of the development works, whom English sources almost invariably render simply as Chief Engineer Walther. Of the senior scientific staff of the establishment, exactly two men died. That single sentence disposes of a great deal of what was believed in London for the following two years.

The rest of the dead have no such biographies, and the manner of their dying can only be reconstructed from the architecture. A splinter tube in the Karlshagen gardens was a concrete pipe a few paces from a front door. A barrack at Trassenheide was a timber building in a row of forty-two or forty-five such buildings, inside wire, with no shelter of consequence within reach and a gate a long way off. Onto that arrangement, in a period of about a quarter of an hour, fell the bomb loads of perhaps seventy heavy bombers, and then, an hour later, a proportion of a third wave that had overrun its release point by between 2,000 and 3,000 yards. Twenty-three huts were destroyed. In the days that followed, the surviving prisoners were fed irregularly and set to the heaviest of the clearance work, digging out the camp they had been living in. The stone at the Karlshagen cemetery records the burial place of the two thousand victims of the air raids on Peenemünde and Karlshagen of 17/18 August 1943 and 18 July 1944. That figure is a memorial inscription of the East German period, aggregating both raids and rounded upwards for the purposes of commemoration. It is not evidence, and it should not be quoted as though it were. What it does mark, accurately, is that the great majority of the people buried there had no say in being at Peenemünde at all.

What Hydra achieved

Searby’s report supplies the wartime verdict, and it is wrong three times in four lines. As a kind of bonus, he wrote, it was discovered at a later date that 250 scientific workers of varying degrees of importance had lost their lives in the attack; the effect of the raid was to delay Germany’s V weapon programme by six months; Bomber Command losses in this operation were 41 Lancasters. The scientific dead were two, not 250. The losses were forty aircraft of three types. And the delay was not six months. It is not a criticism of the man to say so. He was writing from within a system that had no access to German records, that had been given a target it was not told the purpose of, and that had every organisational incentive to believe its own bomb plots. But the six month figure entered the literature from this document, and it has been reprinted ever since.

The figure that most people now quote is two months, and its provenance deserves to be understood, because it is a good deal softer than its repetition suggests. It comes from the British official history, Webster and Frankland, which wrote that the attack may well have caused a delay of two months, noted that Dornberger himself claimed between four and six weeks, or possibly two months once other attacks were counted in, and then described that supporting evidence, in the historians’ own word, as anecdotal. A hedged phrase resting on evidence its authors distrusted has hardened, by way of a thousand retellings, into a round number. Goebbels, writing in his diary at the time and with every reason to minimise, thought preparations had been thrown back by four or even eight weeks. Set against this is the single most awkward fact in the argument: test launches at Peenemünde resumed on 6 October 1943, some seven weeks after the raid. That date, if it holds, disposes of six months by itself.

Modern assessments range about as widely as it is possible for professional opinion to range. The United States Strategic Bombing Survey, reporting in 1945 with access to the German records and no British reputations to protect, judged the raid not effective, on the grounds that V-1 production was already established near Kassel and that V-2 production had gone underground. At the other end, Adam Tooze, whose account of the German war economy is the standard modern work, calls the bombing highly successful, but for a different reason than the one usually advanced: not because it delayed the rocket’s development, but because it forced the transfer of the production of twelve thousand A-4 missiles to Thuringia, and that transfer was a Herculean task. This is the strongest defensible claim available. Hydra did not stop the rocket. It moved the rocket, at enormous cost in German administrative effort, at a moment when Germany had none to spare.

By its own stated primary objective the raid failed. Operation Order No. 176 had set out to kill the men who were building the weapon, and the museum on the site confirms from the German end that this was understood to be the aim. Of the senior scientific staff, two died. Von Braun survived; Dornberger survived; Rudolph, Steinhoff, Debus, Rees, Gröttrup, the entire directing intelligence of the programme survived and was working again within weeks. The wind tunnel establishment and the measuring equipment came through, as Dornberger noted with some satisfaction, which meant that the research on which everything depended was barely interrupted. A decapitation strike that kills two of a leadership of dozens, at a cost of forty aircraft and between 215 and 280 aircrew and between 600 and 735 people on the ground, cannot be called a success on its own terms, whatever its incidental consequences.

