by Andrew Brown
Yours sincerely,
J.D. Bernal33
At the board meeting, Bernal gave his progress report on Habbakuk, and then listened to Admiral King’s team attack the project. Bernal thought they had not made the best of their case, and with typical impartiality and eloquence, put forward the arguments against on their behalf. When word of this Jesuitical performance reached Churchill, he sent a telegram to Mountbatten saying: ‘The next time you come to a Combined Chiefs of Staff Conference, you must not bring your scientific advisers with you.’34
A day or so later, Perutz arrived in Washington and was surprised to be greeted at Union Station by the whole British Habbakuk team, who ‘seemed in no hurry to get back to their desks’.35 He imagined they would be ‘busy sixteen hours a day with the planning of the bergship’s construction’ but soon learned that work was at a standstill until the US naval engineers issued their final report. The tide was ebbing fast for Habbakuk. Mountbatten listed several reasons36 – the great demand for steel; permission from Portugal to use airfields in the Azores which facilitated the hunting of U-boats in the Atlantic; long-range fuel tanks that allowed British fighters extra time to operate over France; and the American preference for conventional aircraft carriers. To these should be added Mountbatten’s own withdrawal from the project. The final Habbakuk Board meeting took place in December 1943, and Pyke’s unsinkable vessel was melted down in a few tepid words: ‘The large Habbakuk II made of Pykrete has been found to be impracticable because of the enormous production resources required and technical difficulties involved.’37
Mountbatten succeeded in bringing Bernal back to London, after he had spent about one month in Washington. Although Sage had expressed some interest in joining the SE Asia command, he was soon fully committed to Overlord or more precisely to Operation Neptune, which was the codename for the maritime part of the invasion. Never before had such a huge seaborne force been assembled, and the problems of landing thousands of troops with their equipment on defended beaches were completely novel. The most infamous amphibious landing in military history, at Gallipoli in the First World War, was the only precedent for attempting to put large numbers of troops ashore without first capturing a port. The long shadow of that tragic debacle, apart from causing Churchill continual nightmares, had provided the impetus for the inter-service collaboration that culminated in the formation of Combined Operations. There must never again be such amateurish preparation, tactical dithering, lack of troop protection and lack of proper equipment when putting an attacking force ashore. To make Overlord successful, there needed to be detailed knowledge about the nature of the Normandy shore – safe approaches for the ships and landing craft that were free from mines or rocks, the range of the tides, mean high and low water marks, underwater obstacles and fortifications on the beach, beach gradients and dimensions, exit routes from the beaches that would support tanks and heavy equipment, the topography of the hinterland. To preserve the vital element of surprise, all this intelligence needed to be collected without alerting the enemy.
When Sage returned to COHQ in late September, there was no specific information available on any of the above issues. How he came to be in charge of providing the necessary data is not clear, but having been the only scientist present at Quadrant, he was perhaps the obvious choice. He certainly had Mountbatten’s full confidence and had made a positive impression on the Chiefs of Staff. Pyke was fond of quoting Chesterton’s Father Brown stories to the CCO, especially the saying: ‘It isn’t that they can’t see the solution. It is that they can’t see the problem.’ Faced with numerous imponderables in the planning of Overlord, who better than Sage to define and elucidate the problems? His strength as a scientist was always to have the first word on a subject, before handing it over to someone else with more doggedness, for the ultimate solution. Here was a role that would test his multi-faceted brilliance, where it was vital that no potential pitfall should go undiscovered, and where every dimension of each problem needed to be examined and related to one another. By 9th October, Bernal was able to report to Mount-batten that ‘good progress is being made on the provision of harbours, information as to swell etc.’38
