The public inquiry into the loss of the Comets G-ALYP and G-ALYY sat for twenty-two days in the autumn of 1954, sixteen of them at Church House, Westminster, and the last six at the Institution of Civil Engineers across Parliament Square. Lord Cohen presided, with Sir William Farren, Professor W. J. Duncan and Air Commodore A. H. Wheeler as assessors. Sixty-eight witnesses were heard, 145 exhibits put in, and the shorthand transcript ran to 1,600 pages and some 800,000 words. The daily transcript is at the National Archives; the fullest account in print is the weekly report in Flight, written by its representative at the hearings, and the words of the witnesses and counsel below are taken from it. The entry gives the story; this piece and its sequel, “The makers’ case”, give the hearings in order. This one covers the first three weeks, in which the Crown put its case through the Establishment, the airline and the two boards.

The first week: 19 to 22 October

Sir Lionel Heald, for the Crown, opened on Tuesday 19 October with a complete explanation of the Elba accident, which he proposed to prove “by means of a closely reasoned and documentary argument, which started from the known facts and proceeded by a process of elimination and deduction to a conclusion which could actually be shown to be sound by visual inspection of certain parts of the wreckage”. He said that time after time in the history of science “some radical new departure” had led to a disaster that seemed to close the road, and that men had never been afraid to re-examine the whole project afterwards. No aircraft in history, he said, had been subjected to such an examination. He gave the facts of both flights, the recovery of 70 per cent of the Elba aircraft by weight, the engines’ “completely clean bill of health”, and the chain that led to “the demon fatigue”: “Everything in Sir Arnold Hall’s investigation supports the theory that this was the cause of the failure of the cabin, and nothing contradicts it.” On the makers he said: “No one suggests for one moment that anyone could have known this in 1946 or 1951 in Messrs. de Havillands or the Air Registration Board.” The Board had been satisfied that the stress at the worst points would be “only about 40 per cent to 50 per cent of the ultimate stress of the material”; the Establishment had found 70. De Havilland’s comparable section had “broke down only after 54,000 hours”, but the result “might have been misleading, due to the possible strength-increase of the material caused by earlier static loads.”

The second day built up the flight of Yoke Peter from the ground staff at Ciampino, the log of her radio messages, the Argonaut captain who heard her last words, and the affidavits of the people of Elba. On the third, Thursday 21 October, Commander Forsberg described the naval search, Professor Fornari of Pisa the bodies, and Hall began. He explained the Establishment’s four lines of attack: the wreckage, the fatigue tests on Yoke Uncle, the flight tests on G-ANAV, “operated at all times in company with a Canberra aircraft which flew behind it in order to keep general observation on it”, and the models. From the wreckage, “we first formed the impression that the aircraft broke up probably into about six main pieces.” The source of the cabin failure was a point near the rear aerial window, from which every fracture ran away and into which none ran. A piece of Yoke Uncle’s fuselage with the same kind of compression buckle was shown beside it. “Is it fair to say that there is a very close correspondence between those two fractures?—Yes, I think that is correct.” The one-tenth Perspex cabin and its dummies had shown “complete chaos” a tenth of a second after bursting. Sabotage, an accidental explosion, high tail loads, the autopilot, over-control and gusts were each taken up and set down.

On Friday 22 October Hall gave the cause its name: “a phenomenon known as fatigue. The essence of the phenomenon is that whereas the structure will stand one application of the load quite satisfactorily, it may not stand many hundreds, thousands or millions of applications of the load satisfactorily, and may in the end fail under a load which it is well capable of bearing when new.” Could one know what was going on beforehand? “So far as I am aware it is not possible to know what is going on. I know of no physical test one can apply to a piece of material which will tell one how much it is fatigued.” Identical specimens, he said, scattered: a weak one might fail at a third of the average life and a strong one last three times as long. At the escape hatch of Yoke Uncle the peak stress was 45,700 pounds a square inch against an ultimate of about 65,000: “failure occurred at 70 per cent of the ultimate.” “The margin or safety factor is not a very large one?—It is not.” Round the aerial windows of Yoke Peter he had seen cracks drilled at their ends during manufacture. “Although you may see a crack which is formed by such a process, one is always a little worried as to whether there are some you do not see. . . . I cannot totally eliminate that they might have accelerated the onset of fatigue in this aeroplane . . . it might have brought the accident a little forward in time; it certainly was not in any sense a primary cause.” On the three aircraft: “If we take YU as unity, YP was 2.4 times younger and YY was 3.4 times younger . . . At the time of the Elba and Naples accidents, both aircraft had flown longer than the safe fatigue life we could give in the light of the evidence obtained from YU.” Dr Walker described the tank, the failure at 9,000 hours and his opinion that it was fatigue at a rivet hole. The Corporation’s fleet inspector, asked whether he would have expected the makers to point out manufacturing cracks when delivering an aircraft, said “we would hardly expect to get a new aeroplane with cracks in it.”

