Caged in an Asylum: Aaron Kosminski.

IT WAS ONCE SAID of Sir Melville Macnaghten that his head was 'crammed full with official secrets'. The description might have been applied with even greater justification to his predecessor, Sir Robert Anderson. For when he retired and accepted a knighthood from a grateful King Edward VII in 1901 he had notched up twenty years in gathering and processing intelligence on Fenian activities for the Home Office and another thirteen as a.s.sistant Commissioner of the Metropolitan Police in charge of CID. Anderson's was a retirement punctuated by only infrequent references to the Jack the Ripper murders. But the few he made have given rise to immense speculation.

Readers of his book Criminals and Crime (1907) were told that the ident.i.ties of the murderer and of the author of the infamous letter to the Central News had both been established. There Anderson categorically a.s.serted that the killer had been 'safely caged in an asylum' and that the 'Jack the Ripper' letter had been penned by an 'enterprising journalist'. No names were given, no proof cited, but three years later Sir Robert fed more details to the public.

His memoirs, 'The Lighter Side of My Official Life', were then being serialized in Blackwood's Magazine. Part VI (March 1910) contained some remarkable revelations: One did not need to be a Sherlock Holmes to discover that the criminal was a s.e.xual maniac of a virulent type; that he was living in the immediate vicinity of the scenes of the murders; and that, if he was not living absolutely alone, his people knew of his guilt, and refused to give him up to justice. During my absence abroad the Police had made a house-to-house search for him, investigating the case of every man in the district whose circ.u.mstances were such that he could go and come and get rid of his blood-stains in secret. And the conclusion we came to was that he and his people were low-cla.s.s Jews, for it is a remarkable fact that people of that cla.s.s in the East End will not give up one of their number to Gentile justice.

And the result proved that our diagnosis was right on every point. For I may say at once that 'undiscovered murders' are rare in London, and the 'Jack-the-Ripper' crimes are not within that category. And if the Police here had powers such as the French Police possess, the murderer would have been brought to justice. Scotland Yard can boast that not even the subordinate officers of the department will tell tales out of school, and it would ill become me to violate the unwritten rule of the service. The subject will come up again, and I will only add here that the 'Jack-the-Ripper' letter which is preserved in the Police Museum at New Scotland Yard is the creation of an enterprising London journalist.

A footnote added: Having regard to the interest attaching to this case, I should almost be tempted to disclose the ident.i.ty of the murderer and of the pressman who wrote the letter above referred to, provided that the publishers would accept all responsibility in view of a possible libel action. But no public benefit would result from such a course, and the traditions of my old department would suffer. I will only add that when the individual whom we suspected was caged in an asylum, the only person who had ever had a good view of the murderer at once identified him, but when he learned that the suspect was a fellow-Jew he declined to swear to him.

When the memoirs were published in book form later in the year a few modifications were made to this pa.s.sage. The reference to 'low-cla.s.s Jews' was expanded to 'certain low-cla.s.s Polish Jews'. And the footnote was rewritten and incorporated into the main text: Having regard to the interest attaching to this case, I am almost tempted to disclose the ident.i.ty of the murderer and of the pressman who wrote the letter above referred to. But no public benefit would result from such a course, and the traditions of my old department would suffer. I will merely add that the only person who had ever had a good view of the murderer unhesitatingly identified the suspect the instant he was confronted with him; but he refused to give evidence against him.

In saying that he was a Polish Jew I am merely stating a definitely ascertained fact. And my words are meant to specify race, not religion. For it would outrage all religious sentiment to talk of the religion of a loathsome creature whose utterly unmentionable vices reduced him to a lower level than that of the brute.

Finally, in the introduction he wrote to H. L. Adam's The Police Encyclopaedia (1920), Anderson reaffirmed his belief that the Ripper case had been solved: 'Despite the lucubrations of many an amateur "Sherlock Holmes", there was no doubt whatever as to the ident.i.ty of the criminal, and if our London "detectives" possessed the powers, and might have recourse to the methods, of Foreign Police Forces, he would have been brought to justice.'1 For another half-century nothing more was known about Anderson's low-cla.s.s Polish Jew. But in 1959 his ident.i.ty was revealed by Dan Farson's discovery of Lady Aberconway's copy of Macnaghten's draft report of 1894. According to the draft the second suspect against whom the police had reasonable grounds for suspicion was: No 2. Kosminski, a Polish Jew, who lived in the very heart of the district where the murders were committed. He had become insane owing to many years indulgence in solitary vices. He had a great hatred of women, with strong homicidal tendencies. He was (and I believe still is) detained in a lunatic asylum about March 1889. This man in appearance strongly resembled the individual seen by the City PC near Mitre Square.

The official version, in the Scotland Yard case papers, is briefer: (2) Kosminski, a Polish Jew, & resident in Whitechapel. This man became insane owing to many years indulgence in solitary vices. He had a great hatred of women, specially of the prost.i.tute cla.s.s, & had strong homicidal tendencies; he was removed to a lunatic asylum about March 1889. There were many circ.u.mstances connected with this man which made him a strong 'suspect'.2 A last police fragment concerning Kosminski came to light at the height of the publicity surrounding the centenary of the murders. Chief Inspector Donald Sutherland Swanson died in 1924. In 1980 or 1981, upon the death of his unmarried daughter, some of his books and papers pa.s.sed to James Swanson of Peaslake in Surrey, her nephew and the chief inspector's grandson. Among his new acquisitions James found a copy of Anderson's memoirs, annotated in pencil by Chief Inspector Swanson himself. At the bottom of page 138, on which Anderson had a.s.serted that the murderer had been identified but that the witness had refused to give evidence, Swanson had written: because the suspect was also a Jew and also because his evidence would convict the suspect, and witness would be the means of murderer being hanged which he did not wish to be left on his mind.