Mountbatten’s departure from COHQ in October brought the opportunity to review the structure and function of Combined Operations. The COS appointed Air Marshal Bottomley to carry out the assessment, which he quickly did. To the annoyance of the Admiralty, he recommended that, with minor pruning, COHQ should continue much as before. The new CCO, Major-General Sir Robert Laycock, arrived in December. In the light of Bottomley’s report, Laycock decided that Pyke should be transferred to the Admiralty, and that he would make do with Bernal as his sole scientific adviser (Zuckerman was still in the Mediterranean, where he had proved himself indispensable to the American and British airforce chiefs in the invasion of Sicily). In fact, the Admiralty had no work for Pyke, and he was to sit, ever more despondently at his desk, before quitting COHQ a few months later.39 Sage’s central role at COHQ was not diminished by the change in leadership; for example, when the Graham Report to the COS on ‘all existing means of providing fire support when landing forces on a heavily defended coast’ surfaced just before Christmas, he was one of two members of the Assault Warfare Committee at COHQ charged with gauging its impact and making sure its findings were being incorporated into planning for Overlord.40
When Bernal took his first glance at the charts for the Overlord landings in Washington at the end of August, his immediate reaction was to recall a holiday visit he had made to Arromanches, ten years earlier. He remembered swimming in turbid water, due, he thought, to a suspension of peat. Starting his research in the obvious place (one previously overlooked by British Naval Intelligence), Sage consulted Le Guide Bleu where he read: ‘The beach at Arromanches is very sloping, indeed the sea goes out a long way and at extreme low tides the peasants go out to the end of the beach and pick up a material they call “gourban” which they use to manure their fields.’41 Sage asked a geologist in London whether any British beaches were laid on peat, and was told that Brancaster, on the north Norfolk coast, was certainly one example.
Bernal quickly came to realize that virtually nothing was known about the interaction of waves and beaches, and was fortunate to enlist the help of Brigadier Ralph Bagnold, who had recently returned from service in North Africa. In the 1930s, Bagnold had planned and carried out a number of expeditions through the great deserts of the Middle East, and in 1941 published a book, The Physics of Blown Sand and Desert Dunes. Recalled to the Army at the outbreak of war, Bagnold had set up the Long Range Desert Group in 1940, which operated behind enemy lines in the Libyan Desert providing intelligence on Rommel’s Afrika Corps as well as engaging in sabotage.42 It is probable that Bernal heard about the exploits of Bagnold’s Long Range Desert Group when he was in Cairo. Now that Bagnold was back in England, Bernal approached him to carry out experiments on the action of waves on the formation and erosion of beaches. The work was carried out at the Hydraulics Laboratory of Imperial College, London and resulted in Bagnold showing, as Bernal explained:
… on a normal gentle day the waves coming on to the beach do not produce simple circular movements of the water but have a residual progressive movement. In other words the water at the top moves forward and so does the water at the bottom. The excess water is carried away by a current flowing between the top and bottom – the undertow which is often fatal to swimmers.
In gentle weather, the current along the bottom carries grains of sand and builds up the beach and normally a beach will continually grow or, windblown, form dunes. However, in a storm, an entirely opposite phenomenon occurs. The sand is stirred up by the waves and lifted into the returning stream, to be carried right out to sea and deposited in deep water. A beach may be built up as much as three to four metres in a season and then swept clean in a single night of storm in winter. Fishermen know this very well and so do people living by the sea, but the ordinary seaside visitor only sees the beach at its
best, when it is fully built up in the summer.43
The critical unknown factor about the beaches chosen for the landings was whether they would support the weight of armed vehicles, and in particular whether trucks and tanks would be able to drive across them without getting stuck. Opportunities to take direct samplings from the beaches around Arromanches were obviously going to be limited, and so Bernal decided that he would do his best to infer the beach structure by studying its geological and recent history. To do this he needed to read obscure volumes in the British Museum and in Oxford. An arrangement was made with the Principal Keeper of Printed Books at the Museum that Sage would have unfettered access to all their books and take away whatever he wanted, without any ticket being issued, on the understanding that he would eventually return the volumes to their correct places.