The second week: 25 to 29 October

On Monday 25 October Hall was recalled and said that whether or not the cracks had hastened it, “there is no doubt that the aeroplane YU showed a proneness to fatigue in the cabin, which would have overtaken it quite independently of any other manufacturing crack there might be.” Pressed by Shawcross on which corner the failure had begun at: “I get to this square foot of this aeroplane near the A.D.F. window and I say, ‘That is where it first failed.’ It is now, when we start to get down to the square inch from the square foot that we have difficulty.” Dr Moult of the de Havilland Engine Company accounted for the four engines: no seizure, no heat, no burst disc; the hub shaft of one had broken under a sudden rotation of the whole wing, a gyroscopic load, and the other three were cracked in the same places. The witnesses for the Naples flight followed, and the captain of Eagle. On Tuesday the Court went to Farnborough; on Wednesday it considered in private what it had seen.

Thursday 28 October was the day of the Air Safety Board. Sir Frederick Bowhill read the minute of 5 March recommending the return to service, and Mr J. M. Shaw, for the relatives, went through it. “When you passed this Minute, what did the Board consider was the probable cause of the Elba disaster?—We did not know.” “Did that include some failure of the pressure cabin?—We had considered that.” “So at the time when you passed this Minute the position remained that the cause of the Elba crash might have been due to some defect in the design or manufacture of the aircraft?—We did not think so.” “If you did not think it was due to that what did you think it was due to?—We did not know. I mean to say we were not certain. We thought the modifications which were introduced would cover all the points which were humanly possible to prevent any further accident.” “You had in mind the possibility of a common failure on both these aircraft?—There might be, yes.” “And yet you recommended the Minister that they should be allowed to fly again?—That is what I have done in this Minute.” “How many aircraft would have to crash from an unknown cause before the Air Safety Board recommended they should not be allowed to fly any more?” Lord Cohen stopped it: that sounded to him “very like a question” for a jury “rather than a proper question to address in this Inquiry.” Mr MacCrindle, for the pilots, asked whether “everything humanly possible” was not “putting it much too highly”, and whether it would not have been humanly possible to wait for more wreckage. “Well, we could have.” “It would have been humanly possible, would it not?—Yes, it would have been humanly possible.” Sir Frederick added that at the time he had thought it rather unlikely that any further wreckage would be recovered.

Charles Abell of BOAC then gave the history of the London Airport examination: the six possible causes, the modifications, the eighty-two recorded pressurisation defects on Yoke Peter, none of them serious, the fuselage cracks round the door caused by passenger steps, the twenty-two loose bolts in Yoke Yoke and the reprimand of the two men concerned. The committee’s decision to resume, he agreed, had been joint, and the cause was still unknown. On Friday Mr Ripley explained the reconstruction: water damage and fire damage that did not cross the same breaks, the distinctive edges of bending, tension and fatigue, and the marks of one piece passing over another. He had found “no signs whatever in the wreckage which would suggest that the failure of Redux contributed to the cause of this accident.”