In the margin he continued: And after this identification which suspect knew, no other murder of this kind took place in London.

On the back end-paper of the book Swanson had added a further note: Continuing from page 138, after the suspect had been identified at the Seaside Home where he had been sent by us with difficulty in order to subject him to identification, and he knew he was identified. On suspect's return to his brother's house in Whitechapel he was watched by police (City CID) by day & night. In a very short time the suspect with his hands tied behind his back, he was sent to Stepney Workhouse and then to Colney Hatch and died shortly afterwards Kosminski was the suspect DSS In 1987, when James Swanson revealed the existence of this evidence to the Daily Telegraph, the paper trumpeted its scoop on the front page under the caption 'WHITECHAPEL MURDERS: SENSATIONAL NEW EVIDENCE'. Inside it published a special report by Charles Nevin. But Paul Begg was the first to print the Swanson marginalia in full, in his book Jack the Ripper: The Uncensored Facts, in 1988.3 Macnaghten and Swanson give us the name of Anderson's suspect: Kosminski. The first author to attempt to follow up these leads was Martin Fido, whose book, The Crimes, Detection and Death of Jack the Ripper, was published in 1987. I can find little to say in favour of his theory that David Cohen, a lunatic found wandering at large in December 1888, was the murderer. But Fido is to be congratulated upon his explorations into asylum records at a time when their importance was generally unrecognized and his discovery of Aaron Kosminski in the archives of Colney Hatch Asylum was a find of major importance.

Fido published some details from Kosminski's Colney Hatch record.4 By no means do they tell the complete story. However, when I set out to learn more I quickly discovered that searching Kosminski out in workhouse and asylum records would be no straightforward task. Medical records of individual patients in public asylums are closed to public access for 100 years. Fortunately, after I had explained that my purpose was to write an accurate and objective history, the hospitals in which Kosminski was treated graciously permitted me to examine all relevant files. As a result what survives of Kosminski's story can be told in full for the first time. The records demonstrate that the memories of our police informants were faulty even on the most basic facts. For Kosminski was not committed to Colney Hatch in 1889 but in 1891. And far from dying shortly afterwards, he lived for another twenty-eight years.

The records of Mile End Old Town Workhouse show that Aaron Kosminski, an unmarried Jewish hairdresser, was admitted to the workhouse on Sat.u.r.day, 12 July 1890, from 3 Sion Square, the home of Woolf Abrahams, his brother-in-law. He was able-bodied but insane. Three days later he was discharged into the care of his 'brother'. The 'brother' referred to was probably Woolf Abrahams. However, when Aaron was re-admitted to the workhouse on Wednesday, 4 February 1891, it was from the home of Morris Lubnowski, another brother-in-law, at 16 Greenfield Street. On 6 February Aaron was examined at the workhouse by Dr Edmund King Houchin of 23 High Street, Stepney. The doctor concluded that he was of unsound mind 'and a proper person to be taken charge of and detained under care and treatment' so Henry Chambers, a JP for the County of London, accordingly made an order committing him to the county lunatic asylum at Colney Hatch. Aaron was discharged from the workhouse to Colney Hatch on 7 February. There is a slight discrepancy in the workhouse records as to his age. In July 1890 the year of his birth is noted as 1865, in February 1891 as 1864.

Today the only doc.u.ments in the records of the Mile End Old Town Board of Guardians which actually shed light on Kosminski's mental state are the medical certificate and the committal order made out on 6 February by Houchin and Chambers respectively.

The medical certificate sets out the grounds for Dr Houchin's opinion of insanity. It rested partly upon his personal examination of Kosminski: 'He declares that he is guided & his movements altogether controlled by an instinct that informs his mind; he says that he knows the movements of all mankind; he refuses food from others because he is told to do so and eats out of the gutter for the same reason.' But in addition Jacob Cohen of 51 Carter Lane, St Paul's, had informed Houchin that Kosminski 'goes about the streets and picks up bits of bread out of the gutter & eats them, he drinks water from the tap & he refuses food at the hands of others. He took up a knife & threatened the life of his sister. He says that he is ill and his cure consists in refusing food. He is melancholic, practises self-abuse. He is very dirty and will not be washed. He has not attempted any kind of work for years.'

On the reverse of the committal order some particulars about Kosminski, prepared by Maurice Whitfield, Relieving Officer for the Western District of Mile End Old Town, were recorded for the benefit of the receiving doctors at Colney Hatch. They tell us that none of Aaron's close relatives were known to have suffered from insanity and that the cause of his illness was unknown. His first attack had occurred at the age of twenty-five and he had been treated at the Mile End Old Town Workhouse in July 1890. The present attack had lasted six months. It is particularly significant that, despite Jacob Cohen's mention of a knife threat, Whitfield's statement explicitly a.s.serted that Kosminski was not suicidal or dangerous to other people.5 The records of Colney Hatch confirm that Kosminski was admitted to the asylum on 7 February 1891. On the day he came Mr F. Bryan, one of the a.s.sistant medical officers, reported that he was clean and of fair bodily health. Kosminski was held at Colney Hatch for the next three years and his progress there was doc.u.mented in brief case notes made two or three times a year. At first doctors found him difficult to deal with because of his obedience to his guiding 'instinct'. This, they thought, was probably aural hallucination (i.e. he was hearing voices). Whatever, Kosminski's 'instinct' forbade him to wash and notes of 10 February and 21 April 1891 tell us that he was objecting to weekly baths. By January 1892 his habits were cleanly. But the doctors failed to cure him of another symptom mentioned in Houchin's medical certificate a refusal to work. Between 21 April 1891, when he was described as 'incoherent, apathetic, unoccupied', and 18 September 1893, when we are told that he was 'never employed', all but one case note referred to his unwillingness to work. Kosminski's general health remained satisfactory but his mental condition seems to have deteriorated. Although tending to be reticent and morose, he could answer questions fairly when first admitted. In November 1892, however, he was only speaking German. And two months later it was noted that he was suffering from chronic mania and that his intelligence was impaired. As late as September 1893 he could still answer questions about himself. But on 13 April 1894, just six days before he was discharged to Leavesden Asylum near Watford, he was described tersely as 'demented & incoherent'.