Sage read every volume of the Proceedings of the Linnaean Society of Caen, starting in the 1840s. He ‘paid particular attention to the accounts of the summer excursions, which were usually made to a part of the seaside. It’s true that much of the material was concerned with the dinner and with the speeches made there, but in the daytime there were excursions of geologists, of botanists, of zoologists, who brought back information very valuable to me: the character of a snail or of a plant would indicate the nature – marshy or otherwise – of the particular piece of country.’44 While reading these prosaic descriptions of the naturalists’ rambles, in his mind’s eye Sage was picturing tanks and army lorries making their way over the land.
One of the most significant contributors to the Linnaen Society of Caen was Abbé Huc, a local priest, who achieved fame as a missionary to the Far East. He was fascinated by the geology of the Bay of Calvados, and described in one article how he awoke after a tremendous winter’s storm to find that all the sand had been stripped off the underlying peat. The Abbé wandered around at low tide and found ‘a coin which bore the distinct imprint of the features of the Empress Crispina, thereby dating a period at which this had been farmland as no later than 230 AD’.45 From this account, Sage pictured the Calvados reef as marking the headland of a pleasant wooded valley, rocky only at the edges and marshy in the middle. Much of the beach and shoreline was laid down on a base of clay and peat, the residues of an ice-age forest. The land was likely to be treacherous and Bernal’s notes show that ‘soft ground’ and ‘beach tests’ were his top concerns in November 1943.46
There were other unlikely sources of recent shoreline history. The Roman de Rou, a twelfth century Norman epic by Maistre Wace, Sage regarded as a ‘somewhat propagandist account of the escape of William the Conqueror, before he had earned that name, from a castle near Cherbourg to a friend of his at Ries’.47 The journey was carried out at night and alone, and Sage was impressed by the detailed knowledge of the local countryside shown by the author. William ‘crossed, at low tide, what the writer calls Le Grand Gué, and which is now called La Baie des Ryes. The writer indicated that a ridge of rock and gravel crosses this muddy bay, and enables one at extreme low tides to avoid a detour of some forty kilometres.’48 Bernal was able to arrange for the RAF to take reconnaissance photographs of the bay at low tide ‘which amply confirmed the existence of the submerged causeway that William took’.49
As he looked at old maps, Sage was puzzled by the name given to a small low-lying promontory – Hable de Heurtot. Thinking that this name should indicate a harbour, he investigated further and came across a fourteenth century lawsuit between the Le Sieur de Courseulles and the King, who claimed the Hable de Heurtot was being used to evade the tax levied at royal harbours. Bernal then found a report of a great storm in the sixteenth century that had completely silted up the mouth of the harbour so that it could never be cleared.50 This ground was clearly unfit for tanks.
While the military planners were no doubt impressed by Sage’s erudition and scholarship, with so much at stake they wanted some direct evidence from the Normandy coast. If the ground were too soft to support wheeled vehicles, miles of metal tracking would have to be manufactured in the USA and shipped in time for the planned summer invasion. In early November, Sir Bertram Ramsay, the Scottish admiral in charge of Neptune, issued a memo setting out the urgent need for a survey of the proposed positions of the Mulberry harbours.51 He thought reconnaissance should be carried out during the moonless nights at the end of the month by naval force J, with Combined Operations providing suitable cover. Data should be recorded on the depths of water and nature of bottom, with samples and borings taken at frequent intervals if possible. The strength of tidal stream, any corrections to the silhouette of the coastline, any information about the nature of the beach and its rocky outcrops should be noted.
One of the men picked for this crucial and dangerous reconnaissance was Logan Scott-Bowden, a 24-year-old major in the Royal Engineers. He went to COHQ to be briefed by Professor Bernal, who struck him as a venerable professor.52 On his overflowing desk were some Latin documents. Bernal explained to the observant young officer that they were reports on fuel reserves for the Roman Empire. The Romans used peat as a fuel and had carried detailed surveys of where it could be found in northern Europe. Secrecy about Overlord was imperative and Churchill would not authorize any reconnaissance of the French shore until the exercise had been tested in England first. Bernal was skeptical that commandos like Scott-Bowden could swim ashore and carry out a meaningful beach survey without detection by sentries. It was decided to hold a trial on Brancaster beach with Sage as one of the sentries.
