David N. Thomas and Dr Simon Barton

First published in 2004 in Dylan Remembered 1935-1953, Seren

Alcohol, diabetes and drugs have all been indicted in the death of Dylan Thomas in St Vincent's hospital, New York, on November 9 1953. But our investigations lead us to conclude that Dylan's death was brought about by a serious chest infection in which chronic bronchitis developed into terminal pneumonia before his admission to hospital.

Death was also a consequence of cerebral hypoxia i.e. insufficient oxygen being delivered to the brain cells, causing them to die. This caused cerebral oedema – swelling of the brain tissue and cells. The hypoxia was a result of Dylan’s inefficient respiratory system, which had been impaired by the bronchitis and pneumonia, and then by injections of morphine given by his New York doctor, Milton Feltenstein.

Alcohol certainly played a part in Dylan's death but it did not directly kill him. Over the long-term it contributed, together with poor diet, smoking and lack of sleep, to the general debilitation which allowed bronchitis and pneumonia to take hold and flourish.

A critical factor in Dylan's death was medical negligence. Dylan's bronchitis and pneumonia went undiagnosed and untreated in the days before he was admitted to St Vincent's hospital as an emergency case. There was also a significant delay in getting Dylan to St Vincent's. He was not examined by a specialist or consultant until some thirty-seven hours after his admission.

We begin with an outline of Dylan's medical history. This is followed by a brief note on Dylan's last days in New York. We then describe his admission to St Vincent's hospital, and examine his medical data. We look at the three causes of death established at the post-mortem - pneumonia, pial oedema and fatty liver - followed by a discussion of how, if at all, alcohol, drugs and diabetes were implicated in Dylan's death. The post-mortem report is reproduced on this website, along with a Chronology of the events in Dylan's final illness and death. Before continuing with this chapter, it may be helpful to read the Chronology because it offers useful diagnostic clues.

1. Dylan's Medical History

Dylan's history can be assembled from his letters, biographical accounts and the interviews carried out by the radio journalist, Colin Edwards.[1] From early childhood, it is a chronicle of bronchitis, influenza and colds. Dylan was a 'chesty' boy, teenager and adult, and smoking aggravated his list of bronchial complaints. There were endless, nagging worries about TB, not least because his family, on both sides, had a history of bronchial problems and related deaths.

Throughout his life, Dylan tried as hard as he could to avoid doctors, and to avoid taking their advice when he did see them. The family doctor at Laugharne, David Hughes, told Colin Edwards that "I have no recollection of Dylan ever consulting me about his health...I was never asked to attend Dylan medically. Although I believe that occasionally he did consult a doctor in London."[2]

Bronchial Woes

In his 1955 book, John Brinnin has helpfully documented how Dylan "coughed and flushed" his way through his American tours. Here are two typical examples, taken from Dylan's first visit in 1950:

"Several times already that morning he had fitfully broken into spells of coughing that racked the whole length of his body, brought tears to his eyes, and left him momentarily speechless...These attacks were as a rule brief, did not seem to alarm him, and he recovered always within a few minutes, seemingly undisturbed by a collapse that would have sent almost anyone else to bed."

"...barely five minutes before he was to go on stage, he was overtaken by a coughing attack so violent I had to hold him to enable him to keep on his feet...As we moved on, he became ill again and began to cough in a spasm so blinding it seemed he would break asunder."[3]

Dylan’s wife, Caitlin, also has described his coughing fits at the Boat House in Laugharne, as have other friends and visitors who talked with Colin Edwards. Dylan's letters from the "bronchial heronry", as he once put it, continue the tale of complaints: at the edge of pleurisy (November 11 1950); gastric influenza (March 20 1951); bronchitis (October 1951); pleurisy (June 28 1952); pneumonia (October 8 1952); bronchitis, no voice only croaking (November 21 1952); influenza, bronchitis, croaking and snuffling, no voice (February 6 1953). Both Dylan and Caitlin blamed the damp Boat House for these ailments, as well as the rain and wet mists of the Laugharne estuary.[4]

The 'pleurisy' and 'pneumonia' are almost certainly Dylan's exaggerations for bronchitis. All these pulmonary infections are associated with Dylan's breathlessness that was observed in Laugharne by his mother and by some of his friends. Aeronwy Thomas has also recalled her father's condition: "I can remember my father always huffing and puffing up the steep path to the cliff walk. Physically, he was a bit like a turtle who’d been stranded."[5]

There is consensus amongst Dylan's biographers about his "chesty complaints", though sometimes asthma is referred to rather than bronchitis. Nashold and Tremlett also conclude that Dylan's chronic bronchitis and emphysema were all conditions that he suffered from for many years, as he stumbled breathless around Laugharne "on his walking stick, starved of oxygen." They note that Dylan was using an inhaler to help his breathing on his visit to America in the spring of 1953.[6]

Dylan was rejected for war service in 1940 because, it is said, he had "weak lungs". This condition cannot be confirmed because the records of the war service tribunals have been destroyed. When he was examined by a doctor in 1946 at St Stephen's hospital in London, nothing abnormal was found in his lungs.

Both Dylan's maternal and paternal relations had suffered a range of pulmonary illnesses. As death certificates show, Dylan's maternal grandfather, George Williams, died of bronchopneumonia and his paternal grandfather, Evan the Guard, of chronic bronchitis. In 1953, D.J. Thomas, Dylan's father, died of bronchopneumonia, and Dylan's letters describe DJ's earlier pneumonia in 1949 and 1950. Two of DJ’s sisters died in their thirties from pulmonary TB.[7]

Broken Bones and other Woes

There is a record from childhood of falls and broken bones. From at least early manhood, there is a history of poor diet, as well as weight gain leading to obesity, linked presumably to Dylan's fondness for sweets and beer. His disinterest in food is described in many interviews with his friends.[8] Brinnin, too, has described in his 1955 book Dylan's lack of interest in eating. Dylan's poor and irregular diet is associated with his alcohol consumption.

There is at least one report, from the late 1930s, of Dylan having a venereal disease but there is no confirmation.

There is a history of vomiting, sometimes without blood and often with.[9] He was admitted to St Stephen's hospital in 1946 after vomiting blood. Both Brinnin and the poet Patrick Boland observed Dylan vomiting blood on his first American trip in 1950, followed immediately by another vomiting attack in the street, one in the theatre before giving a reading, and another afterwards. Brinnin does not implicate alcohol or drunkenness in these attacks.[10]

The vomiting seems sometimes to be associated with anxiety or apprehension, as it was in Dylan's stay in St Stephen's. Severe nausea and vomiting were a major characteristic of the hours leading up to Dylan's hospitalisation in St Vincent's hospital. Brinnin's account also tells us that vomiting and retching sometimes occurred after a coughing bout - "such shattering fits of coughing, often followed by frightful retching and vomiting, went on through all the time I was to know him." A letter from Dylan in 1943 to Donald Taylor suggests that the coughing-vomiting condition was of long-standing:

"I've got laryngitis or bronchitis or asthma or something: a complaint, whatever it is, that makes your chest like a raw steak, prevents breathing, & produces a food-losing cough."

Dylan experienced blackouts and deep sleeps in his later years from which he could not easily be roused. These are summarised in the section on diabetes below. Dylan was unwell before he left for America - his condition is described in a letter from Vernon Watkins:

"It is a great tragedy that Dylan who had been drinking much less before he left for America, was allowed to drink so much when he got there. He was not well before he went, and had had several short black-outs, and had even written for a diet-sheet; he was anxious to give up drinking. But I am sure he did not tell anyone he was ill, and I know how difficult it was to stop Dylan. Yet if one did stop him he was grateful, because he really didn't like it, except for beer."[11]

According to Brinnin, Dylan suffered from gout and alcoholic gastritis in America, but there is no medical confirmation of either. Dylan certainly believed that he had gout, and there are numerous reports of his having a very painful toe. Both Brinnin and Dylan report that he was treated for gout with ACTH, a cortisone secretant, whilst on his last two American trips. The ACTH was also used to relieve his breathing and as a general tonic. Dylan received three injections in May, two in October and three in November 1953 - the dosages of all but one are unknown.[12] There seems little to support Nashold and Tremlett's assertion that Dylan received twice-weekly injections of ACTH in May 1953 - we deal with this in the following Note.[13] There is no information to suggest that Dylan received ACTH or cortisone itself between June and early October 1953 when he was in Laugharne and London. We do not accept that the ACTH or cortisone could have played a part in producing Dylan's puffy demeanour, and we explain the reasons in the following Note.[14]

Nashold and Tremlett have suggested that Dylan had diabetes but this is speculative and is not confirmed by hospital tests. We return to this matter later in the paper.

Hospital Treatment

Jack Lindsay reports that Dylan went to a psychiatrist at a London hospital and asked to be cured of "drinking obsessions." The doctor joked and bantered with him about being a poet. Dylan did not turn up for the next appointment.[15]

Dylan was admitted to St Stephen's Hospital, London, in the early afternoon of March 10 1946. He was taken to Ward 7-B where he gave his address as 39, Markham Square, Chelsea, his religion as non-conformist, and his occupation as writer, poet, journalist and BBC worker. His hospital notes, to be found in the FitzGibbon collection at Texas, are perfunctory and incomplete. The admitting doctor's letter is not with the notes but it appears, from a letter to FitzGibbon from a Consultant, that Dylan was admitted to the hospital as a case of haematemesis i.e. vomiting blood. The notes record:

"Urine sugar NIL

Recurrent bouts of depression. Recurrent bouts of alcoholic & smokers [next word unclear]. Recurrent bouts of vomiting & paroxysmal morning cough. Recent attack 3 weeks duration. Apprehensive ++ Pulse high tension BP [blood pressure] 160/98. Tremor. Shaking hands. Obesity. Watery flesh. Dilated pupils. Sympathetictonia[?]. Chest: NAD [Nothing Abnormal Discovered when examined by a doctor]. Heart NAD - Tachycardia [rapid pulse]. CNS [central nervous system] overactive ++. Liver ↓ 2 fingers Tender. Throat Dry granular

= Anxiety."

The liver was not only tender but also enlarged since it was found two fingers down from the ribs. The phrase "watery flesh" is suggestive of oedema, and we note the presence of a generalised oedema found in 1953 at the post-mortem.

A chest X-ray and liver tests were done but the results are not reported in the medical notes. Both the chest and heart were normal when examined by a doctor. Dylan's urine was tested for sugar but none was found. No mention is made in the medical notes of diabetes, nor of any drugs or diet to treat diabetes.

As for medication, Dylan was given only phenobarbitone, as a way of calming him. The notes record that he was put on a normal diet.

By March 12, Dylan's blood pressure had fallen to 132/96. He was also seen that day by a psychiatrist who recorded: "Depression consists mostly of reactive depression with elements of hypercritical attitude. Finds sleep aided by drink - habit +".

Reactive depression is that occasioned directly by events in the external world e.g. a loss of a loved one, and is an extension of normal feelings of loss or grief. That the psychiatrist underlined "reactive" may suggest he had found sufficient reasons in Dylan's external circumstances for a state of depression, though we do not know what these might have been. Neither is it clear whether the phrase "hypercritical attitude" means that Dylan was hypercritical of himself or of people or circumstances around him.