There is one further loose end, and it goes to the heart of what the raid was for. Peenemünde-West, the Luftwaffe establishment where the Fi 103 flying bomb was being tested, was not an aiming point. The RAF official history states that the west plant did not appear to be hit. Yet the RAF Museum holds a post-raid reconnaissance photograph, accession AC71/14/7/22, whose own catalogue description records a concentration of bomb craters on the airfield and damage to technical buildings of the Luftwaffe test facility at Peenemünde-West. The two statements are in direct conflict and this account does not attempt to resolve them. It is possible that the interpreters distinguished between the airfield and the plant proper; it is possible that a proportion of the overshooting third wave fell across the western side; it is possible that the official history simply summarised too briskly. What is beyond argument is intent. Britain flew to Usedom that night for the rocket, and the weapon that would kill 6,184 people in England over the following year was developed on a part of the peninsula that nobody had put on the list.

The suicide of Hans Jeschonnek

Generaloberst Hans Jeschonnek, born in April 1899, had been Chief of the General Staff of the Luftwaffe since February 1939. He was young for the post, able, and by 1943 trapped between a Führer who demanded offensive results and a Reichsmarschall who would neither support him nor let him go. The Hamburg firestorm of late July had broken something; when he asked Göring afterwards to release him to a field command with Luftflotte 4, he was refused. His father, his brother and his brother-in-law had all died recently. He had come close to a breakdown in November 1941 after Udet shot himself. The account offered by the Dupuy Institute treats the night of 17 August as the final catalyst rather than the cause, which is almost certainly the right weighting.

What happened that night, from Jeschonnek’s chair, was a humiliation of a very specific kind. Deceived by the Mosquito feint, he ordered Berlin’s anti-aircraft defences to open fire while German night fighters were orbiting overhead, endangering his own aircraft in order to defend a city that was not being attacked. Meanwhile the actual raid, some hundred and fifty miles to the north, was destroying the most secret installation in the Reich and killing several hundred of its people. The same twenty-four hours had already seen the American Eighth Air Force strike Schweinfurt and Regensburg in daylight, the deepest penetration yet attempted, and although it cost the Americans severely it demonstrated that no part of Germany was now beyond reach in either darkness or daylight. It would be an overstatement to make the double blow the explanation of what followed, but it is fair to say that on that day the air defence of the Reich failed twice, in two different mediums, and that the man formally responsible for it was already at the end of his resources.

He shot himself in his office aboard the Luftwaffe command train, at what is now Goldap in Poland, close to the Wolfsschanze. The date is contested. Göring’s own announcement gave 19 August 1943, and gave the cause of death as a stomach haemorrhage. Most accounts now give 18 August, and hold that the later date was put about to obscure the connection with the Peenemünde raid. The safest formulation is that Jeschonnek shot himself on 18 or 19 August 1943, that the official date was the 19th, and that the case for the 18th is a serious one. He left a note that read, in the version generally quoted, that he could no longer work together with the Reichsmarschall, followed by a salute to the Führer, and a memorandum to Hitler critical of Göring.

One story attached to his death should be flagged and set aside. It is sometimes said that Hitler telephoned Jeschonnek on the night of 17/18 August and said, in effect, that he knew what to do. The anecdote is dismissed as untrue by Professor Richard Suchenwirth in the USAF Historical Division’s study of German air force command, who traced it to Erhard Milch’s contested postwar testimony. Milch had reasons of his own for wanting the responsibility placed at the top. The story is too good and too convenient, and it does not belong in an account of what Hydra caused.