Of particular interest to us is any disposition Kosminski may have exhibited towards violence. Our evidence is pretty conclusive on this point.

When Kosminski first came to Colney Hatch the information provided about him by Whitfield was dutifully copied into the male patients' casebook. But as a result of their experiences with the patient and, presumably, regular contact with his relatives, the staff at the asylum subsequently made alterations in red ink to some of these entries. We thus find the cause of Kosminski's illness altered in the casebook from 'unknown' to 'self-abuse' and the duration of the present attack corrected from six months to six years. Obviously, though, the doctors learned nothing to persuade them that Kosminski was a homicidal patient. For Whitfield's statement that he was not dangerous to others was allowed to stand unamended.

The case notes strongly suggest that their a.s.sessment was right. Nine notes in all cover the three years Kosminski remained at Colney Hatch. Only one, dated 9 January 1892, explicitly mentioned violence: 'Incoherent; at times excited & violent a few days ago he took up a chair, and attempted to strike the charge attendant; apathetic as a rule, and refuses to occupy himself in any way; habits cleanly; health fair'. Another, entered on 18 January 1893, recorded that at times he was 'noisy, excited & incoherent'. It is apparent, then, that Kosminski could be excitable. But more frequently he was described as quiet, apathetic or indolent. And there is no evidence of malice or cunning.

One last piece of evidence on Kosminski's behaviour at Colney Hatch exists. On 18 April 1894, a day before he was discharged to Leavesden, a statement giving brief details about him for the receiving doctors was signed by William J. Seward, the Medical Superintendent at Colney Hatch. In it Seward reiterated Whitfield's a.s.sessment that Kosminski was neither suicidal nor dangerous to others and commented simply: 'Incoherent; usually quiet; health fair.'6 Leavesden was an asylum for adult imbeciles established in 1870. And on 19 April 1894 W. Thacker, Clerk to the Board of Guardians, Mile End Old Town, signed an order for Kosminski's admission there as a 'chronic harmless lunatic, idiot or imbecile'. The order named his mother, Mrs Kosminski of 63 New Street, New Road, Whitechapel, as his nearest known relative.

Leavesden was Kosminski's home for the remaining twenty-five years of his life. No case notes before 1910 appear to have survived. But from that date eight entries afford us glimpses of his behaviour. Two (1 April and 16 July 1914) noted that he was excitable and 'troublesome' at times, one (17 February 1915) that he was occasionally 'very excitable' and another (2 February 1916) that he could be 'very obstinate'. None referred to him as a violent patient or as one that represented any risk to staff or other patients. Clean, but untidy and slovenly in his habits, he did no work and seemed unable to respond rationally to the simplest questions. This last point is mentioned in all but two of the eight case notes. 'Patient is morose in manner. No sensible reply can be got by questions. He mutters incoherently.' So ran a typical entry in January 1913. 'Patient merely mutters when asked questions,' another reported in February 1915. From four entries between 1 April 1914 and 2 February 1916 we learn that Kosminski was hearing voices and seeing things that were not there. By the latter date he had become a sad sh.e.l.l of a man, dull and vacant, and locked in a secret world of his own: 'Patient does not know his age or how long he has been here. He has hallucinations of sight & hearing & is at times very obstinate. Untidy but clean, does no work.'

Of Kosminski's general health we know a little more. Dr Henry Case, Medical Superintendent at Leavesden, informed Thacker upon Kosminski's arrival at his asylum in 1894 that his bodily condition was 'impaired'. Detailed medical records after 1910 note that it ranged from weak to good and record such mundane facts as Kosminski sustaining a cut over the left eye from an encounter with a wash-house tap in November 1915 and being twice put to bed with swollen feet in January and February 1919. It was in 1918, however, that his general health seems to have entered into terminal decline. On 26 May he was put to bed suffering from diarrhoea and 'pa.s.sing loose motions with blood & mucus'. Eight days later, his diarrhoea having ceased, he was ordered up by Dr Reese. In May 1918, too, his weight fell below seven stone. In May 1915 it had exceeded seven stone eight pounds. By February 1919, the last time he was weighed, it stood at six stone twelve pounds. It is from such arid medical data that we must of necessity reconstruct the last days of the man Sir Robert Anderson insisted was Jack the Ripper. From late February 1919 Kosminski was more or less permanently bedridden with erysipelas. On 13 March it was reported that his right hip had 'broken down'. On 22 March he was very noisy but took little nourishment. The next day he again took little nourishment and appeared 'very low'. Then, at five minutes past five on the morning of 24 March 1919, he died at the asylum in the presence of S. Bennett, the night attendant. There was a sore on his left hip and leg. Some of Kosminski's symptoms suggest that he may have been suffering from cancer but the male patients' medical journal and Kosminski's death certificate both record the cause of death as gangrene of the left leg.7 To judge by Anderson's comment that there was 'no doubt whatever as to the ident.i.ty of the criminal' one would think the case against Aaron Kosminski cut and dried. In one respect, certainly, Kosminski was unique among major Whitechapel murder suspects he was the only one against whom any direct evidence linking him with the crimes was ever adduced. That evidence, of course, was the positive identification of a witness mentioned both by Anderson and Swanson and the credibility of the case against the Polish Jew rests almost entirely upon it.

So who was the witness? Neither Anderson nor Swanson tell us his name but there are sufficient clues in the police evidence for us to determine his ident.i.ty with reasonable certainty.