The third week: 1 to 5 November

Hall came back on Monday 1 November to explain the Board’s decision of March. The withdrawal of a certificate was the Registration Board’s business, and the Safety Board had started from its view that the aircraft was airworthy. “The first one was that the cause of the disaster at Elba had not been established,” and the question was whether it ever would be, “and that, I suppose, depended (it seemed to me at the time) on what was the possibility of recovering a substantial part of the wreckage.” The Calcutta Comet, breaking up at 7,000 feet, had been spread over five miles; from 30,000 feet Yoke Peter’s should have lain along twenty, back towards Rome, where the water deepened and there was an unswept minefield. In the event the wreckage lay close together, “an exceptional feature caused by the exceptional nature of the pressure-cabin disruption”, on a hard sandy bottom. He described the resources: “I deployed very large resources in terms of men and money and material. I drove my staff, and incidentally myself, well beyond normal limits. The Court may be interested to know that we had a very large deployment of industrial labour working 80-hour weeks and we had a certain number of men that we drove to 100- and 120-hour weeks in order to get this matter forward.” The first failure in the tank had been on 24 June; the aerial-window piece had come up five weeks later, on 12 August. For the future he proposed partial tests on half a dozen specimens of the weakest part, then a whole aircraft in a tank until it failed, then half a dozen of whatever had failed: with six specimens one could put the scatter at three to one either side of the average, whereas “if you only work on a single specimen, you would have to give a safety life of about one-ninth of what the specimen comes to, because the specimen might by chance have been the strongest.” “I am sure it can be cured, and I hope it will.” After the first failure Yoke Uncle had been repaired and run on to a second catastrophic failure at about 17,000 hours. The main new knowledge “was that a specimen which passed a static test of twice its normal loading could nevertheless have in it stress concentrations of 70 per cent of the ultimate.” If he had known of the manufacturing cracks in March, would his advice have been the same? “No, it would not.”

On Tuesday 2 November Shawcross took him to the test of 1953, and he gave the answer with which the entry opens: the 54,000 hours had been “achieved because the specimen had been pre-loaded in a static test to 13½ lb on several occasions, and to 16 lb on some occasions”, and “if one got 54,000 hours on a pre-loaded specimen, I do not know of any basis on which one could then predict that the fatigue life of an unpre-loaded specimen would be X at all.” The company had been “in no sense in serious violation of good practice or of general opinion, although they were probably in violation of certain areas of opinion.” In the afternoon Lord Brabazon told the Court that the Calcutta accident had been “a strange accident”, the Elba accident “a completely unexplained accident to us at that time”, and that “if you grounded the type of every aeroplane that had an unexplained accident you would scarcely have a machine in the air today.” Two Tudors had been lost unexplained, modified on every possibility, and never lost again. Then he addressed the Commissioner. “You may even try to find a scapegoat. I am quite prepared to be that scapegoat, because it is true to say, is it not, that the Air Registration Board, of which I am Chairman, could have forbidden the Comet to fly. You know and I know the cause of this accident. It is due to the adventurous, pioneering spirit of our race.” “Of course we gave hostages to fate, but I cannot believe that this Court or our country will censure us because we ventured.” The aircraft “deteriorated, so I am led to understand, by a slowly developing molecular metallurgical fault. It is metallurgy, not aeronautics, that is in the dock. . . . in every step in progress we have paid for it in blood and treasure, God knows that in this case we have paid in full.” Lord Cohen made “only one criticism”: the Court’s duty was not to search out troubles but to find the cause, and the far more important thing was “whether the evidence discloses that the trouble can be put right and the Comet can take the air again”.

On Thursday 4 November Hall said of cracks: “The whole question of what we might call crackology depends entirely on what is the stress in the vicinity of the crack and in what sort of stress field the crack finds itself. If it is pointing towards a rising stress, plainly it is getting encouragement to go ahead.” A properly designed aircraft would not be imperilled by an accidental crack of a fifth of an inch; “if an aircraft was such that a crack of 0.2in did imperil it, then there must be something wrong about it.” Walker, cross-examined on the tank, would not treat the eighteen windows as eighteen specimens: “That is really cheating . . . as far as I am concerned, the aircraft is the specimen.” Had he known nothing of Elba and Naples, what would he have made of the tank? “We should have said that any Comet aircraft at that number of hours (2,700) was not safe to fly.” Knowing both, what safe life would he give? “I think I would say it was not worth flying at all.” Lord Cohen put it for him: the design must be modified, and no life could be given until the modifications were known. “Exactly.” Allowing for everything the tank could not represent, “If my life depended on giving an estimate I think I would pull that 9,000 hr down to perhaps 7,500.” Sir Victor Tait gave the fleet’s hours, 25,408 up to Naples, and the four oldest aircraft: Yoke Uncle 3,521, Yoke Peter 3,495, and two others. On Friday Walker said of the company, “I think they could have come and consulted us more than they did. I am not saying it was a serious default”, and Walter Tye of the Registration Board began the history of the requirements: the full certificate issued on 22 January 1952, the six minor deviations accepted, and the advice given to makers that proof and ultimate factors of 1.33 and 2 on the working pressure “were advisable in order to combat the relatively unknown factors of fatigue, crack propagation, creep, etc.”