Dylan was discharged as "recovered" on March 15 with a final diagnosis of "gastritis", and was given a prescription for phenobarbitone to be taken three times a day.

We can find nothing in these incomplete notes to support a diagnosis of gastritis, let alone one of alcoholic gastritis as put forward by FitzGibbon.[16] If the hospital had suspected alcoholic gastritis, they would certainly have put Dylan on a special diet and given him appropriate medication.

The vomited blood could have been the result of a Malloy-Weiss tear i.e. a tear in the lining of the stomach wall that can be caused by vigorous vomiting. Ingestion of excess alcohol and drugs can produce bouts of violent vomiting. Vomited blood could also come from a rupture of an oesophageal vein or from a bleeding stomach ulcer.

The collection of signs recorded in the medical notes (apprehension, raised blood pressure, tremor, shaking hands, dilated pupils, fast pulse rate, overactive central nervous system and dry throat) led the doctors to conclude "anxiety". They may well have been right but they are suggestive of other factors including the use of amphetamines, such as Benzedrine.

We conclude on the facts available to us about this admission that Dylan had

- an enlarged liver probably due to long-term alcohol abuse.

- an anxiety state of unknown cause, possibly induced by amphetamines (which also cause loss of appetite and can cause vomiting).We also note that when Dylan lived in South Leigh in 1947-49, one of his neighbours, Dosh Murray, found him to be "rather nervy." [17]

- raised blood pressure, which is not unexpected in an obese, sedentary smoker who drank.

- depression.

We also conclude that Dylan did not have diabetes in 1946.

2. The Last Days

Dylan's friends thought he looked unwell when he left London for New York. He has described how he felt "as sick as death all the way over." He arrived at Idlewild Airport on October 20 1953 to be met by his lover, Liz Reitell. He slept for most of the next few days in his room in the Chelsea Hotel, complaining occasionally that he felt unwell. He became very ill on the 23^rd - he was feverish, gasping for breath and exhausted. There was little improvement the next day ("Never this sick, never this much before") and Dr Milton Feltenstein gave him an injection of ACTH. According to Nashold and Tremlett's account, Feltenstein did not examine Dylan before giving these injections. Dylan was "desperately ill" again on October 25 and collapsed at a performance of Under Milk Wood, when he was given another injection.

Dylan remained unwell for another two days, including his birthday when he was too sick to stay at the party given for him. After making a temporary recovery, he started drinking heavily on the weekend of October 31st. The first few days of November consisted of parties, drinking and hangovers interspersed with bouts of sickness, exhaustion and sleeping. As for much of his adult life, Dylan smoked heavily and ate very little whilst in New York.

His chest problems would also have been aggravated by the city’s air quality. New York was one of the most polluted of American cities. In the four days before Dylan collapsed on November 4, air quality worsened appreciably, especially on November 2nd when the city’s smog reached a level that was a threat to those with respiratory problems. By the end of November, New York had experienced an exceptionally severe bout of air pollution that killed some two hundred people.[18]

Around 2am on November 4, Dylan left Reitell in the hotel room and went to a local bar and drank eight large whiskies before returning to the hotel. When he woke later that morning, he complained of difficulties with his breathing - "suffocating". A couple of his New York friends rang him, including Harvey Breit. Dylan sounded “bad”, he said, his voice was “low and hoarse.” Breit wanted to say “You sound as though from the tomb”, but he told him instead that he sounded like Louis Armstrong.[19]

Dylan went out with Reitell, drank two beers but returned, feeling unwell, to his hotel. He stayed in bed for the rest of the day, coughing and retching, with severe attacks of nausea and vomiting, and was looked after by Reitell. She summoned Feltenstein to the hotel on three occasions, and each time he injected Dylan with ACTH and morphine. Within half an hour of the third injection of morphine, Dylan stopped breathing properly and fell into coma around midnight on November 4/5. About an hour later, an ambulance was summoned and it delivered Dylan at 1.58am to the emergency room of St Vincent's hospital, based in the Seton Building on 11^th Street and 7^th Avenue.

3. St Vincent’s

St Vincent's hospital had been set up in 1849 by the Sisters of Charity, a religious community of women within the Roman Catholic church. Its mission had once been to provide free care but by the 1950s most patients were required, as in other private and voluntary hospitals, to meet all or some of their care costs, either by drawing on their own resources or through medical insurance. Patients with enough money or insurance could choose to be in a private room or in a semi-private room with just four or two beds. Those with few or uncertain resources were consigned to the general wards.

After leaving the emergency room, Dylan was taken in the elevator to St Joseph's East, a mens' medical ward on the third floor. In his 1955 book, John Brinnin gives the impression that Dylan was placed in a private room. In fact, there were no private rooms on St Joseph's East. It was a large, general ward divided into sections with eight, six or four beds in each, with some thirty patients overall. It had high ceilings and big windows; it was "well kept, institutional and sterile-looking, not cosy." [20]

We talked with Dave Slivka, who was one of Dylan's visitors. He does not have a clear memory of Dylan's accommodation but recalls "not that large a room, there may have been other beds but there were no other people in the beds." This suggests that Dylan was probably in a four-bed section in the general ward. He had very little privacy. Many friends, as well as strangers who came in off the street, turned up to "view" him through a glass partition that separated the corridor from the ward. Dylan's wife, Caitlin, has described the scene after Brinnin had escorted her up to the third floor for her first visit:

"I came to a corridor, crowded with people - twenty or thirty of them. I didn't know who they were or where they came from, but I realised that they were all looking through a glass partition down the side of the corridor into the room where Dylan lay..."

Caitlin then entered the ward and went up to Dylan's bedside: "I talked to him, but he didn't respond and I felt so embarrassed with all those people gazing at me through the glass; I felt as though I were on a stage."[21]

Oscar Williams, Dylan's unofficial agent in America, visited Dylan on several occasions and he, too, was extremely critical of Dylan's accommodation:

"They allowed Dylan to be in a public ward; and America’s public wards in the hospitals are terrible places...he lay there for three or four days. And he was never really moved to private quarters. And the result was that he was virtually on display and hundreds of people came, all wanting to see him, and there was no defence...I think that was one of the most terrible things that happened to Dylan, the fact that he was not protected at all from the rabble, or from the consequence of being so famous." [22]

Brinnin was Dylan's sponsor and friend. In Caitlin's absence, he took responsibility for decisions about Dylan's care in St Vincent's, and how it was to be paid for. Discussions certainly did take place about whether to move Dylan out of the general ward, as Rollie McKenna has indicated.[23] The reasons not to move him were probably to do with the cost of doing so (a private room was some $4 per day more than a ward bed), and with weighing the benefits of giving a comatose patient a private room.[24]

After all, Dylan was not expected to recover - even on November 5, the day of his admission, his friends were being told by the doctors that he "had a cerebral haemorrhage...with no better than an even chance of life."[25] By the morning of November 7, doctors were advising Brinnin that Dylan "would die within the next few hours", that he was "sinking rapidly" and that his death was not only inevitable but that "it was now also to be desired" because of the extent of brain damage.[26] The next day, Brinnin started fundraising to provide money for Dylan's medical and funeral expenses, and to provide support for Caitlin and the children.

St Vincent's was neither the best nor the worst hospital in New York, lying somewhere between the elite institutions such as New York Hospital and the municipal hospitals such as Bellevue. It's reputation at the time lay partly in the quality of its nursing. St Joseph's East was supervised by Sister Marie Consilio Lillis, a highly experienced nurse who had trained at the London Hospital, before emigrating and entering the Sisters of Charity in 1927. The ward was usually well staffed on the day shift between 8am and 4pm but, as in most hospitals at that time, there were difficulties on the evening and night shifts, for which there were just three staff for each shift. Holidays and sickness could make the situation even worse. In 1954, Priscilla Sassi worked the 4pm to midnight shift in the adjoining St Joseph's West ward, which was also a mens' medical ward with some thirty patients:

"On the evening shift, St Joe's West was usually staffed by a registered nurse, a student nurse and an orderly or a nurse's aid. During my time at St Joe's West, the registered nurse was on vacation. That is why I was there, a senior student nearing graduation. I was helped by a first year student. I don't remember any orderly - maybe he was also on vacation...there was a nursing supervisor who made rounds throughout the hospital to check up on us and who was available for any unusual problems.

Many of the patients were desperately ill. They were the victims of strokes, partially paralysed and unable to speak. Others were recovering from heart attacks or dying from cancer, cirrhosis or other diseases. The work was brutally hard, especially since we had no orderly to help lift the heavier patients. We had to prepare them for the night., change their sheets if necessary, and clean them up. I gave out all the medications, we changed dressings, irrigated bladders and colostomies, managed the IV's and on and on. I never got to write my charts until the night shift came on at midnight to care for the patients." [27]

There was a general shortage of nurses throughout New York hospitals in the early 1950s. Both the numbers of nurses available, as well as their experience, was a particular problem on general wards. We note that Dylan was a "weekend patient", admitted to a general ward on a Thursday and dead by the Monday. Staff shortages and absences, including supervisors, technicians, nurses, doctors and specialists, were a serious worry on weekends - weekend work was unpopular amongst staff and more expensive for a hospital to fund. Not surprisingly, Brinnin and Rollie McKenna paid for private nurses to come to St Vincent's to look after Dylan on the weekend of his admission.[28]

The two doctors who admitted Dylan and looked after him throughout much of his stay were Dr William McVeigh and Dr F. Gilbertson, second year medicine residents from New York university. They should have been supervised by a chief medical resident, but no such doctor appears in the various accounts of Dylan's stay in St Vincent's. Indeed, McVeigh and Gilbertson were at first supervised and directed by Milton Feltenstein, who had a private practice in the fashionable Grammercy Park area of New York. He had no admitting rights to St Vincent's nor any position of authority in the hospital. His involvement continued until the evening of November 6 when he was instructed to take no further part in supervising Dylan's treatment.[29]

Dylan was not examined by a specialist until the afternoon of November 6, when he was seen by Dr C. G. Gutierrez-Mahoney, who had been head of neurosurgery and neurology at St Vincent's since 1945. He had been trained at Harvard Medical School and in Germany and had co-authored the standard text on neurological nursing. Dylan was in good hands, even if they had arrived rather late in the day. The delay, however, was not the fault of the hospital or Gutierrez-Mahoney. When Feltenstein decided in the afternoon of November 5 that Dylan needed to be seen by a brain specialist, he advised Brinnin to hire Dr Leo Davidoff, chief of neurosurgery at Beth Israel hospital. It took twenty-four hours for Davidoff to reply, saying that he was unavailable and that he unreservedly recommended none other than Gutierrez-Mahoney, who was immediately contacted and, as Brinnin notes, arrived shortly after at Dylan's bedside.