First, we have Macnaghten's comment in the draft version of his 1894 report that 'this man [Kosminski] in appearance strongly resembled the individual seen by the City PC near Mitre Square'. Now, as we have seen, Macnaghten's draft and official report are factually weak. This particular statement is quite erroneous for despite Major Smith's orders that couples be kept under close observation no City policeman saw the Ripper with his victim near Mitre Square and this led to speculation in the force that they might have met there by prior appointment.8 The Mitre Square witness, in fact, was Joseph Lawende, the commercial traveller who saw a man with a woman who may have been Kate Eddowes at the entrance of Church Pa.s.sage, leading into Mitre Square, ten minutes before Kate's body was discovered in the square itself. Macnaghten's 'City PC' was undoubtedly a hazy memory of PC William Smith. Smith, however, was a Metropolitan, not a City, constable, and he reported seeing a man with Liz Stride in Berner Street, not one with Kate Eddowes near Mitre Square. In short Macnaghten confused two separate sightings made on the night of the double murder: those of PC Smith in Berner Street at about 12.35 and Joseph Lawende near Mitre Square an hour later.

It may seem difficult to believe that a senior police officer could have botched his facts as badly as this. But Macnaghten's report shows every indication of having been largely compiled from memory. In the last chapter we noted several errors in his account of Druitt and that of Kosminski is similarly flawed by its a.s.sertion that this suspect had been committed to an asylum about March 1889. The correct date was February 1891. There are also errors in Macnaghten's remarks on the Tabram, Chapman and Stride murders. The last is particularly revealing in that, like the reference to the City PC, it seems to have arisen from a transposition of the events surrounding the Berner Street and Mitre Square killings. Macnaghten's draft avers that Stride's killer was disturbed when 'three Jews drove up to an Anarchist Club in Berners Street'. Now the Berner Street killer might very well have been disturbed but if he was it was by just one Jew Louis Diemschutz, the steward of the International Working Men's Club, who drove his barrow into Dutfield's Yard, next to the club, within minutes of the time Long Liz must have been killed. Macnaghten's reference to three Jews, then, was probably inspired by the story of Joseph Lawende, Joseph Hyam Levy and Harry Harris, the three Jews who, upon leaving the Imperial Club in Duke Street later that same night, chanced upon the couple subsequently believed to have been the Ripper and Kate Eddowes. Haste and a disposition to trust too much in the memory are the causes of the Chief Constable's lapses. 'I never kept a diary, nor even possessed a notebook,' he confessed in his autobiography in 1914, 'so that, in what I write, I must trust to my memory, and to my memory alone.'9 If there is any truth at all in Macnaghten's statement the witness who identified Kosminski was either PC Smith or Joseph Lawende. Other clues, though, clearly rule Smith out. Both Anderson and Swanson were emphatic that the witness was a Jew. And Swanson's revelation that it was the City CID who watched Kosminski's house points unmistakably at Lawende. Whether Sion Square or Greenfield Street was meant is immaterial. Both were well within the jurisdiction of the Metropolitan Police. So why was the surveillance being undertaken by the City force? There seems only one plausible explanation. The witness who had identified Kosminski was Lawende so the police were seeking to charge the suspect with the murder of Kate Eddowes in Mitre Square. Since the investigation of this crime, the only one in the series which occurred in the City, was the responsibility of the City detectives they had, of necessity, to be involved in the inquiry.

If the witness was a Jew the only alternative to Lawende is Israel Schwartz. But Schwartz does not fit the bill anything like as well. If Schwartz was the witness then Macnaghten was completely wrong and the City Police would have had no business trespa.s.sing into Metropolitan Police territory in order to watch a man suspected of a crime (the Berner Street murder) committed within the jurisdiction of the Metropolitan force. Furthermore, if the man Schwartz claimed to have seen attacking Stride in Berner Street really did call out 'Lipski!' he is unlikely to have been, as Kosminski unquestionably was, a Jew.

The witness mentioned by Anderson and Swanson was almost certainly Lawende and it is upon his identification of Kosminski that the case against the Polish Jew largely hinges. This single piece of positive evidence seems to mark Kosminski out as a more likely suspect than Druitt but in other ways, too, he sounds a more plausible Whitechapel murderer than the ill-starred barrister.

His known addresses in Sion Square and Greenfield Street were within walking distance of all the murder sites. Additionally, a killer making his way there from Mitre Square could have traversed Goulston Street, where the bloodstained portion of Kate's ap.r.o.n was found. Evidence of a violent disposition, lacking altogether in the case of Druitt, is there for all to read in Kosminski's record. He was said by Jacob Cohen to have threatened his sister with a knife. When he was conveyed to Mile End Old Town Workhouse the authorities are said by Swanson to have felt obliged to restrain him by tying his hands behind his back. And in Colney Hatch he attacked an attendant with a chair. In his personal appearance, too, Kosminski arguably displayed some of the characteristics reported by witnesses of the Whitechapel killer. Mrs Long spoke of a man of 'shabby genteel' appearance. Three witnesses (Marshall, Smith and Schwartz) described one who was respectably or decently dressed. One (Marshall) said that the man he saw reminded him of a clerk rather than a manual worker. A hairdresser like Kosminski would scarcely have been an affluent man but he would not have been accustomed to dress like a labourer. 'Shabby genteel', 'decent' and 'respectable'. These terms do not seem inappropriate ones to describe the general appearance of a poor immigrant barber. Then, of course, two important witnesses (Mrs Long and George Hutchinson) explicitly reported a man of foreign or Jewish appearance.

So have we found Jack the Ripper?