Underground: Mittelwerk, Dora and Blizna

The consequence that mattered began nine days after the raid. On 26 August 1943 Albert Speer convened a meeting with Hans Kammler of the SS, Dornberger, Gerhard Degenkolb and Karl Otto Saur to settle the future of A-4 production, and the answer they arrived at was to put it underground and out of reach of Bomber Command. The chosen site was a complex of tunnels driven into the Kohnstein, an anhydrite ridge near Niedersachswerfen, just outside Nordhausen in the southern Harz, which had been excavated before the war for fuel and chemical storage. It became the Mittelwerk. From September 1943 machinery and personnel began to move, Arthur Rudolph among them with a party of around ten engineers, and on 13 October the concentration camp prisoners who had been living in the basement of Fertigungshalle 1 were entrained for the Kohnstein.

To dig and equip a rocket factory inside a mountain required labour that could be worked without limit, and the SS supplied it. The first 107 prisoners arrived from Buchenwald on 28 August 1943 to what began as a subcamp and became, on 1 November 1944, the independent concentration camp of Mittelbau-Dora with 32,471 prisoners on its books. For the first months the workforce lived underground in the tunnels themselves, in unventilated galleries thick with gypsum dust, sleeping in tiers of bunks in side chambers, without daylight, without adequate water and without sanitation. The mortality of that first winter was catastrophic and is the reason Dora acquired the reputation it has. Around sixty thousand prisoners passed through the camp system between August 1943 and March 1945. Something over twenty thousand of them died in it, which is to say that around one in three of those confined there did not survive; some 350 were hanged, including two hundred for sabotage, with as many as thirty executions in a single day in the last months.

The rest of the establishment scattered across the Reich in the same months. The valve laboratory went to a castle near Leutenberg, materials testing to the airbase at Anklam, the wind tunnels to Kochel in Bavaria, engine testing to Lehesten and to Zipf in Austria with its own liquid oxygen production. Further underground works were begun in the Austrian Alps, the Zement complex near Ebensee from January 1944 and the Salamander test stands in early 1944. What had been a single, coherent, dangerously concentrated research and production centre on one Baltic peninsula became a dispersed archipelago, harder to bomb, harder to photograph, and immeasurably harder to run.

Test firing had to move too, since the whole point of Peenemünde had been the four hundred kilometres of empty Baltic to the north east, and that was now a range under British surveillance. The programme went east, to the SS training ground at Heidelager near Blizna in south eastern Poland, a site built by two thousand concentration camp prisoners. It was there that the Germans learned that Poland was not a safe place to fire rockets. The Home Army’s air reconnaissance service, under the engineer Antoni Kocjan, put observers around the range; they noted the German deception measures, including plywood models of houses and plaster dogs and dolls set out to make an evacuated village look inhabited from the air. Foresters at Wola Osiecka helped obtain rocket components, Polish scientists examined them in secret and forwarded their findings to London, and in the end an intact unfired rocket was recovered from the Bug and flown out in Operation Most III. The dispersal that Hydra forced thus produced, by a route nobody in London had planned, the best technical intelligence on the V-2 that Britain ever received.

This is the raid’s real legacy, and it is not a comfortable one. Hydra did not stop the rocket and it did not kill the men who made it. What it did was drive the programme off a peninsula where prisoners at least worked in daylight and into a mountain where they did not. The decision to go underground was German and the crimes committed there were German, and no share of the moral responsibility for Dora belongs to the crews of Nos. 3, 4, 5, 6 and 8 Groups. But the causal chain is plain enough, and the historian who omits it is not telling the story.

The ledger: what the rocket did, and what it cost to build

The first Mittelwerk built V-2 was launched successfully against London on 8 September 1944, thirteen months after the raid that was supposed to have prevented it. Production reached six hundred units a month by that September and continued into February 1945; the total built between August 1943 and April 1945 is usually given as around six thousand. They fell on London and on the Home Counties without warning, at several times the speed of sound, arriving before the sound of their own approach, which had a particular effect on civilian nerve that the flying bomb, for all its greater destructiveness, never achieved. The last to fall on London struck Hughes Mansions in Vallance Road, Stepney, on 27 March 1945, six weeks before the end of the war in Europe. Antwerp, whose port the Allied armies needed and whose civilians had no shelter policy worth the name, was the other principal target, and took a heavier share of the campaign than is usually remembered in Britain.