Dylan's hospital doctors knew virtually nothing of his medical history. Little could be gleaned from Feltenstein because Dylan had told him very little about his previous illnesses. On the night of November 6/7, Daniel Jones gave Gutierrez-Mahoney information about Dylan's blackouts, drinking, possible diabetes and a haematoma (blood clot) on the right temple that, he said, Dylan had experienced two months previously. The absence of a medical history was significant: a few days after Dylan's death, his friend George Reavey talked with the doctors:

"Hospital could not establish how long Dylan had been in coma before being brought. Nor could they get any case history. Doctor said 'everyone was too emotional.' Perhaps they were hiding something. Doctor said Dylan 'had a bad chance in the beginning' because they could get no clear picture." [30]

4. Dylan's Medical Data from St Vincent's: the Intern Notes

No biographer from John Brinnin in 1955 to Andrew Lycett in 2003 has had access to Dylan's hospital data. We, too, requested the case notes but had no reply from St Vincent's. Neither were Nashold and Tremlett allowed to see this information, though they were permitted to ask questions which were answered by St Vincent's staff by reference to the case notes. Nashold and Tremlett also interviewed some of the doctors who had looked after Dylan in hospital, but not Milton Feltenstein who had died in 1974.[31]

The only person we have identified who was given access to Dylan's medical notes was Dr William B. Murphy. Having first obtained Caitlin's permission, he visited St Vincent's in the morning of December 3 1964 to read the notes and to discuss Dylan's case with Gutierrez-Mahoney. Murphy then wrote a memorandum about what he had found in Dylan's medical notes, as well as the outcome of his talk with Gutierrez-Mahoney. The following summarises Murphy's memorandum.[32]

When Dylan was admitted in coma to hospital in the early hours of November 5 he was examined by Drs McVeigh and Gilbertson. Murphy refers to their notes as the "intern notes" and we shall follow him in this. The Intern Notes record that prior to admission Dylan

- had delirium tremens before falling into a coma, and that he had had delirium tremens "several times in the past."

- had been given half a grain of morphine and eight units of ACTH approximately half an hour before lapsing into coma [i.e. sometime between 11pm and midnight on November 4.]

The Intern Notes then went on to say that, on admission, Dylan:

- had seizures and was placed on Dilantin. (This is the brand name of phenytoin, an anti-epileptic drug. It was given to calm seizures, which in Dylan's case were induced by hypoxia.)

- was "profoundly comatose", in that "he had extensor plantar response bilaterally." (This is a sign that both sides of Dylan's brain were malfunctioning - see the following Note for more detail.[33])

- had coarse crepitant rales audible in all areas of his chest. (These are noises associated with bronchitis.)

- had bronchial pneumonia as shown by a chest film.

- had a red blood cell count of 5,100,000 and a white cell count of 21,050. (The red cell count was normal but the white cell count is raised, which is associated with an infection, in this case, bronchopneumonia.)

- tested negative for diabetes. Tests also ruled out poisoning by barbiturates or other drugs.

Finally, the admitting doctors noted "in their rather scanty history", as Murphy described it, that Dylan

- suffered from a stomach ulcer three years prior to admission.

- attempted suicide with Siconal in 1951. (Sicanol is the trade name for chemicals, made by a Belgium company, that are used in the beverages industry. We believe that it is a typing error, and that the writer intended Seconal, a barbiturate often used in suicides in combination with alcohol. Murphy records that the medical notes give no further information about this incident. There is more on Dylan and suicide in the following Note.[34])

- had old scars, about four, "over the dorsal aspect of both wrists."

- had drunk "smoke", as well as "wood alcohol in the past". (Smoke is a cocktail that was once served in some New York bars; it mixed alcohol with Sterno, a stove fuel.[35])

The other main source of data available to us on Dylan's condition is the Medical Summary completed on November 9 by McVeigh for the purposes of the post-mortem. It described the circumstances of Dylan's admission, diagnosis and treatment:

"Pt. brought into Hosp in coma at 1.58am 11/5/53. Remained in coma during Hosp stay. History of heavy alcoholic intake. Received gn.1/2 of M.S. [morphine sulphate] shortly before admission. CSF [cerebrospinal fluid] clear on 2 occasions, press 260mm/H₂O, Protein 34 mg %. Urine neg for barbiturates. Serum Amylase normal.

No history of injury.

LUMBAR PUNCTURE 11/5/53 11/8/53

Impression on admission was Acute Alcoholic Encephalopathy - for which patient was treated without response. Expired after 4+ days in coma."

What does all this mean, particularly for an understanding of the cause of Dylan's coma? Dylan was injected with half a grain (= 30mg) of morphine shortly before admission. Morphine can cause coma, and can contribute to it. It can also cause respiratory depression, which can lead to hypoxia (insufficient oxygen in the blood stream), especially in someone with a pre-existing lung disease. Hypoxia can lead to coma, permanent brain damage and death. Morphine is usually metabolised by the liver, but this would be delayed in a person whose liver was damaged so that a course of injections in one day, such as Dylan received, would lead to a build-up of morphine in the body.

The CSF pressure was at the upper limit of normal. The fact that the CSF was clear rules out many other causes of coma, including sub-arachnoid haemorrhage, major trauma to the head and infection i.e. meningitis. The protein measurement in the CSF of 34 mg % was within the normal range, 18 to 58 mg per dL, but it is also within the range known to occur in acute alcoholism, 13 to 88 mg per dL, so it does not rule in, nor rule out, acute alcoholism, nor most other diagnoses for coma.[36]

Dylan's urine had tested negative for barbiturates, so that an overdose of barbiturates was not the cause of his coma.

Raised serum amalyse is diagnostic of pancreatitis, often a complication of acute and chronic alcohol abuse. But the result was normal.

The lumbar punctures were carried out to extract the cerebro-spinal fluid for diagnostic tests. The second was done on November 8, the day before Dylan died though, curiously, Nashold and Tremlett date it to the 6th.

The hospital treated Dylan for "acute alcoholic encephalopathy" i.e. it suspected that Dylan's brain had been directly damaged by a rapid intake of a large amount of alcohol immediately prior to admission. This "impression" or diagnosis was based only on information given by Milton Feltenstein and Liz Reitell. The record of delirium tremens could only have come from Feltenstein and Reitell, as probably did the information on drinking "smoke" and wood alcohol. We can confirm that Reitell and Feltenstein, who persisted throughout with a diagnosis of alcoholic coma, also passed on the false information that Dylan drank "eighteen whiskies" in the early hours of November 4 - Guttierrez-Mahoney has referred to an "acute episode" of drinking "which began in the early morning of 4 November 1953" and this information could have come only from Reitell and Feltenstein. [37]

This drinking "history" should be treated with caution, and we express our doubts about delirium tremens in the following Note.[38] But we can be sure of one consequence of the stories about Dylan's drinking eighteen whiskies, smoke and wood alcohol - they would have deflected attention away from Feltenstein's own treatment of Dylan, and lent authority in the eyes of the two junior doctors to his diagnosis of alcoholic coma.

In modern medical practice, a patient's blood alcohol level would be measured to confirm any diagnosis made simply on the basis of information provided by others, particularly for a patient from overseas whose medical and social history was not known to the hospital. A surprising feature of both the Intern Notes and the Medical Summary is that they contain no information on the amount of alcohol, if any, found in Dylan's body. CSF fluid, urine and blood samples were taken some eighty minutes after Dylan was admitted, as soon as his breathing had been restored to normal. It is conceivable that in 1953 a patient's body fluids were not routinely tested for alcohol. It is also possible that Milton Feltenstein so overawed McVeigh and Gilbertson with his diagnosis of alcoholic coma that they felt it unnecessary to measure Dylan's alcohol level.

We have gleaned further data from other sources which we draw upon in this chapter. Nashold and Tremlett discovered from the questions they put to hospital staff that on admission Dylan was severely dehydrated (p172) and that blood sugar tests on November 5 "showed a level of over 500 milligrammes against a norm of between 100 and 120 milligrammes" (p159). They were also told that Dylan's heartbeat was faint; his pulse weak but steady; his face "dry and clammy with blue lips and a splotchy, red and white, bloated face"; and the pupils of his eyes were small, not dilated (p155). Nashold and Tremlett report that Dylan was found to have anaemia (p115), but this is not borne out by the blood cell count described in the Intern Notes.[39]

We know little about Dylan's treatment in hospital. He was given artificial respiration on admission but after some fifteen minutes he was still not breathing spontaneously. After about an hour's compression and pumping, he began to breathe shallowly, and was given an oxygen mask. These difficulties in restoring Dylan to spontaneous breathing indicate how seriously ill he was at this early stage.

A blood transfusion was given on the morning of November 5, noted by Brinnin, and a tracheotomy late on November 6.[40] Dylan was then placed in an oxygen tent. He was put on a saline and dextrose drip which, according to Nashold and Tremlett, was changed to insulin on the 6th because the hospital suspected, for a while, that Dylan may have had diabetes. However, this information is not in the Medical Summary or the Intern Notes.

We do not know what drugs, if any, Dylan was given to fight the pneumonia. In 1953, most common forms of pneumonia could be treated with penicillin which was being strongly marketed by Pfizer - presumably a sick patient in a hospital would have received it. However, we have no bacteriologic data to establish whether penicillin would have been the correct drug. We have to consider, too, whether Dylan's friends and doctors might even have thought it merciful, given the extent of Dylan's brain damage, to withhold drugs to allow the pneumonia to run its course.

5. The Causes of Death

Mortuary attendants collected Dylan's body from St Vincent's and took it and the Medical Summary to the City Morgue at Bellevue hospital for the post-mortem. The examination was carried out on the afternoon of Tuesday November 10 by Dr Milton Helpern, New York's deputy chief medical examiner and an associate professor of forensic medicine. He became America's leading forensic expert, known as the "World's Greatest Medical Detective". His biography became an international bestseller, though it said nothing about Dylan's post-mortem.

Milton Helpern would have first read the Medical Summary noting that Dylan was described as a patient with a "history of heavy alcoholic intake" who had been diagnosed as having acute alcoholic encephalopathy. He would have seen there were no body alcohol measurements on the form, nor other test results which offered positive confirmation of this diagnosis. We have checked Dylan's file in the Office of the Chief Medical Examiner and established that it neither contains nor records any other medical data sent by the hospital to the post-mortem.

Helpern first examined the head. He found "considerable" pial oedema over the vertex (the crown). The brain was heavy and congested (cerebral oedema) but there were "no areas of softening or haemorrhage made out anywhere", nor blood clots (haematomata) nor tumours. The pons, cerebellum and brain stem were grossly i.e. macroscopically, normal. There was no gross pathology observed (as opposed to microscopic pathology) to explain the blackouts Dylan had been experiencing in Laugharne and London. The damage that was undoubtedly done to Dylan's brain cells by the hypoxia was not noted in the report because Helpern did not carry out a microscopic examination of cellular tissue.

As for the lungs, Helpern found that bronchopneumonia was "very evident" and "extensive" with "markedly diminished aeration" with "All of the lobes [appearing] to be involved in this bronchopneumonic process." Both lungs were affected, a condition commonly known as double pneumonia. The position of the diaphragm at the level of the fourth rib and fourth rib space was much higher than normal, suggesting that the lungs were contracted by the disease process.