Scott-Bowden and his sergeant, Bruce Ogden-Smith, were brought offshore by a landing craft and swam a mile or so to the beach in freezing conditions. They wore primitive wetsuits made of canvas-like material, which covered their heads and were sealed at the ankles. They saw Bernal standing guard next to an Army truck, and ‘crawled around him, making notes on what he was doing and collecting soil samples’.53 As the night wore on, both soldiers felt admiration for the professor who was prepared to endure the cold, windswept beach. After about four hours, they stood up and shouted at Sage, who had not seen them at all; they were able to tell him what he had been doing, with great precision. Exhausted, they all climbed into the Army truck and drove down to London, dropping Bernal off at 6 am. He reported to Churchill, through Cherwell, that the Brancaster rehearsal had been a success and Churchill decided that the Normandy beach reconnaissance could go ahead. He also instructed that it should take place on New Year’s Eve, even though this was not a completely moonless night, because the Germans would be celebrating.
One of Mountbatten’s personal appointments to COHQ was Sir Malcolm Campbell, the land-speed record holder. In Sage’s opinion, he was an inspired amateur with unrivalled knowledge of beach surfaces, as a result of his speed trials on sand and salt flats. Campbell had devised various formulae and reckoned that a depth of fourteen inches of sand was sufficient to support wheeled vehicles, even if there was soft clay underlying. Campbell was a talented mechanical engineer with his own private workshop, and he constructed various gadgets for the raiders to use. One difficulty was that Campbell thought he should go on the reconnaissance and took some persuading by Scott-Bowden that he would probably not make it to the beach swimming in the frigid water.54
After crossing the Channel in a Motor Torpedo Boat (MTB), Scott-Bowden and Ogden-Smith swam ashore. Both men carried ‘a dozen long tubes, numbered in phosphorous on the top, in a bandolier on our backs, and the idea was to make a sample from various parts of the beach, noting on the underwater writing tablets which were strapped to our arms approximately where each sample had been taken’.55 After crawling along the beach below the high water mark so as to leave no tracks, and avoiding detection in the sweeping beam from a lighthouse, they found the area where they were due to carry out a detailed survey. Scott-Bowden described the data collection:
We came across a large outcrop of peat, which had been suspected from aerial photographs, but by and large we thought the area would turn out all right from the point of view of bearing capa
city. When we had each filled about eight tubes I reckoned we had got enough and so I said ‘Let’s go.’ That’s when the trouble really started, because the breakers were quite heavy and we were positively bogged down with our bandoliers and all our other kit.56
After struggling to break through the incoming breakers, the two soldiers swam out and managed to rendezvous with the MTB; they were pulled on board, cold and exhausted, with their precious samples. The samples confirmed what Bernal had been saying about the subsoil of peat and clay.
Writing to Mountbatten a week later to explain why he could not visit the new HQ in Delhi, Bernal mentioned ‘the necessity to attend personally to detailed briefings of recce parties which depend on moon and tides’.57 For the next operation, at the end of January 1944, the Combined Operations Pilotage Party arrived off the French coast in a midget submarine that had been towed about a third of the way across the Channel by a trawler. For several days, detailed observations of the cliffs at Pointe du Hoc that would be such a deadly obstacle for the Americans on Omaha beach were made through the submarine’s thin periscope. Their eel’s-eye view was used to supplement information from aerial photographs. At night, further beach samples were taken, following which Scott-Bowden had to struggle out of his wetsuit in the battery compartment of the submarine, where there was only eighteen inches of headroom.
Bernal’s 1944 desk dairy lists ‘trials at Brancaster’ beginning on 11th January.58 The trials confirmed that as long as there was a layer of sand 18 inches deep covering the gourban, vehicles would not get stuck, providing that they did not follow in each others’ tracks.59 It would be necessary, therefore, to clear mines over a wide front, and not just be content to provide a narrow passage. It was concluded, nevertheless, that the beaches at Arromanches would support tanks, providing the mines were cleared successfully and the tank drivers carefully briefed.