The British figures are the ones most often cited and they are worth stating precisely. V-2 attacks killed 2,754 people in London and injured another 6,523 there. For comparison, the flying bomb, the cheaper and cruder weapon, killed 6,184 and seriously injured 17,981. Against that the Allies flew 68,913 sorties in the Crossbow campaign and dropped 122,133 tons of ordnance on the V-weapon system in all its forms, of which Hydra was less than two per cent. As a return on investment for either side the arithmetic is dismal. Michael Neufeld’s summary of the weapon is that its impact was minor and that it was an extraordinarily expensive way to drop a one ton bomb on an enemy city.

The comparison that people reach for, and frequently garble, is between the dead at the receiving end and the dead at the manufacturing end. Handled carelessly it becomes a slogan. Handled correctly it is one of the most damning facts in the history of twentieth century technology, and Neufeld, who has done more work on it than anybody, states it in a specific form that should be preserved. At least ten thousand deaths can be attributed to the V-2 programme itself, roughly twice the number of those killed on the receiving end, which he puts at approximately five thousand Allied civilians and soldiers across all the targets. That is the comparison, and it is the only one the evidence supports. It is not the same as the figure of over twenty thousand deaths in the Mittelbau-Dora system, which covers the whole camp complex across all its projects between 1943 and 1945, and which includes work that had nothing to do with rockets. The two numbers are frequently run together and they should not be. Nor should the London death toll of 2,754 be set against either of them without saying that Antwerp and the continental targets account for the difference.

Stated at its narrowest, then: the V-2 killed about twice as many people in the building of it as it killed in the use of it, and the people it killed in the building of it were prisoners. There is no comparable statement to be made about any other weapon of the war.

Paperclip: the men who went to America

In the last week of April 1945 the United States Army took into custody more than 450 German rocket specialists at Oberammergau, where the SS had concentrated them, and on 2 May von Braun, Dornberger and a number of others formally surrendered at Reutte in the Tyrol. The men who had run Peenemünde had spent the preceding months making a calculation about which of the advancing armies they preferred, and had acted on it deliberately, moving themselves and as much documentation as they could westwards. The technical archive of the programme, buried in a mine in the Harz, was recovered by the Americans; so were enough components to assemble dozens of complete rockets, shipped out of the Mittelwerk in the weeks before the area passed to the Soviet occupation zone.

What followed was Operation Paperclip. Around 127 engineers were eventually contracted to work for the United States Army, the main body based at Fort Bliss in Texas with a detachment at the White Sands Proving Ground in New Mexico. The first shipment crossed on the SS Argentina and arrived on 16 November 1945; a third came on the Victory ship Central Falls in February 1946. A group photograph taken at Fort Bliss that year shows 104 of them, in suits, arranged as for a company portrait. In 1950 the team moved to the Redstone Arsenal at Huntsville, Alabama, and there they stayed. Von Braun became the first director of the Marshall Space Flight Center when it was founded in 1960, and the Saturn V that carried Apollo 11 to the Moon in July 1969 was designed under his direction by an organisation whose senior engineering cadre had learned its trade on the Peenemünder Haken. The Soviet Union took a much smaller share: only a few of the Peenemünde staff, including Helmut Gröttrup and Erich Apel, signed Soviet contracts in October 1946 and were then forcibly transferred east under Operation Osoaviakhim.

The reckoning came late and it came unevenly. Arthur Rudolph, who had run development and fabrication at Peenemünde, who had written in April 1943 that the operation of Fertigungshalle 1 could be conducted with prisoners, and who had been the operations director of the Mittelwerk, went on to manage the Saturn V programme and to receive NASA’s Distinguished Service Medal. The standard accounts record that his wartime role was investigated by the United States Department of Justice decades afterwards, that he left the country and surrendered his American citizenship, under an agreement with the Department of Justice that avoided formal denaturalisation proceedings; the detail of it lies outside the published record and readers who want it should go to the specialist literature. Dornberger, held at Trent Park and later in British custody, was released and went into the American aerospace industry, spending the bulk of his later career with Bell Aircraft, where the work he was associated with fed into hypersonic research programmes.