Dylan's lungs were also "slightly emphysematous", hardly surprising given Dylan's history of smoking, and were "posteriorly somewhat atelectatic" i.e. collapsed, brought about by the pneumonia and also by immobility in coma. Helpern then considered the bronchi - the main air passages to the lungs - and found they were "deeply congested", indicating that Dylan had bronchitis. The bronchi were also covered with a patchy fibrin-and-pus membrane. Then Helpern noted another respiratory disorder - Dylan had "acute tracheobronchitis", described as a reddening and irritation of the upper part of the trachea (windpipe). The condition of the bronchi, taken together with the reddened and irritated trachea, add up to tracheo-bronchitis - inflammation of the trachea and bronchi. This inflammation would almost certainly have been caused by the same infection present in Dylan's lungs. It would also have been aggravated by his smoking. In short, Dylan had a badly impaired respiratory system.

Helpern's findings in Dylan's heart are typical of a 39 year old who led a sedentary life and who drank and smoked a good deal. The insides (lumen) of the coronary arteries had been narrowed by fatty deposit, in one case by fifty percent, with "considerable sclerosis with calcification". This means that Dylan was heading for angina and a possible heart attack as one or both of the arteries became completely blocked.[41]

Helpern then found several fine varices (dilated veins) at the lower end of Dylan's oesophagus. These varices are most common in patients with alcoholic liver disease. They can burst, especially as they get larger, leading to the vomiting of fresh red blood, and even causing a life-threatening haemorrhage.

Helpern did not note any signs of alcoholic hepatitis or cirrhosis in the liver but it was "firmer than normal [with] fairly evident fatty infiltration" and weighed 2,400 grams. This was much in excess of normal which is about 1,500g. There are many causes for fatty infiltration of the liver including long term alcohol ingestion and poor diet. It is not possible to tell from the post-mortem findings how well Dylan's liver was functioning, though the possibility of some liver damage cannot be ruled out.

Dylan's kidneys were fatty, as was the pancreas, which showed "considerable fatty infiltration, surrounded and concealed by fat...no evidence of fat necrosis or old pancreatitis but considerable fat between the lobules." Helpern also noted that the spleen was "moderately enlarged", a condition which is suggestive of liver damage.

Dylan's stomach showed no sign of ulceration, despite the history of vomiting blood in the past. It contained a litre of fluid, even though a tube had been inserted through Dylan's nose on November 6 to drain the stomach contents.[42] Draining is important because vomiting or regurgitation of stomach contents is a risk in a comatose patient.

There are several references to oedema in Helpern's report (pial and cerebral oedema, oedematous foreskin, puffy face). Some oedema is to be expected as a consequence of the coma but we suggest it is also possible that this generalised oedema could have been contributed to, over the long-term, by hypoproteinaemia (low blood protein levels) caused by poor diet and/or a malfunctioning liver, with low protein levels in the blood leading to fluid "leaking" out of the blood vessels into the tissues. Hypoproteinaemia is consistent with Dylan's fatty and enlarged liver, and both in turn are consistent with his alcohol consumption and poor diet. Hypoproteinaemia is consistent with Dylan's flesh being described as "watery" when he was admitted to St Stephen's hospital in 1946. It is also consistent with his puffy demeanour being apparent to Olive Suratgar as early as1951: "an unpleasant pallor and puffy, extremely puffy, for a young man."[43] We stress, however, that we have no blood test results available to confirm or refute a diagnosis of hypoproteinaemia.

Helpern concluded that the causes of death were "Pial Edema: Fatty Liver: Hypostatic Bronchopneumonia." He noted that Dylan had been "In too long for alcohol studies." The last entry on Helpern's report is a hand written note which says "Classify as Acute & Chronic Ethylism [alcoholism]. Hypostatic bronchopneumonia." Helpern would have seen from the Medical Summary that Dylan had been injected with a half grain of morphine just before going into coma, but he made no comment about its likely effects on Dylan's respiratory system.

After the post-mortem, Dylan's brain and liver was sent for chemical analysis. This was done the following day by Alexander Gettler, the City's Principal Chemist, who is acknowledged today as one of the founders of modern toxicology. He found no sign of ethyl or methyl alcohol, barbiturates, other volatile poisons or morphine. This is hardly surprising since all these would have been metabolised in the days since Dylan's admission to hospital on November 5.

i. Hypostatic Bronchopneumonia

"Hypostatic" refers to pneumonia which develops as a result of a lack of movement. This occurs in patients who are not active because, for example, they are in coma. Helpern was wrong about this, but through no fault of his own. The Medical Summary he had been given failed to mention that Dylan had pneumonia on admission to hospital. His immobility in hospital will have exacerbated the pneumonia but it did not cause it.

Helpern's finding of bronchopneumonia as a cause of death has been uncontroversial, and many biographers from FitzGibbon to Lycett have followed Helpern and written that the pneumonia was contracted while in coma. Only Ferris has raised the possibility that the pneumonia could have begun earlier, before Dylan's admission to hospital.

Dr William Murphy's examination in 1964 of Dylan's medical records (set out above as the Intern Notes) confirms that Dylan's pneumonia had indeed started before he entered hospital, as the "chest film" taken on admission showed. Murphy included this information in his 1968 paper, noting that there were "physical and radiological signs" of bronchopneumonia on Dylan's admission. Murphy had also sent this data about the pre-hospital pneumonia to FitzGibbon in December 1964, together with his memorandum containing the information he had gathered from the hospital notes. FitzGibbon had also been given the post-mortem report. Yet he did not draw upon any of this material for his 1965 biography, and thus a knowledge vacuum was created in which myths and falsehoods about Dylan's death were able to flourish.[44]

The confirmation in Dylan's medical notes of pre-hospital pneumonia is consistent with the pervasiveness of the pneumonia identified in the post-mortem report i.e. such severe and extensive pneumonia would have taken longer to develop than the four days Dylan was in hospital. Indeed, it is clear from Brinnin's account (see our Chronology) that Dylan's respiratory system was already in severe crisis by the morning of November 7.

We were unable to see for ourselves the results of Dylan's X-rays, so we asked the distinguished forensic pathologist, Bernard Knight, for a further opinion. Knight is Emeritus Professor of Forensic Pathology at the University of Wales College of Medicine. He was also a friend of Milton Helpern and his biographer. Knight examined the post-mortem report and concluded that

"death was clearly due to a severe lung infection, with extensive advanced bronchopneumonia. This is often a terminal event to many other underlying causes, but here the pre-existing acute-on-chronic bronchitis could be quite sufficient to flare up into a full-blown pneumonitis...given that Dylan had a history of a cough and chest trouble, confirmed by the autopsy finding of some emphysema, he obviously had chronic bronchitis. This often flares up at intervals into acute-on-chronic bronchitis, especially if he persisted in smoking. One would have to go on his history to a large extent, but the severity of the chest infection, with greyish consolidated areas of well-established pneumonia, suggests that it had started before admission to hospital." [45]

The post-mortem report also tells us that in addition to his chronic i.e. long-standing bronchitis, Dylan had acute tracheobronchitis, an infectious disease of the windpipe (trachea). In other words, the infection had taken hold in the whole of his respiratory tract from the windpipe down through the left and right bronchi, and into the lung tissue itself in both lungs (pneumonia). The chronic bronchitis would account for Dylan's longstanding cough and sputum, but when the sputum became infected (acute bronchitis) it developed into full-blown pneumonia.

The observations of pre-hospital pneumonia and acute-on-chronic bronchitis are consistent with many of the details noted in the Chronology: fever and chills, fatigue, paleness, loss of appetite, nausea and vomiting, shortness of breath and excessive sweating. We also note from the Chronology that Dylan was seen "gasping" for breath on October 23, and on other occasions, too. In the afternoon of November 4, he complained he was suffocating and having trouble breathing. His voice was so hoarse that he sounded like Louis Armstrong.

Bronchopneumonia is a common outcome for people who are chronically debilitated through, for example, over-indulgence in alcohol, and who also have diet and sleep deficiencies. Throughout his adult life, Dylan neglected to eat regularly and properly, as he himself mentioned to Herb Hanum on October 24: "I guess I just forgot to sleep and eat for too long." Dylan's chronic debilitation would have made it more likely that pneumonia would occur as a complication from bronchitis.

Dylan had always romantically seen himself as "the tubercular poet", but in truth it was his long-standing condition of bronchitis that did for him. His inability to take his health seriously, to stop smoking, to drink less, and to eat and sleep properly ensured that his chronic and acute bronchitis would lead, through pneumonia, to his death.

ii. Pial Oedema

The pia is one of the meninges, the three layers of thin tissue which are wrapped around the brain. We have never heard pial oedema described as a significant finding, nor of it being associated with any specific disease process. Pial oedema is certainly not a diagnosis that would be much used by pathologists today, and it is commonly observed at post-mortem examinations. Much more significant is the swelling or congestion of Dylan's brain tissue which is described as cerebral oedema.

Dylan's cerebral oedema should not be considered in isolation We have already noted that the post-mortem report suggests that Dylan had a generalised oedema, including not just pial and cerebral oedema but also a puffy face and an oedematous foreskin; some element of his "bloated" appearance would also have been due to oedema. We have suggested that some of this general oedema may have been caused by hypoproteinaemia, which should therefore be considered as something which over a longer period of time may have contributed to Dylan's cerebral oedema.

A more significant factor in causing the cerebral oedema, as well as Dylan's coma, was hypoxia or oxygen starvation. Body cells depend on oxygen to survive. When the brain is deprived of oxygen, cells begin to starve and die. The brain is the largest consumer of oxygen and it is especially sensitive to hypoxia. Brain cells begin to die in large numbers after about five minutes of reducing the oxygen supply, leading to coma, cerebral oedema and even brain death.

It is helpful in diagnosing hypoxia to have laboratory tests on blood oxygen saturation and carbon dioxide levels but this data is not available. Nevertheless, we can be sure of hypoxia because of other evidence: gross breathing difficulties, cyanosis (blue face and lips, a sign of the low oxygen level in his blood), loss of consciousness and the great difficulty the doctors had in restoring Dylan's breathing.

Dylan's hypoxia was caused by his impaired respiratory system, particularly the pneumonia. Sheldon Cohen, one of America's leading chest specialists, has taken this view. In an article in 2000 co-authored with Philip Rizzo, he summarises Dylan's life-long history of "bronchial woes". Cohen and Rizzo conclude that the primary cause of death was "respiratory tract disease with consequently impaired oxygenation." [46]

Again we turned to Professor Knight to look at the evidence, and he concluded that "there was a mild degree of cerebral oedema terminally, related to the mode of death from a severe chest infection and consequent generalised hypoxia, which can cause cerebral oedema." [47]

We concur that Dylan's bronchitis and pneumonia were a significant contribution to his hypoxia and thus to the onset of coma in his hotel room and the development of cerebral oedema.

A second pertinent factor in causing the hypoxia was morphine. Dylan was given three injections on November 4. About half an hour after receiving the last injection, he stopped breathing normally and his face turned blue. In 1977, Paul Ferris was the first to link the third injection of morphine to hypoxia and cerebral oedema. Twenty years later, James Nashold (a neurosurgeon) and George Tremlett agreed that the morphine had caused hypoxia leading to brain damage. Nashold and Tremlett write that the morphine "was enough not only to send Thomas to sleep but to put him into irreversible coma by totally suppressing his brainstem's ability to control his breathing."[48] They make the point on two more occasions:

"Thomas was now suffering from cerebral oedema, a swelling of the brain cells directly caused by the hypoxic injury..."