Well, there is no doubt that at the moment informed opinion regards Kosminski as the leading suspect. In The Secret Ident.i.ty of Jack the Ripper, an American TV doc.u.mentary broadcast in 1988, a panel of experts including Scotland Yard's Bill Waddell unanimously chose Kosminski as the most likely murderer from a list of five admittedly ill-chosen candidates. And Begg, Fido & Skinner, in their influential Jack the Ripper A to Z (1991) contend that nothing Anderson wrote about his suspect has been 'shown to be false', that the doc.u.mentary case against Kosminski is 'very strong indeed' and that it is research into Kosminski which will most likely lead to the identification of Jack the Ripper 'if it has not done so already'. Such comments led at least one reviewer to observe that the answer to the mystery seemed 'tantalizingly close'. Unfortunately it isn't. For the facts established in the present work about Kosminski and the Ripper prove that there is no credible evidence against the Jewish hairdresser, that there are important objections to attempts to identify him with the Ripper and that Anderson, the 'rock' upon which these accusations have been founded, is repeatedly and demonstrably inaccurate and misleading.

We had better start with that crucial identification of Kosminski by Lawende.

Just how incriminating this identification was depends upon the answers to three questions. When did it take place? Under what circ.u.mstances? And how confident was Lawende in the result? We do not have the information to furnish precise answers but the little that we do know, or can deduce, sheds great doubt upon the worth of Lawende's evidence.

The important clue to the date of the identification is Swanson's statement that it took place at the 'Seaside Home'. This is a reference to the Convalescent Police Seaside Home at 51 Clarendon Villas, Hove, officially opened by the Countess of Chichester in March 1890, and its use proves that the identification was made between March 1890 and 4 February 1891, when Kosminski was last committed to the workhouse. If Swanson is to be believed we can narrow it down still further. For his statement that Kosminski was committed to the workhouse, and from thence to Colney Hatch, 'in a very short time' after his return strongly suggests that we should place the identification nearer the latter than the former date. The upshot of all this is clear; Lawende did not identify Kosminski until two years or more after his original sighting.

It is difficult to understand why it was considered necessary to take Kosminski to the Seaside Home unless it was to escape the attentions of the London press. The venue, however, does give cause for disquiet concerning the circ.u.mstances of the identification. Was Kosminski picked out from a line-up, as he would have been, for example, at Leman or Commercial Street, or were suspect and witness, as Anderson perhaps implies, simply confronted with each other? The last method would have ensured a measure of secrecy but the value of any identification made under such circ.u.mstances would have been extremely doubtful. Victorian detectives do not appear to have been consistent in their approach to the identification of suspects. The line-up, as in the case of John Pizer, may have been usual, but circ.u.mstance sometimes dictated variations in the routine. In the Lipski case of 1887 Charles Moore was taken to the London Hospital to see if he could identify Lipski, then a patient, as the man to whom he had sold a bottle of nitric acid on the morning of the murder. He picked Lipski out but doubts about the validity of his identification were not settled during the trial when Moore admitted that before he had been allowed into the ward one police officer had already told him that he would most likely find his customer there and, worse, that although he was permitted to walk from bed to bed and look at the various patients Lipski had been the only one guarded by a policeman! The argument did not cease with Lipski's conviction. Inspector Final insisted that the constable guarding Lipski was 'a young man in plain clothes, and not brash like a constable', and that the identification was 'not open to exception'. A hospital nurse, on the other hand, contended that when Moore, Final and Detective Sergeant Thick came into the ward they 'were all together and, it seemed to me, went altogether direct to the foot of Lipski's bed.'10 Anderson claimed that his witness 'unhesitatingly' identified Kosminski 'the instant he was confronted with him'. But when we look at the circ.u.mstances of Lawende's original sighting of 1888 it is impossible to understand how this could have been so. He saw the Ripper, if indeed it was the Ripper, at 1.35 on the night of the double murder, 2930 September 1888. The distance between them was not great, perhaps fifteen or sixteen feet, but the observation was a fleeting one, Lawende neither spoke to nor took especial interest in the man, and it was dark. More important, at a time when the incident was still fresh in his mind Lawende insisted repeatedly that he would not be able to recognize the man again. The contemporary records leave us in absolutely no doubt of it. 'I doubt whether I should know him again,' Lawende told the Eddowes inquest on 11 October 1888, less than two weeks after the sighting. He told the police the same thing. We know because Inspector McWilliam and Chief Inspector Swanson in their reports of 27 October and 6 November 1888 both explicitly said so. And Major Smith, writing in his memoirs of 1910, remembered it the same way. 'I could not "lead" him [Lawende] in any way. "You will easily recognize him, then," I said. "Oh no!" he replied; "I only had a short look at him".'11 It should now be clear why Lawende's identification of Kosminski cannot possibly be considered a conclusive or even persuasive piece of evidence. Anderson's book is seriously misleading on this point. He is expecting us to believe that Lawende, who saw the murderer but once, fleetingly and in a dimly lit street, and who admitted within two weeks that he would not be able to recognize him again, made a cast-iron identification of the culprit more than two years later! The notion, of course, is quite unacceptable and inevitably one wonders what pressures were brought to bear upon Lawende and why, despite them, he ultimately refused to give evidence against the suspect. Anderson and Swanson have it that he backed out when he learned that the suspect was a fellow Jew. Perhaps. But why, if the identification was sound, could not Lawende have been served with a subpoena compelling him to testify? One cannot help but speculate that Lawende was less sure of his identification than the police wanted him to be and that, when he discovered that they had no other evidence against Kosminski, his misgivings overwhelmed him. If so he was right to withdraw. His identification would not have stood up in court for a moment.

Nowhere does Anderson so much as hint that the witness who identified his suspect was used on any other occasion. Yet we have evidence that he was twice.