Neufeld’s judgement on von Braun is the one that has worn best, and it is deliberately unresolved. He probably joined the Party and the SS under pressure rather than out of ideological conviction; that does not sever his connection to what was done in the programme he directed; to what extent we can hold such men responsible for those crimes is debatable, but we simply cannot remember them only as space heroes. It is worth adding one plain observation to that. The men who walked out of Reutte in May 1945 had, between them, watched a labour camp burn a mile and a half from their houses in August 1943, and had then transferred their production to a mountain in the Harz. Whatever the correct apportionment of guilt, none of them was ignorant.

Peenemünde today

The last A-4 was fired from Peenemünde in February 1945. On 31 January that year Kammler’s order to relocate to Thuringia had come in by teleprinter, and from 17 February the evacuation trains ran, the first carrying 525 people to Bleicherode, Sangerhausen and Bad Sachsa. On 5 May 1945 Soviet troops of Rokossovsky’s 2nd Belorussian Front took Swinemünde and the whole of Usedom, and Soviet infantry under Major Anatole Vavilov walked into the installations to find them, by their own estimate, seventy-five per cent wreckage, with the research buildings and the test stands demolished by the departing Germans. Between 1948 and 1961 the Soviets destroyed most of what remained of the technical facilities. The power station, the airfield and the railway to Zinnowitz survived because they were useful. The peninsula spent the next forty years as a closed military zone of the German Democratic Republic, its Soviet and East German garrisons living among ruins whose purpose was not much discussed.

The power station outlasted everything. Built by the Siemens Schuckert power plant division from late 1939 and designed for 33 megawatts, it went on generating until 1990, and it is now the largest industrial monument in Mecklenburg-Vorpommern and the home of the Historisch-Technisches Museum Peenemünde, established in 1991. The museum received something in the region of 222,000 visitors in 2008, a great many of them school parties, and has taken around 6.5 million euros of renovation investment with further work planned. It holds the Coventry Cross of Nails, awarded in 2002, and a Europa Nostra award of 2013, and it is an anchor point on the European Route of Industrial Heritage. The turbine hall is used for concerts during the Usedom Music Festival. A museum in a power station that was built to make rockets, holding a peace prize from a city Germany destroyed, is a fairly exact emblem of how modern Germany has chosen to handle this material.

What a visitor sees is a landscape rather than a building. The museum’s Denkmal-Landschaft, its monument landscape, is a circular route of about 25 kilometres with twenty stations across the 25 square kilometres of the subdistrict, taking in structures from the seventeenth century to the twentieth and presenting the relationship between people, nature and technology rather than a heroic narrative of engineering. The liquid oxygen plant stands in ruins near the entrance. At Test Stand VII, where the A-4 first flew, the elliptical earthwork, foundation walls, traces of the rail approach and the water filled exhaust shaft all survive, and the centre of the stand is marked by a surviving hydrant and a memorial stone recording that this was the launch base of the A-4 rocket. In the museum’s permanent exhibition are the engine and bent propeller of Lancaster DV 202, lifted from the Kölpinsee in 1991. The rest of the aircraft is still in the lake.

The memorial that matters most is not at Peenemünde but at Karlshagen. A group of prisoners who died in captivity, probably from the Luftwaffe camp, had been buried in a mass grave beside the Peenemünde cemetery; they were exhumed in 1968 and reinterred at the edge of the Karlshagen cemetery, where the Mahn- und Gedenkstätte was built in their honour, with a mosaic by Klaus Rößler executed between 1969 and 1970. Through the East German decades this was the central site of antifascist commemoration for the research establishments, and the mosaic carries the visual language of that period plainly on its face. Since 1995 stone crosses have stood alongside it commemorating the victims of the air raids as well, and the memorial stone records the two thousand dead of 17/18 August 1943 and 18 July 1944 together. A site that was designed to make one argument now makes several at once, which is arguably the most honest thing about it.