"It was brain swelling that ultimately caused Thomas's death."

An experienced doctor in his late forties, Milton Feltenstein could have been expected to know about the depressing effects of morphine on respiration. Cohen and Rizzo point out that these consequences of morphine have been understood since at least the nineteenth century, most notably described in Salter's medical text book published in America in 1864. A former colleague of Milton Helpern, forensic pathologist Dr Cyril Wecht, has confirmed for us that the impact of morphine on breathing was known to medical practitioners in America in 1953.

It would be useful to pause here, and examine the chain of events that led to Dylan, a sick man with a severe chest infection, being given morphine and why, when he fell into an hypoxic coma, he was not taken immediately to hospital to be revived.

Dylan woke up in his room at the Chelsea Hotel on the morning of November 4 complaining of problems with his breathing. His distress was hardly surprising - he had acute and chronic bronchitis, emphysema and pneumonia. Part of his lung would have started to collapse (atelectasis) because of the blockage of smaller airways by infected secretions. If Dylan's liver were damaged, his breathing would also have been depressed by any alcohol that would have accumulated in his system instead of being excreted by the liver.

Reitell called Milton Feltenstein for help. During the day and evening of November 4, he gave Dylan two injections of morphine and ACTH. On both occasions he failed to examine Dylan, but had he done so he would have observed that Dylan's respiratory system was in a parlous state and that he should have been admitted to hospital. Instead, he gave the injections, as well as a lecture about how Dylan should look after himself, and suggested a new diet and treatment regime.[49]

Soon after, Dylan had bouts of incoherence and began "seeing things". Reitell wrongly thought this was delirium tremens, and once more called Feltenstein, who did not realise that the "delirium tremens" were just mild hallucinations brought on by the combined effects of fever, impaired oxygenation and morphine. Even if Dylan did have delirium tremens, Feltenstein would have been wrong to contemplate another injection of morphine - its use in treating delirium tremens was unequivocally contraindicated in leading American texts at the time e.g. Noyes (1953, 4^th edition), as well as in reference books used by doctors such as the Merck Manual of Diagnosis and Therapy.[50]

But Feltenstein took out his needle yet again, and injected Dylan with a further half grain of morphine and eight units of ACTH, sometime between 11pm and midnight. The half-grain was a substantial dose, three times the amount usually given for the relief of pain. It would not normally be lethal but in this case it was given to a man with breathing difficulties. The morphine served to depress Dylan's respiration even further. It was the final insult to a beleaguered respiratory system already close to failure.[51]

Jack Heliker, who was in the hotel room with Dylan and Reitell, has timed Dylan's fall into unconsciousness at around midnight on November 4/5. Brinnin, drawing upon what Reitell had told him, has also said that when Dylan "passed into a comatose condition" it was "a few minutes after midnight." Ruthven Todd, one of Dylan's close friends who had moved from London to New York in 1947, wrote to Louis MacNeice describing what Reitell had observed:

"Shortly before midnight, she noticed a sudden change in him. He was blue in the face and she and Jack (Heliker) could find only the faintest tremor of a pulse."

Dylan was now in crisis: his chronic chest disease and pneumonia were affecting his ability to get sufficient oxygen into his blood through his lungs. His body would have tried to compensate by increasing the respiratory rate (breathing faster). But the morphine was working in the other direction, reducing the respiratory rate, further hindering the body's attempts to maintain an adequate oxygen saturation of the blood. If Dylan's liver were damaged, then the effects on respiration of the third injection of morphine would have been compounded by the liver having not fully metabolised the two previous injections. Dylan thus drifted into an hypoxic coma, the most significant causes of which were pneumonia and morphine.

Dylan's life could possibly have been saved if he had been given immediate medical treatment, but it took almost two hours to get him to a hospital. The story of this botched emergency is confused but the main details are clear. Although Dylan went into coma around midnight on November 4/5, St Vincent's hospital did not receive a request for an ambulance to be sent to the Chelsea until about 1am. What had caused the delay? Reitell and Heliker had probably tried to revive Dylan themselves. More precious time would have been spent phoning Feltenstein, who at the beginning of the emergency seems to have been unavailable. Reitell later confessed that she should have immediately called the police.[52]

But there was to be yet more delay in getting Dylan to hospital. St Vincent's was just a few blocks from the Chelsea, some five minutes drive away for an ambulance at night time. To this must be added the time it would take ambulance staff to go to a room at the Chelsea, stretcher the patient down and return to the hospital - the whole collection and delivery should not have taken more than fifteen minutes. Yet it took almost an hour, and Dylan arrived at the hospital at 1.58am, as the Medical Summary notes.

Todd's letter to MacNeice mentions that the ambulance personnel immediately gave Dylan oxygen when they arrived at his hotel room. If this is true, it may explain some, but not all, of the delay in getting Dylan to the hospital. Unfortunately, the oxygen would have made little difference because Dylan had stopped breathing properly.

Drs McVeigh and Gilbertson asked the driver what had caused the hour's delay since the hospital received the request for an ambulance. He blamed Feltenstein, Reitell and Heliker for holding him up: "They insisted on coming with me." This is hardly a convincing explanation, and Nashold and Tremlett wonder if Feltenstein had first tried, but failed, to get Dylan into Beth Israel on East 18^th Street, where he had admitting rights.

Some two hours had been lost since Dylan first became unconscious. The delay was probably enough to cause such hypoxic injury that Dylan's brain may have been already damaged when he was brought to St Vincent's in deep coma, his lips blue and his face blotched red and white with fever. For the next hour or so all effort was put into getting him to breathe spontaneously again.[53]

Drs McVeigh and Gilbertson found it difficult to establish how long Dylan had been in coma, and both Reitell and Feltenstein failed to tell them that he had not been breathing properly for some time.

iii. Fatty Liver

Dylan's liver was not only fatty, but its weight of 2,400 grams was very much in excess of normal (about 1500 grams). The mere presence of excessive fat in a liver is not on its own a serious problem and it rarely causes illness. A fatty liver is usually associated with the consumption of alcohol, though it can also be caused by diabetes, obesity and inadequate diet, particularly one deficient in protein.

We suggest that Dylan's fatty liver was a result of his drinking, poor diet and obesity. Both the fat and the weight of the liver indicate the possibility of liver damage and malfunction. We cannot be certain of the degree of any liver damage, which can only be assessed during life by liver function tests.

In the 1950s, the inclusion of "fatty liver" as a cause of death was a euphemism for "chronic alcoholism" often used by American medical examiners to spare the feelings of grieving relatives. It is likely that today a pathologist would not include Dylan's fatty liver as a cause of death.

iv. Alcohol

The Notice of Death issued by Milton Helpern's office said that the hospital's diagnosis of toxic brain damage was not confirmed:

"Comatose upon admission. Never regained consciousness. Hist: (1) Heavy alcoholism (2) 1/2 grain morphine administered by a private doctor. Treated in hospital for toxic encephalopathy but diagnosis unconfirmed."

In other words, there was no confirmation that Dylan's brain had been directly damaged by alcohol or any other toxic agent. But was alcohol an underlying cause of death, contributing to the primary causes identified by Milton Helpern? Did alcohol, for example, play any part in bringing about Dylan's pneumonia? Helpern clearly thought so. Whilst he did not list alcohol, alcoholism or alcoholic poisoning as a cause of Dylan's death, he included the euphemism "fatty liver" as part of the data that would be put on the death certificate. But when he came to classify the death for official statistics there was no need to pull his punches. He duly wrote "Acute and Chronic Ethylism [alcoholism]. Hypostatic Bronchopneumonia."

Acute Alcoholism

This means that Helpern believed that Dylan had drunk a large amount of alcohol in the hours immediately prior to going into coma. What evidence did Helpern have for this? He himself had not been able to carry out any alcohol studies of the body, as his post-mortem report declares. The Medical Summary sent to Helpern with the body did not contain data on Dylan's body alcohol level. Nor does this form report what other diagnostic steps the doctors took to establish acute alcoholism e.g. smell of alcohol on the breath, flushed face, dilation of pupils, low blood pressure and rapid pulse. In fact, the doctors noted small pupils (which fits with the record of morphine injections), a steady pulse and a splotchy red-and-white face.

The present history of events for the evening of November 4 does not support Helpern's classification of acute alcoholism. Dylan drank moderately on November 3 and went to bed early. In the early hours of November 4, he drank heavily - eight large whiskies at the White Horse between 2 and 3-30am. He then had two glasses of beer at lunchtime. But even with a damaged liver, much of this alcohol would have been metabolised from his body by late evening on November 4. According to Brinnin, Dylan did not drink anything alcoholic after the two lunchtime beers - Reitell was with him all the time, nursing him whilst he slept fitfully in bed for the rest of the day, and denying him alcohol.

This history depends wholly on Reitell's account of the evening, as passed onto Brinnin and published in his book. Could Reitell have been mistaken or was she lying as part of a cover-up? As Dylan lay ill in his hotel room in the early and late evening of November 4, did he wheedle any alcohol from her? We accept Reitell's account that he did not. Reitell was a feisty, efficient and determined woman who had been a wartime lieutenant in the Women's Army Corps. Brinnin's book describes other occasions when she had successfully controlled Dylan's drinking and broken up parties to protect him - see the Chronology for details. No biographer has yet found any other information to challenge Reitell's account, which has been corroborated by Heliker - he told Ferris that Dylan "wanted a drink very badly, but she wouldn't give him one." When Ruthven Todd went to Dylan's hotel room on November 7 to remove Dylan's possessions, he noticed that the whisky in the bottle of Old Grandad was at exactly the same level as it had been after he, Todd, had opened it on November 3 and poured the very first measure from it for the hotel maid. It had not been touched by Dylan or anyone else on November 3 or 4.[54]

In short, there is at present no evidence from body fluid tests or from the accounts given by Reitell, or any other person, to support Helpern's classification of acute alcoholism. Were such evidence to become available, indicating that Dylan had been drinking on the evening of November 4, then it would confirm that alcohol had played a part, together with pneumonia and morphine, in depressing Dylan's breathing and bringing on hypoxia.

Chronic Alcoholism

In coming to a judgement of chronic i.e. long-term, alcoholism, Milton Helpern presumably drew on the post-mortem evidence of an enlarged and fatty liver, an enlarged spleen and several fine varices (dilated veins) in the lower part of the oesophagus. The condition of the liver alone supported long-term and frequent alcohol use, but, as we have argued above, it was not in itself a direct cause of death. Helpern had also been told on the Medical Summary, that Dylan had a history of "heavy" alcoholic consumption.

Helpern's classification of chronic alcoholism indicates he believed that Dylan's lifetime of drinking played a part in bringing about his death. We broadly agree with this view. We have already described how Dylan's long-term alcohol consumption would have been a factor, along with smoking, inadequate nourishment and sleep deficiency, in the chronic debilitation that allowed his bronchitis and pneumonia to flourish, both of which contributed to the hypoxia that led to cerebral oedema. Alcohol is also implicated if it had damaged Dylan's liver, perhaps leading to hypoproteinaemia which may have contributed in the long term to the cerebral oedema.