In February 1891, after Kosminski had been lodged in Colney Hatch, Tom Sadler, the suspect in the Coles case, was put in a line-up to see if Lawende could identify him. He couldn't. But then, in the spring of 1895 the CID came up with another Ripper suspect in the person of William Grant Grainger. The Pall Mall Gazette, reporting their inquiries, says: 'there is one person whom the police believe to have actually seen the Whitechapel murderer with a woman a few minutes before that woman's dissected body was found in the street. That person is stated to have identified Grainger as the man he then saw. But obviously identification after so cursory a glance, and after the lapse of so long an interval, could not be reliable.'12 It would be fascinating to know who this witness was. For if the police considered Lawende an honest witness and were in touch with him as late as 18901 it is by no means unlikely that they were able to use him again in 1895. Indeed, careful scrutiny of the Gazette report indicates that the 1895 witness probably was Lawende. Mrs Long and Mary c.o.x can be ruled out because the witness was a man. Hutchinson is eliminated by the reference to the victim being found in the street. The best candidates, then, are PC Smith, Israel Schwartz and Joseph Lawende. But if the victim was 'dissected', as the Gazette says, we can discount the first two. They saw Liz Stride and her body, as is well-known, was not mutilated. All this leaves Kate Eddowes the victim and Lawende, once again, the witness.

It would be wrong to make too much of unsubstantiated news reports. Time and the loss of police records have buried the truth, probably forever. Nevertheless, the fact that Lawende was confronted with suspects after he had identified Kosminski demonstrates that the first identification was anything but conclusive. And at the very least it is food for thought that he may have identified different men at different times. If it can be proved that Lawende did identify Grainger in 1895 his previous identification of Kosminski will be completely discredited.

Besides all this, the facts we have established on Kosminski raise serious difficulties for those who would identify him with the Ripper.

Take the question of his appearance. The little we know about this is soon said. His weight, recorded in May 1915, was only seven stone eight pounds and ten ounces, which suggests that he was small and slight of stature (he was said to have been in good bodily health as late as 1916). In 1888 he was twenty-three or twenty-four years old. He was a Polish Jew. And he was a hairdresser. Now it is true that these details are consistent with some of the observations reported by witnesses who may have seen the killer but, on the whole, what we know about Kosminski does not match their descriptions particularly well.

Our best witnesses concur that the murderer was of average or below average height. This fits our data on Kosminski. Build, however, is a problem. Only two important witnesses mention build. Lawende said that the man he saw was of medium build, which could fit Kosminski, but Schwartz described a stout, broad-shouldered man, which doesn't sound like him at all. Age is very difficult to estimate. Nevertheless the witnesses consistently described men who looked older than Kosminski is known to have been. And although two of the key witnesses did, indeed, accuse a foreigner neither can be said to have described Kosminski. He could fit Mrs Long's description of a 'shabby genteel' foreigner but not her estimate of age over forty. George Hutchinson's man looked Jewish. But in this case Kosminski is seemingly ruled out both by the man's age (thirty-four or five) and by his obviously prosperous appearance. It might justifiably be argued that Mrs Long be discounted on age because she did not see her suspect's face. Hutchinson's middle-aged 'toff', however, would seem a world away from Anderson's young 'low-cla.s.s' Polish barber. The other important witnesses (Smith, Schwartz and Lawende) did not indicate at the time that they had seen foreigners.

Although Kosminski's earnings as a hairdresser would not have enabled him to tog himself out like Hutchinson's man they might, as we have said, have financed the shabby genteel or respectable appearance reported by Long, Smith and Schwartz. There is, though, a real problem here in the case against Kosminski: we cannot be certain that he was working at hairdressing or anything else in 1888.

Among the symptoms of Kosminski's mental illness was a refusal to work and wash. We do not know when these particular symptoms first manifested themselves. In February 1891 Maurice Whitfield stated that Kosminski's first attack of insanity had occurred in 1890 and that the present attack had lasted six months. These details were entered in the male patients' casebook at Colney Hatch. But at some time during Kosminski's stay at the asylum the entry recording the duration of the present attack at time of admission was altered in the casebook from six months to six years and six years is also the period specified in the male admissions register. In other words Kosminski may have been exhibiting symptoms of mental illness since 1885. Jacob Cohen's statement in February 1891 certainly indicates that refusal to work had been a long-standing characteristic of his behaviour: 'He has not attempted any kind of work for years.' It is entirely on the cards, then, that in 1888 Kosminski was already dirty, dishevelled and out of work. In this condition his appearance would be irreconcilable with the descriptions given by Long, Smith or Schwartz, to say nothing of Hutchinson, and his circ.u.mstances would not square with our deduction, suggested by the dates and times of the murders, that the killer was in regular work.13 Kosminski's known acts of violence no more qualify him for the role of Jack the Ripper than his physical appearance. Certainly they fall far short of proving Macnaghten's claim that he was a homicidal lunatic with a deep hatred of women. His altercation with his (or Cohen's) sister, in which he is said to have menaced her with a knife, can hardly be considered significant by itself in the light of the great amount of domestic violence evidenced weekly in the Victorian police courts. And the workhouse and asylum records are very revealing. In 1891, when Kosminski was committed to Colney Hatch, Maurice Whitfield, the Relieving Officer at Mile End, explicitly stated that he was not considered dangerous to other people. Three years later, upon Kosminski's discharge from Colney Hatch to Leavesden, William Seward, Medical Superintendent at Colney Hatch, did the same. As late as 1910 the records of Leavesden Asylum were reiterating this belief. Clearly the authorities in these inst.i.tutions never knew that their patient had been suspected of the Whitechapel murders and, notwithstanding the incident of the chair, nothing in his behaviour while under their care gave them reason to believe that he had homicidal tendencies. Typically he languished indolent and apathetic.