For anyone intending to go, the practical position is that the museum sits on the northern tip of Usedom, reachable by road and by the island railway from the resorts along the Baltic shore, with the memorial landscape walkable or cyclable from the site and the Karlshagen memorial a short distance to the south. Opening hours, seasonal closures, admission arrangements and the accessibility of individual stations on the monument route all change from year to year, and some of the outlying structures are on ground that is fenced or unstable. Check before travelling.

The anniversary clock

Hydra has an unusually well defined anniversary, and it falls on a date that behaves oddly in print. Because the operation straddled midnight, it is always written as 17/18 August, and both halves of it belong to the event: the briefing, the take off and the outward flight on the 17th, the marking, the bombing, the fighters and the losses on the 18th. Commemoration at Peenemünde and at Karlshagen has therefore tended to settle on the 18th, the day the dead were counted, while British commemoration has tended towards the 17th, the day the crews walked out to their aircraft. Both are correct. The eightieth anniversary in August 2023 produced the usual crop of features, museum events and veteran recollections, and it was the last such anniversary at which any first hand British memory was likely to be available in quantity. The eighty fifth falls on 17/18 August 2028.

That matters more than it sounds, because the way anniversary material is produced has its own calendar. Broadcast and publishing commissioning for a round anniversary is typically settled between twelve and eighteen months in advance; museum exhibition planning runs longer still, often two to three years; and academic conference proposals and edited collections are usually locked a year out. In practice this means that the shape of what will be said about Peenemünde in August 2028 is being decided in 2026 and 2027, by people making choices about which of the contested figures to use, whether to lead with the Master Bomber or with Trassenheide, and whether to treat Dora as a footnote to the raid or the raid as a prologue to Dora. Those are editorial decisions with consequences, and they are being taken now.

One constraint on all of it is worth stating, because it will shape the visual record of every future anniversary. The photographic archive of this operation is finite and it is closed. There will be no new pictures. The pre raid reconnaissance frames from April and June 1943 exist and are in the public domain; the target photograph and the site plan from the National Archives exist; there are Bundesarchiv images of the settlement, of the forced labourers and of rockets on the stand; there are IWM photographs of the aircraft types and of the men who flew them. What is conspicuously scarce, and what every editor discovers in turn, is a freely usable post-raid damage photograph of Peenemünde. The reconnaissance cover flown a few days later, which Searby described as revealing extensive damage to all sections of the establishment, survives in institutional collections whose licensing is restrictive, and none of it has reached the open repositories. The result is that the raid is illustrated, again and again, by pictures of the target before it was hit. There is something apt in that. Hydra is a story about people looking at photographs of Peenemünde and trying to work out what they were seeing, and more than eighty years later that is still, in the most literal sense, what we are doing.

Licensing the imagery of the Peenemünde raid

The photographic record of Operation Hydra is finite, closed and unusually concentrated. There are the reconnaissance frames that made the case for the raid, the vertical photographs taken during the attack itself with bombs bursting among the buildings, and the Mosquito coverage flown afterwards that showed the craters, the wrecked test buildings and the flattened housing estate. High resolution, print-ready scans of much of this material, including the post-raid verticals of the destroyed and damaged buildings, the No. 540 Squadron Mosquito photography of the damage to the Luftwaffe test buildings, bombs exploding beside the production works during the raid, and the destruction of the Karlshagen estate, are available for licensing from the Piemags collections on Alamy, together with archive photography of the V-2 itself. Enquiries for publishing, broadcast and exhibition use are welcome directly through piemags.com, where identifications are checked against museum and archive records rather than repeated from the captions that circulate online, and where pricing is negotiable, and we will match or beat what you pay other aviation agencies for comparable historical imagery.

The same archive supplies the modern era. Aviation Photo Crew’s air-to-air photography covers the warbird and front-line fleets, including the last flying Lancasters’ contemporaries and the aircraft that carry the bomber offensive’s memory today, all licensable through professionalaviationphotography.com with terms on the licences page. Picture researchers working on the bomber war may also want our epics on the Dambusters raid and the Doolittle raid, the type histories of the Avro Lancaster and the Handley Page Halifax, and our guide to aviation picture research. For the 2028 anniversary cycle, early enquiries are sensible; the strongest frames are already being asked for.

Frequently asked questions