There was no evidence, however, that Dylan's long-term alcohol consumption directly caused the cerebral oedema or damaged the brain in any other way, as the Notice of Death makes clear. Helpern did not find any signs at the post-mortem of damage or changes to the structure of the brain associated with long-term alcoholism. Whilst chronic alcoholics can suffer from "wet brain" disease or alcoholic dementia, it is impossible to make such a diagnosis without histology i.e. microscopic examination of the brain tissue. Neither Helpern nor anyone else carried out such an examination.

It is also worth noting here that alcoholic brain damage would not suddenly strike Dylan down in those two weeks in New York. It is a long-term degenerative process, and its range of symptoms, if Dylan was afflicted by it, would have been very apparent back home in Laugharne. In New York, Dylan's mind was functioning perfectly well during his professional and social engagements, including revising Under Milk Wood, taking part in two performances of the play, giving a poetry reading and participating in a symposium on film art. He displayed none of the signs of the neurocognitive deficits of alcoholism, not even those associated with mild alcoholism such as impaired memory, a declining capacity for abstract thought, attention deficit and concentration difficulties. [55]

When the post-mortem report and Notice of Death were issued, Gutierrez-Mahoney withdrew his previous diagnosis that Dylan's brain had been directly damaged by alcohol. Later, he went so far as to tell his colleague, George Pappas, that "Dylan Thomas did not have any neurological problems and I doubt whether he was truly alcoholic." [56]

In summary, alcohol was not a direct or immediate cause of Dylan's cellular brain damage or death, though over the long-term it had contributed, together with food and sleep deficits, to such a deterioration in his overall health that bronchitis and pneumonia were able to take hold and kill him. There is one last point to make about alcohol, and that is the intriguing history of the theory that Dylan died of "chronic alcoholic poisoning." This is dealt with in the following Note.[57]

v. Drugs

We know that Dylan took Benzedrine (amphetamine) and sleeping pills on his last American trip, but these have never been implicated in his death, and there is no record of his having taken them prior to his admission to hospital. Nashold and Tremlett allege that Dylan was also taking barbiturates and that when he was admitted to St Vincent's "the pupils of his eyes were small, not dilated, which indicated the presence of narcotics." [58]This allegation is contradicted by the hospital Medical Summary and the Intern Notes which show that Dylan tested negative for barbiturates. In examining Dylan's medical notes, Dr William Murphy also noted that "There was nothing to suggest he had swallowed a fistful of barbiturates, or any other drug."[59] Dylan's small pupils had probably been caused by the morphine.

vi. Diabetes

It is hardly surprising that when Colin Edwards started to talk to his interviewees about Dylan's drinking, many also brought up the matter of his eating habits. It is clear from Edwards' interviews that for much of his adult life Dylan neglected to eat regularly, and went for long periods either without food, or sustained only by sweets and fizzy drinks and, in America, canapés and other party nibbles. Dylan's taste for sweet things (including beer) and the blackouts he experienced, led Edwards to explore with interviewees whether Dylan suffered from hypoglycaemia - low blood sugar levels. A number of interviewees were with Dylan when these blackouts occurred, including Philip Burton, Richard Bell Williams, John Dark and Billy Williams. Burton also mentions a blackout that happened when Dylan was with Louis MacNeice, and FitzGibbon reports a blackout at the McAlpine's house in London in August 1953. Dr David Hughes recalls being consulted by Caitlin about the blackouts. Robert Pocock did not witness any blackouts but told Edwards that "I've often thought that the way in which he used to pass out after drinking a certain amount, suggested that he might be a diabetic, or eventually become one."[60]

Brinnin has recorded that the first lumbar puncture done in the early hours of November 5 had produced "some evidence that Dylan had sustained a diabetic shock", though by late evening, he records, all discussion of diabetes had finished and the doctors were talking of alcoholic poisoning of the brain. Nashold and Tremlett argue that Dylan was an undiagnosed diabetic, writing that the sugar levels in the blood taken from Dylan at around 3am on November 5 were "sky high."[61] They claim that the levels went even higher as a result of the dextrose that was administered from the early hours of November 5. We note that no sugar levels were recorded either on the Intern Notes or the Medical Summary - neither mentioned raised blood sugar, nor glucose being detected in the urine. And whilst a high sugar level of the kind reported by Nashold and Tremlett would certainly suggest diabetes was a possibility to be investigated, no such single random reading would prove it.

If Dylan's blood sugar levels were "sky high", Ferris has pointed out that these could have occurred as a result of the damage to the brain, something that was not appreciated in 1953. More importantly, the raised levels in the blood taken at 3am would almost certainly have been the result of the three injections of ACTH that Dylan had received from mid-day onwards on the previous day, November 4. In other words, the high sugar levels reported by Nashold and Tremlett may have been a temporary condition, and in themselves did not mean that Dylan was diabetic or in a state of diabetic shock.

We conclude that there is no evidence yet available that Dylan had diabetes. FitzGibbon has pointed out that diabetes was not found in 1940 when Dylan was medically examined for war service. Nor was sugar found in Dylan's urine when it was tested at St Stephen's hospital in 1946. The Intern Notes from St Vincent's record that in 1953 Dylan tested negative for diabetes. Dr William Murphy, writing in December 1964 two weeks after examining Dylan's hospital records, told FitzGibbon that "Laboratory findings prior to death ruled out diabetes...there was nothing to suggest he had...diabetes or other toxic condition."[62]

Nashold and Tremlett claim that "details of Thomas's diabetes were sanitised from his medical records", though they do not say what evidence or sources they have for this allegation.[63] In the same December letter to FitzGibbon, Murphy reported: "I saw no evidence that portions of the medical records had been removed or altered."

There is certainly enough material in both the Edwards interviews to caution us to be open-minded about the possibility that Dylan had problems with his blood sugar levels, though this does not necessarily mean he was an undiagnosed diabetic. Even if he were, diabetes did not bring about his death. Both Dylan's coma and death were due entirely to respiratory tract disease, morphine and oxygen starvation, as Nashold and Tremlett have themselves conceded at three separate points in their book.

vii. Medical Incompetence and Care Standards

It is inconceivable today that a doctor would inject any morphine, let alone half a grain, to relieve discomfort from gout or to treat delirium tremens. Yet this is what Milton Feltenstein did to Dylan in 1953. And he did this without knowing much about Dylan's medical history, or carrying out any diagnostic tests. He failed to detect that Dylan was a patient with chronic chest disease, suffering from bronchitis and pneumonia. Feltenstein later acknowledged to a colleague that Dylan "had died as a result of his failure to conduct diagnostic tests and his consequent failure to prescribe appropriate treatment." [64]

It is difficult to understand why Feltenstein did not admit Dylan to hospital when he first came to see him at mid-day on November 4. He certainly felt Dylan's condition was serious because he told Reitell that his principal concern was that Dylan might slip into a coma, as Gittins has reported. Gittins also notes that a doctor who had studied Dylan's hospital treatment later said that "Dr Feltenstein's failure to have Dylan admitted to hospital earlier was even more culpable than his controversial course of injections." Dr William Murphy makes the same point in his 1968 paper: "It remains a mystery how so obviously and gravely ill a person, mentally and physically, could have remained outside a hospital." Gutierrez-Mahoney concluded that "Feltenstein's treatment was extraordinarily inadequate on several counts." [65]

Reitell and Heliker failed to call an ambulance immediately, and wasted almost an hour in trying to contact Feltenstein. The fact that the ambulance then took another hour to deliver Dylan to the hospital was a further significant failure of care that contributed to his brain damage and death.

These initial errors were compounded by Feltenstein's failure, and Reitell's, to tell the admitting doctors that Dylan had been unconscious for some two hours before his admission to hospital.

For the first critical day and a half of Dylan's time in hospital, Feltenstein directed Dylan's treatment, overawing and overruling the two house doctors, McVeigh and Gilbertson. These two young doctors appear to have been left during this period without adequate supervision and Feltenstein, who had no position of authority in the hospital, assumed that supervision. This appears to be a serious breakdown in the management system of the hospital.[66]

No specialist or consultant examined Dylan until Gutierrez-Mahoney arrived in the afternoon of November 6, some thirty-seven hours after Dylan's admission.

6. Our Conclusions

Dylan Thomas died from a severe chest infection with extensive and advanced pneumonia, complicated by morphine, leading to oxygen starvation and cerebral oedema. Self-neglect and medical negligence were both implicated. Dylan's smoking, drinking and poor diet, as well as his failure to seek medical help for his various pulmonary ailments, created the circumstances in which bronchitis and pneumonia could flourish in the days before he entered hospital. But Dr Milton Feltenstein took no steps to examine his patient or carry out diagnostic tests, and he failed to detect Dylan's pulmonary and respiratory problems. In ignorance of these, he injected morphine which depressed Dylan's damaged respiratory system still further.

Reitell and Feltenstein failed to get Dylan to hospital immediately when he went into coma. Feltenstein and the hospital doctors quarrelled with each other. There was no effective management of Dylan's treatment in the first vital hours of care, nor specialist examination.

The story of the death of Dylan Thomas is one of a chain of neglect. It was an early death that could so easily have been prevented.

© Simon Barton and David N. Thomas, 2004, 2021.

Notes, Bibliography and Acknowledgements follow below:

Notes

[1] Edwards’ tape recorded interviews were carried out in the 1960s and 1970s, They are archived at the National Library of Wales. The interviews have been edited and published in two volumes. See D. N. Thomas, Dylan Remembered, 2003 and 2004, Seren.

[2] See Thomas (2004) Dylan Remembered vol 2, p193.

[3] J.M.Brinnin, (1955) p19, pp32 and 34.

[5] Caitlin and coughing: 1986, 174-76. "Bronchial heronry" is in Dylan's letter to Oscar Williams of October 8 1952.

See Thomas (2004), p201.

[6] J. Nashold and G. Tremlett pp85, 114-115 and the inhaler on p19. P. Ferris (1999, p325) also mentions an inhaler that was amongst Dylan’s possessions that were returned from New York after his death.

[7] "...my father is dangerously ill with pneumonia." (to Margaret Taylor, August 5 1949). "My father is off the danger list. He had pneumonia as well, &, though the muck on his lung has not cleared up yet, the Doctors are optimistic." (to Bill McAlpine, February 12 1950).

[8] Including those with Fred Janes, Mervyn Levy, Mably Owen, Edward Evans, Bill Mitchell, Alban Leyshon, Oscar Williams and Ralph Wishart. The interviews are in Thomas (2003, 2004) Dylan Remembered 2 vols.

[9] See the interview with Elizabeth Ruby Milton in Thomas (2004) p72.

[10] Brinnin (1955) pp31-34. Shattering fits of coughing: p19.

[11] To John Berryman, November 25 1953 in the Berryman Collection, University of Minnesota.