Then, again, we have no evidence that Kosminski possessed even an elementary degree of anatomical knowledge. This point would not, perhaps, have troubled Anderson because he was much influenced by Dr Bond, and Bond, in his report of 10 November 1888, a.s.serted unequivocally that the murderer did not possess anatomical knowledge. Unfortunately Bond's opinion was not shared by any other medical expert who saw the wounds inflicted by the Ripper and whose judgement is on record. The rest seem to have subscribed to some degree of expertise, with Phillips and Brown attributing a great deal to the killers of Annie Chapman and Kate Eddowes respectively. Anderson's suspect thus retains credibility only if we are prepared to select the evidence that can be used to incriminate him and discount what remains.

Finally, whereas Druitt's death might explain the cessation of the crimes so neatly, Kosminski's incarceration took place more than two years after the Miller's Court murder. If Kosminski was the killer, therefore, we have to accept that after committing five if not six murders in three months he quietly went to ground and remained inactive for another two years three months. If we add Alice McKenzie to the Ripper's toll it still leaves us with a period of nineteen months between the last murder and Kosminski's committal to Colney Hatch to account for. It is by no means impossible that the Ripper remained at large and refrained from murder in order to avoid detection. But does Kosminski, foraging for bread in the gutter, drinking water from taps, dirty, unwilling if not unable to work and listening to his voices, sound like the type of man with the necessary cunning and discipline to have done so? Lawende's identification, it should be noted, cannot satisfactorily account for Kosminski's inactivity because, according to Swanson, it occurred only 'a very short time' before the committal to Colney Hatch.

If only because of Lawende we have to take Kosminski seriously. Nevertheless, the more we have discovered about this sad and pathetic suspect the less plausible the case against him has appeared. Its central strut mentioned by Anderson, Swanson and Macnaghten was the alleged identification but this, for all the reasons we have discussed, cannot possibly be regarded as satisfactory. What else the police had upon Kosminski it is impossible now to say. But to judge by the vagueness, even falsity, of the other circ.u.mstances alleged against him it was not very much. Macnaghten's claim that Kosminski had strong homicidal tendences is not substantiated at all by the medical record. And Swanson's that the murders ceased with Kosminski's identification is patently untrue. On the present evidence the case against Kosminski is so extraordinarily flimsy that we have simply no alternative but to exonerate him.

Sir Robert Anderson thought otherwise and said so in no uncertain terms. 'There was no doubt whatever as to the ident.i.ty of the criminal' '"undiscovered murders" are rare in London, and the "Jack-the-Ripper" crimes are not within that category' 'in saying that he was a Polish Jew I am merely stating a definitely ascertained fact.' This is not the language of compromise. So our dismissal of Anderson's suspect inevitably raises questions about the worth of his writings as a source of historical information.

In accounts written long after the event lapses of memory are only to be expected. Anderson makes his fair share of such slips. His memoirs state, for example, that the police undertook their house-to-house search during his absence abroad. In truth it was conducted after his return to London. Again, he is inconsistent about the date of his return. Anderson's book tells us that, having decided to spend the last week of his holiday in Paris, he arrived in the French capital on the night of the double murder, and that, when the next day's post brought an urgent summons from Matthews to return to London, he complied. A letter he wrote to one of the daily papers in April 1910, however, maintains that he was actually on his way home from Paris the night Stride and Eddowes were slain.14 The interesting thing about these memories is that neither was accurate. Contemporary doc.u.ments prove that Anderson did not return to duty until nearly a week after the double murder.

Reminiscent accounts suffer, too, from the natural tendency of their authors to interpret the past in ways advantageous to themselves. And it is in the interpretation of his memories, rather than in simple errors of fact or chronology, that Anderson most misleads later students of the Ripper case. His book foisted five important myths upon them when it contended that: (1).

his policy of warning prost.i.tutes that the police would not protect them ended street murders in the Jack the Ripper series after the double event; (2).

the house-to-house inquiry led the police to believe that the Ripper was a low-cla.s.s Polish Jew; (3).

subsequently Kosminski was identified by 'the only person who had ever had a good view of the murderer'; (4).

although the witness refused to testify against Kosminski the identification was conclusive and solved the case; (5).

the ident.i.ty of the writer of the original Jack the Ripper letter was conclusively established as that of a London journalist whom Anderson could name.

From contemporary and other evidence, every one of these contentions can be categorically refuted.

In his book Anderson explains that when he returned to London after the double murder he initiated a policy of warning prost.i.tutes that the police would not protect them. He continues: 'However the fact may be explained, it is a fact that no other street murder occurred in the "Jack-the-Ripper" series. The last and most horrible of that maniac's crimes was committed in a house in Miller's Court on the 9th of November.'15 The inference of these words is plain. Anderson is trying to suggest that his policy put an end to street murders by scaring Whitechapel wh.o.r.es off the streets.

The truth is that many of these women were compelled to solicit in the streets in order to raise money for their beds in common lodging houses. Admittedly, the terror unleashed by the double murder did temporarily diminish the number seen out at night as some sought refuge in casual wards and others fled to safer parts of the metropolis. But resilience the Whitechapel sisterhood had in abundance. And by the end of October they were back on the pavements of the East End. Mary Kelly herself picked up at least two clients in this way on the night she was killed. And as far as her murderer was concerned the fact that she turned out to possess a room of her own to which she could take him was probably an unlooked for bonus. It is impossible to find contemporary evidence that Anderson's heartless and politically impracticable proposal was ever implemented by the police let alone that it drove the prost.i.tutes from the streets. The official response to the double murder was to afford them increased protection by drafting extra men into the murder district and maintaining them there throughout the crisis. This subject is not directly linked with Kosminski. But the fact that Anderson was capable of interpreting events so perversely in order to claim credit for himself surely cautions us against accepting his other statements.