[12] Dylan's first injection was just before May 8, and the others were on May 26 and 29, October 24 and 25 and three on November 4. The first four of these injections are detailed by Ferris (1999) and Nashold and Tremlett (1997) and some are referred to by Brinnin. The interview with Nancy Wickwire is in Thomas (2004), p239, and confirms an injection on October 25 - it was almost certainly ACTH. Brinnin did not report this fifth injection but he was absent on the afternoon of the performance of Under Milk Wood on October 25 1953, when the injection was given .

[13] Nashold and Tremlett's assertion (p124) about twice-weekly injections i.e. about eight in total in May 1953, are based only on Dylan's comment to Stravinsky on May 22: "I prefer the gout to the cure; I'm not going to let a doctor shove a bayonet into me twice a week." This is a highly ambiguous comment, and could just as well mean that Dylan had refused the twice-weekly injections. Feltenstein's bill for his services in May 1953 was "$20...for drugs and medications." Besides ACTH, he prescribed Benzedrine (amphetamine), sleeping pills and probably the inhaler that Dylan was having to use. (Taken from Reitell's letter to Brinnin of June 4 1953, reproduced in Nashold and Tremlett, p23.)

[14] We do not know the dosages given to Dylan, apart from the eight units of ACTH given before midnight on November 4. Even if we assume high doses, the course of injections was too brief to have produced Cushingoid features such as a puffy face. Another side-effect that usually appears with a puffy face is striae, or furrows on the skin, which were not noted at the post-mortem. Finally, Dylan's puffy face was observed long before he had been given ACTH - it was apparent to Olive Suratgar in 1951 – see Thomas (2004) p270. It was also noted by Sean Treacy (Thomas (2004) p235) in October 1953, some four months after Dylan's last injection of ACTH on May 29. The puffiness is also evident in Mervyn Levy's portrait of 1950 and Gordon Stuart's of September 1953. We believe Dylan's puffiness may have been caused by hypoproteinaemia, which we discuss later.

[15] J. Lindsay (1968) pp21-22.

[16] C. Fitzgibbon (1965) p276.

[17] Thomas (2004) p133.

[18] L. Greenburg et. al. (1962) Report of an air pollution incident in new York City November 1953, in Public Health Report, January.

[19] Suffocating: Brinnin (1955) p228. Louis Armstrong: H. Breit (1954).

[20] Priscilla Sassi. Information on St Joseph's East was provided by nurses who had worked on the ward and remembered Dylan's time as a patient - thanks to Sister Rita King, the archivist of the Sisters of Charity, and Priscilla Sassi.

[21] C. Thomas (1986) pp182-83.

[22] Thomas (2004)

[23] Thomas (2004).

[24] St Vincent's had 686 beds, of which half were in general wards, 13% in private rooms and 36% in semi-private rooms. The cost per patient day was: private $20.88; semi-private $16.71; general ward $16.03. (taken by Sandra Opdycke from Financial and Statistical Information Relating to Member Hospitals of the United Hospital Fund of New York, and Hospital Statistics for Greater New York, 1955.)

[25] Even chance of life: Jill Berryman to John Berryman, November 5 1953, Berryman Collection, University of Minnesota. Ruthven Todd was also told on the day of admission that "Dylan's condition was so critical that he might die at any moment." (unpublished memoir, National Library of Scotland)

[26] Brinnin (1955), pp283-84.

[27] Email to David Thomas, March 2004.

[28] Brinnin's receipt shows he paid $24 for daytime nursing on Saturday November 7 and McKenna's receipt shows eight hours of nursing for $14.50 on November 9, probably from midnight to 8pm. Brinnin's receipts for private nursing care are in his archive at the University of Delaware; McKenna notes her contribution in her Edwards interview - see Thomas (2004) p240. There is also a photograph of her receipt on page 80 of her book on Dylan. In the early 1950s, there was a national shortage of nurses. The New York Times of May 3 1952 ran a front page article called Shortage of Nurses Found a Peril to Health of Nation. Even the prestigious and well-heeled New York Hospital had difficulties in recruiting nurses, as its annual reports show. Minutes of its 1951 Board meeting refer to a four-year survey in five Manhattan hospitals showing the average "quit rate" amongst nurses was 64% per year (research by Sandra Opdycke). The book by P.A. and B.J. Kalisch confirms the chronic shortage of nurses at this period; they write that "On a given day from 7.00am to 3-30pm [a patient] would probably see a professional graduate nurse for about six minutes" (1986). In their study (2001) of Ontario hospital admissions, Bell and Redelmeier suggest that not only are hospitals less well staffed on weekends but those who do work are likely to be less experienced; there are also fewer supervisors on duty.

[29] The accounts of Brinnin and Nashold and Tremlett describe Feltenstein's supervision. According to Nashold and Tremlett, it was not until the afternoon of November 6 that McVeigh and Gilbertson were able to snatch a few minutes discussion with Dr George Pappas, chief resident of neurosurgery at the hospital (pp160-162).

[30] Letter to John Davenport November 1953, NLW 14934E. John Berryman was given much the same picture: "There was really nothing they could do except maintain breathing & nutrition, keep him clear of mucous, and administer anticonvulsants...One of his doctors told me immediately after his death that there had never been the slightest hope...his condition was so extreme that all they could exclude was haemorrhage." (Letter to Vernon Watkins, November 1953, quoted in V. Watkins, 1983, pp151-52.

[31] Telephone conversation with George Tremlett, December 2003.

[32] Murphy's memorandum is in the FitzGibbon collection at Texas.

[33] This sign is found in patients who are deeply comatose. When a doctor runs a hard object down the sole of the foot, the toes curl downwards - a plantar response. If they curl the other way - an extensor response - it means there is damage in the 'long tracts' of nerves which run up to the brain and back to the foot muscles. Somewhere along that distance is a problem, and when associated with coma, the problem is likely to be in the brain. In Dylan's case, it was bilateral (both feet), so both sides of the brain were malfunctioning.

[34] Both Ferris (p278) and E. Lutyens (pp202-204) give accounts of the incident that took place in the autumn of 1951 when Dylan threatened to kill himself after a row with Caitlin. Ferris also notes other threats from Dylan of suicide, including "nicking himself with a razor blade on one occasion, claiming to have drunk a bottle of Jeyes Fluid on another." Daniel Jones, who helped re-unite Caitlin and Dylan in the autumn 1951 incident, was probably the one who gave the hospital the information about the Seconal.

[35] In their book on speakeasies, Kahn and Hirshfeld write that O'Leary's on "the Bowery had a recipe for smoke, made with Sterno. I don't know how anybody survived it." The owner of O'Leary's gave them instructions on how to make smoke. The formula ends with the warning, "For God's sake don't drink it." (G. Kahn and A. Hirschfeld (2003) The Speakeasies of 1932, Glenn Young Books.)

[36] mg%. This nomenclature is no longer used. It would now be written as mg per dL (milligrams per decilitre)

[37] Letter to Daniel Jones, January 10 1954, in the FitzGibbon collection, Texas.

[38] Since Ferris in 1977, Dylan's biographers have given little credence to Reitell's observation that Dylan experienced delirium tremens, a disorder caused by abruptly stopping the use of alcohol. It most commonly begins between forty-eight and seventy-two hours after the last drink - Dylan's last drink was just some ten hours before the "delirium tremens" incident. People with DTs commonly hallucinate snakes, animals and bugs; Dylan hallucinated not animals, he said, but abstract shapes. Neither did Dylan show other symptoms associated with DTs: hallucinating family and friends in the room, picking at the bedclothes, staring wildly and shouting at, and fending off, hallucinated people. Dylan showed no mental confusion or disorientation. DTs is most common in people who have a history of experiencing withdrawal symptoms when alcohol is stopped. But Dylan's own history indicates that he could stop drinking for periods of time without experiencing withdrawal symptoms. We believe that what Reitell described as DTs were mild hallucinations brought on by morphine, fever and insufficient oxygen. The hospital staff may well have considered that Dylan's seizures and dehydration supported a diagnosis of delirium tremens; but the seizures could have been induced by Dylan's hypoxia and the dehydration could have resulted from Dylan's prolonged vomiting.

[39] The blood cell count on admission was done on concentrated blood, since Dylan was dehydrated. After the blood volume had been restored by administering fluids, that figure would have dropped because the blood would have been more dilute. This fall in red blood cell concentration could have brought the figure below normal, which is the definition of anaemia.

[40] Brinnin (1955) p279.

[41] Dylan's mother, Florence, died of "(a) Coronary Occlusion (b) Coronary arterio sclerosis and Mitral Stenosis". His maternal grandmother, Hannah, died of "Heart failure accelerated by a broken leg result of accidental fall in bedroom." (death certificates)

[42] Noted by Brinnin (1954) p236

[43] Olive Suratgar: in an interview with Colin Edwards. See his archive at the National Library of Wales.

[44] FitzGibbon's biography was published in the autumn of 1965. Murphy's letter to FitzGibbon of January 4 1965 re-iterates that he has sent the hospital information to FitzGibbon on the understanding that he would not disclose it in his biography: "I hope you don't intend to make an issue of Feltenstein's efforts, good, bad, or indifferent...I owe Feltenstein nothing but I did promise Dr Mahoney I would not stir up unpleasantness based on access to his hospital records." But in the same letter, Murphy notes that "I know you have other sources of medical data e.g. the autopsy report", so it is a puzzle why FitzGibbon did not at least use material from the autopsy in his biography. FitzGibbon certainly had the autopsy report by November 1964, because he refers to it in his article in the Spectator, where his comment about the phrase "insult to the brain" being in the autopsy report first appears. Indeed, there is a reference to the post-mortem report in correspondence between FitzGibbon and a doctor friend as early as August 4 1964. (Letters to FitzGibbon from Murphy, December 16 1964 and January 4 1965 are at Texas. FitzGibbon had obtained the autopsy report via Ruthven Todd and the Dylan Thomas Estate)

[45] Letter to David Thomas, December 17 2003.

[46] Dr Sheldon Cohen is Scientific Advisor, National Institute of Allergy and Infectious Diseases, and a Scholar at the National Library of Medicine and National Institutes of Health. Philip Rizzo is Emeritus Professor of English at Wilkes University, Pennsylvania.

[47] Letter to David Thomas, December 17 2003.

[48] Nashold and Tremlett (1997) pp152, 162 and 168.

[49] Brinnin (1955) pp228-229.

[50] Noyes, p190: "Morphine should never be given." Merck Manual, p1098: "Morphine and depressing hypnotics are contraindicated." Neither Brain (1947, third edition) nor Cecil and Loeb (1955, ninth edition) mention morphine in the treatment of delirium tremens.

[51] Another factor to be considered in relation to hypoxia and Dylan's breathing difficulties are the conditions in Dylan's "small dark" hotel room. It would certainly have been hot and stuffy with the window closed, for the outside temperature dropped on November 4 and hovered around 5º C from late evening to midnight. (Temperature details are taken from the New York Times.) Dylan had had trouble with stuffy American hotel rooms, as he made clear in his letter to Caitlin of February 25 1950.

[52] "'Oh, John, why didn't I call the police?' No one calls the police, I told her, no one calls the police." Brinnin 1955, p278. A. Lycett's account (2003, p370) that Reitell called the police via the hall porter, and that Dylan reached the hospital "within minutes", is plainly wrong. Reitell continued to feel guilt-ridden by her role in Dylan's death. Many individuals and organisations would have felt fearful of being sued for damages. Reitell certainly was: a 1954 letter from her to Brinnin says "No lawyers after me yet." (The letter is in the Brinnin papers at Delaware; it is quoted by Nashold and Tremlett, p233). George Reavey thought that Feltenstein, Brinnin and Reitell were engaged in a cover-up.

[53] The sources for this narrative are: Heliker and midnight: Ferris p122; Brinnin and midnight: letter to John Davenport November 15 1953 in NLW 14934E; letter from Todd to MacNeice, November 23 1953 - thanks to Andrew Lycett for this information; hospital receives phone call about 1am: Ferris, p322 and Dr Gilbertson in Nashold and Tremlett p 154; ambulance driver blames Feltenstein for delay: Nashold and Tremlett, p154 after interviewing Gilbertson; Feltenstein and Beth Israel: Nashold and Tremlett p233.

[54] Heliker: Ferris, p322. Todd: unpublished memoir, National Library of Scotland.

[55] A. Lycett (2003) suggests (p368) that Dylan was suffering from Korsakoff Syndrome, a psychosis that affects chronic alcoholics, whose many symptoms include a decline in the capacity for thinking and learning, mental confusion, confabulation, impaired vision, eyelid drooping, muscle atrophy, speech impairment, incontinence and memory loss. Lycett produces only one sign, an apparent confabulation about a girl in a taxi. This is problematic for a number of reasons: one sign does not make a syndrome, and this particular sign could just as well be related to other clinical conditions. For example, Nashold and Tremlett (p145) use it and other signs to support their case that it was Dylan's diabetes that was affecting his brain and cerebellum.

[56] Quoted in Nashold and Tremlett, p166.

[57] On November 6, John Malcolm Brinnin was told by Milton Feltenstein that Dylan had sustained "a severe insult to the brain" and this was due to direct alcoholic poisoning. The next day, Dr Gutierrez-Mahoney saw Brinnin and confirmed a diagnosis of "direct alcoholic toxicity in brain tissue and brain cells". On November 15, the opinion reached British shores in a letter from Brinnin to John Davenport which repeated Gutierrez-Mahoney's statement about direct alcoholic toxicity in the brain which had been confirmed, wrote Brinnin falsely, by the autopsy. (Letter to John Davenport, NLW 14934E.)

When Brinnin published his book in 1955, he simply repeated what he had been told, erroneously, by Gutierrez-Mahoney in the days before the post-mortem. In 1965, FitzGibbon failed to clear the air, writing that "according to his autopsy the cause of death was: 'Insult to the brain'", a phrase never used in the post-mortem report which FitzGibbon had in his possession.

There was no clinical evidence for the statements from Feltenstein and Gutierrez-Mahoney about direct toxicity because they were made, of course, before Dylan's death and therefore before the post-mortem. When Helpern's office issued the Notice of Death that toxic brain damage could not be confirmed, Gutierrez-Mahoney was forced to back-track. When he wrote to Daniel Jones on January 10 1954, he told him that Dylan had died from "chronic alcoholic poisoning" (FitzGibbon collection, Texas). Gutierrez-Mahoney was now choosing his words very carefully. He could no longer report that Dylan had suffered direct or even acute alcoholic poisoning of the brain, because that initial diagnosis had not been confirmed by Helpern or Gettler. Now Gutierrez-Mahoney was suggesting that there had been a chronic poisoning of Dylan's body i.e. Dylan's body had been poisoned by alcohol over a long period of time. It is difficult to know what he meant by this; perhaps he was going along with Helpern in suggesting that long-term alcohol consumption had brought Dylan to a state of chronic debilitation, providing a fertile breeding ground for bronchitis and pneumonia. In any case, Gutierrez-Mahoney's phrase "chronic alcoholic poisoning" fell out of use amongst doctors many decades ago and is now considered arcane.

[58] Nashold and Tremlett: barbiturates and narcotics: p140 and p155.

[59] Letter to FitzGibbon, December 16 1964, in the FitzGibbon collection, Texas. There is also a copy in the Edwards archive, NLW.

[60] FitzGibbon (1965) p342. Hughes and Pocock: Thomas (2004).

[61] Nashold and Tremlett (1997) p159.

[62] FitzGibbon (1965) on war service: p107. Letter from B.W. Murphy to FitzGibbon December 16 1964, Texas archive with a copy in the NLW Edwards archive.

[63] Nashold and Tremlett (1997) p179.

[64] The words are those of Nashold and Tremlett (pp167-68) who had talked to Feltenstein's colleague, Dr Joseph Lehrman (a pseudonym) in 1996. Lehrman was a neurologist and psychiatrist at Beth Israel Hospital, New York. Feltenstein discussed Dylan's case with Lehrman in the spring of 1954.

[65] R. Gittins (1986) pages 154 and 163. Gutierrez-Mahoney: in discussion with William Murphy, December 3 1964.

[66] Overawing and overruling: as described by Nashold and Tremlett, pp156-168.

Bibliography

H. Breit (1954) Letter from New York in London Magazine, February

J. M. Brinnin (1955) Dylan Thomas in America, Avon paperback

C. Edwards (1968) Dylan Remembered, an unfinished biography, National Library of Wales

P. Ferris (1977 and 1999) Dylan Thomas, Dent

C. FitzGibbon (1964) The Posthumous Life of Dylan Thomas, in the Spectator, November 27

C. FitzGibbon (1965) The Life of Dylan Thomas, Little, Brown

R. Gittins (1986) The Last Days of Dylan Thomas, Macdonald

D. Jones (1977) My Friend Dylan Thomas, Dent

J. Lindsay (1968) Meetings with Poets, Muller

A. Lycett (2003) Dylan Thomas: A New Life, Weidenfeld and Nicolson

E. Lutyens (1972) A Goldfish Bowl, Cassell

R. McKenna (1982) Portrait of Dylan, Dent

J. Nashold and G. Tremlett (1997) The Death of Dylan Thomas, Mainstream Publishing

C. Thomas with G. Tremlett (1986) Caitlin: Life with Dylan Thomas, Secker and Warburg

D. N. Thomas (2003) Dylan Remembered 1914-1934 vol. 1, Seren

D. N. Thomas (2004) Dylan Remembered 1935-1953 vol. 2, Seren

R. Todd (n.d.) Dylan Thomas, an unpublished memoir, National Library of Scotland

G. Watkins (1983) Portrait of a Friend, Gomer

Medical

C.M. Bell and D.A. Redelmeier (2001) "Mortality among Patients Admitted to Hospitals on Weekends as Compared with Weekdays" in the New England Journal of Medicine, August

M. Bell et al (2004) “A Retrospective Assessment of Mortality from the London Smog Episode of 1952: The Role of Influenza and Pollution”, Environmental Health Perspectives, January

S. G. Cohen and P.L. Rizzo (2000) "Asthma Among the Famous: Dylan Thomas" in Allergy and Asthma Proceedings, May-June

L. Greenburg et al. (1962) “Report of an air pollution incident in New York City November 1953”, Public Health Rep. January

C. G. Gutierrez-Mahoney and E. Carini (1948) Neurological and Neurosurgical Nursing, Mosby

M. Helpern with B. Knight (1979) Autopsy: Memoirs of Milton Helpern, Harrap

M. Houts (1968) Where Death Delights: The Story of Dr Milton Helpern and Forensic Medicine, Gollancz

A. Hunt et al (2003) “Toxicologic and Epidemiologic Clues from the Characterisation of the 1952 London Smog Fine Particulate Matter in Archival Lung Tissues”, Environmental Health Perspectives, July

K. Johnson (2002) “You should have seen the air in 1953”, New York Times, September 29

P.A. and B.J. Kalisch (1986) The Advance of American Nursing, Little, Brown

S. Kaye and H.B. Haag (1957) "Terminal Blood Alcohol Concentrations in Ninety-Four Fatal Cases of Acute Alcoholism" in The Journal of the American Medical Association, October

W. P. D. Logan (1949) “Fog and Mortality”, The Lancet, 256:78

B. McCauley (2005) Who Shall Take Care of our Sick? Roman Catholic Sisters and the Development of Catholic Hospitals in New York City, John Hopkins

B.W. Murphy (1968) "Creation and Destruction: Notes on Dylan Thomas" in British Journal of Medical Psychology, 41, 149

A. Noyes (1953) Modern Clinical Psychiatry, Saunders, 4^th edition

S. Opdycke (1999) No One Was Turned Away: The Role of Public Hospitals in New York City Since 1900, OUP

H.H. Salter (1864) On Asthma: Its Pathology and Treatment, Blanchard and Lea

D.A. Seehusen et. al. (2003) "Cerebrospinal Fluid Analysis" in the Journal of the American Academy of Family Physicians, September 15

G.R. Stuart (1938) A History of St Vincent's Hospital in New York City 1849-1938, Stuart

M. Walsh (1965) With a great heart: the story of St Vincent's Hospital and Medical Center of New York, 1849-1964

Acknowledgements

We are particularly grateful to Professor Bernard Knight for his help, advice and encouragement, and to Professor Sandra Opdycke for letting us draw on her knowledge of New York hospitals, as well as her research notes. Likewise, we are indebted to Professor Bernadette McCauley, who drew on her research into St Vincent's hospital, and generously made available her experience and contacts. Thanks to Sister Rita King, archivist for the Sisters of Charity, New York, who provided information about St Joseph's East and contacted former nurses on our behalf; and to Priscilla Sassi, who was a student nurse at St Vincent's between 1951 and 1954. Brenda Maurer kept us all in touch.

We are also grateful for the help of Dr Robert Pitcher and Dr John F. Pickup, consultant histopathologist and principal pharmacist respectively at the Royal Cornwall Hospital. Dr Cyril Wecht, coroner and chief forensic pathologist to Allegheny County, Pittsburgh, advised on a number of points relating to the post-mortem and drug therapies. Thanks, too, to Dr Sheldon Cohen and Dr Stephen Rowlands. Our gratitude also goes to Dr Charles Hirsch, Chief Medical Examiner, New York, and to his staff attorney, Jody Lipton, both for her enthusiastic help and her determination to get to the bottom of things.

Mary Edwards did research for us in America, Paul Ferris shared material from his files, George Tremlett, David Wagoner, Rob Gittins and Andrew Lycett answered queries, Aeronwy Thomas persevered and Dave Slivka described the ward.

Thanks as well to Lisa Richter at the Harry Ransom Humanities Research Center at the University of Texas at Austin who provided material relating to Dylan's last days; as did Iris Snyder, Brinnin Special Collection, University of Delaware; Robin Smith, National Library of Scotland; Rebecca Cape, Lilly Library, University of Indiana; and Barbara Bezant at the John Berryman Collection at the University of Minnesota, as well as Mrs Kate Donahue for permission to copy material from the Berryman Collection.

Staff at the National Library of Wales were, as always, outstanding in their advice, knowledge and helpfuness. Thanks also to Mick Felton and his colleagues at Seren Books, the Wales Arts Council, the Society of Authors and David Higham Associates for their help and support.