The house-to-house inquiry was completed on or about 18 October 1888. It did not persuade the police that the murders had been committed by a Jew. Indirect evidence to this effect is contained in reports Warren and Swanson prepared for the Home Office as late as 6 November. Both officers interpreted the chalked message left in Goulston Street 'the Juwes are the men that will not be blamed for nothing' as a deliberate subterfuge on the part of the murderer to throw the blame for his crimes upon the Jews. And the fact that they subscribed to this view implies, of course, that neither of them believed that the murderer himself was a Jew. Perhaps the clearest evidence that the house-to-house search did not incriminate the Jews, though, is furnished by a CID minute directed to Warren on 23 October, at least five days after the search. It mentioned the co-operation of the people of the East End, especially in permitting their homes to be searched, and acknowledged that, despite five successive murders, the CID were without 'the slightest clue of any kind'. The author of this frank admission of failure? None other than Anderson himself!16 No, for credible evidence that the Ripper may have been a Jew the police had to wait for George Hutchinson.

Anderson's statement that the witness who identified Kosminski was the 'only person who had ever had a good view of the murderer' is absurd. The witness, as I have demonstrated, was probably Lawende. But there were at least four other witnesses who were as likely to have seen the killer as he. And the evidence of three of them Smith, Schwartz and Hutchinson was arguably preferable. Although it is generally supposed that Lawende saw the Ripper talking with Kate Eddowes his evidence is open to the objection that since he did not see the woman's face he could not positively identify her as the Mitre Square victim. This type of criticism cannot fairly be levelled at the other three. Smith and Schwartz both identified Liz Stride's body as that of a woman they had seen in the company of a man in Berner Street shortly before the murder there, and Hutchinson, who described a man with Mary Kelly on the night of the Miller's Court murder, had known Mary for years. These witnesses enjoyed other advantages over Lawende. As a patrolling policeman, Smith was possibly a more careful observer. The great advantage of Schwartz's evidence was that the man he described had actually been seen attacking Stride. And Hutchinson took such an unusual and persistent interest in Mary Kelly's client as to enable him to describe the suspect in exceptional detail. Lawende, then, was merely one of five important witnesses and probably not the best at that. Nevertheless, he was the one who identified Kosminski and as such acquired special significance in the mind of Sir Robert Anderson, anxious as he was in his twilight years to believe that in this Polish Jew he had tracked down the murderer.

Incidentally, it is worth noting that back in 1888 neither Swanson, who synthesized the reports coming in from Abberline and the divisions, nor Anderson, who primarily drew upon Swanson, were in the best position to a.s.sess the relative values of the witnesses. It is to be doubted whether they saw, let alone interviewed, a single one of them. Abberline, who did interrogate them, who looked them in the eye, seems to have been particularly impressed by Hutchinson.

I have already shown that Lawende's identification of Kosminski cannot possibly have been conclusive. It is equally apparent that it was not generally regarded as such among those best qualified to judge.

Melville Macnaghten succeeded to the post of Chief Constable of the CID, the second highest office in the department, in 1890. If he was not himself a party to the Kosminski inquiry he must have been familiar with its findings. Yet, drafting his report of 1894, he explicitly exonerated Kosminski and opted for Druitt. Twenty years later, notwithstanding everything Anderson had written, Macnaghten's belief in Druitt's guilt remained unshaken and he reiterated it in his book Days of My Years.

Frederick George Abberline attained the rank of Chief Detective Inspector on 22 December 1890. His special knowledge of the East End, and of the Ripper investigation in particular, qualified him above all others to lead the Kosminski inquiry and although we have no doc.u.mentary proof that he did it must be regarded as a strong probability. In any event, had conclusive evidence of Kosminski's guilt been procured it is inconceivable that Abberline would not have known about it. The two interviews he gave to the Pall Mall Gazette in March 1903 thus contain a formidable reb.u.t.tal of Anderson's a.s.sertions. 'You must understand,' the detective told the Gazette on 23 March, 'that we have never believed all those stories about Jack the Ripper being dead, or that he was a lunatic, or anything of that kind.' A week later he was even more categoric. 'You can state most emphatically,' he said, 'that Scotland Yard is really no wiser on the subject than it was fifteen years ago.' Warming to his theme, he refuted the Druitt theory, dismissed as 'another idle story' a rumour that the poisoner Thomas Neill Cream had been the Ripper and, in the following remarks, may have referred to Kosminski: 'I know that it has been stated in several quarters that "Jack the Ripper" was a man who died in a lunatic asylum a few years ago, but there is nothing at all of a tangible nature to support such a theory.' Abberline showed the reporter recent doc.u.mentary evidence which proved that the case had never been solved and concluded in language as trenchant as any Anderson would later employ: 'No; the ident.i.ty of the diabolical individual has yet to be established, notwithstanding the people who have produced these rumours and who pretend to know the state of the official mind.'17 Although Macnaghten and Abberline clearly did not share Anderson's view of Kosminski neither are known to have ever been personally critical of their ex-chief. When Anderson's memoirs were serialized in Blackwood's, however, they provoked an immediate and scathing riposte from Sir Henry Smith. Anderson's claim that the East End Jews protected the murderer as one of their own was denounced by Smith as a 'reckless accusation'. 'Surely,' he wrote, 'Sir Robert cannot believe that while the Jews . . . were entering into this conspiracy to defeat the ends of justice, there was no one among them with sufficient knowledge of the criminal law to warn them of the risks they were running.' The latter were then considerable for in murder cases accessories after the fact were liable to penal servitude for life. Smith concluded his diatribe by recommending Anderson to read Bleak House and the Bible. Why? 'In the former book,' he explained, 'Mademoiselle Hortense, to divert suspicion from herself, writes "Lady Deadlock, Murderess" with what result Inspector Bucket tells us. In the latter, Daniel interprets the writing on the wall which brought things to a crisis at Belshazzar's Feast.'18 What Smith was saying, of course, was that if Anderson had correctly interpreted the Goulston Street writing as an attempt to throw the police off the scent of the real culprit he would have known that Jack the Ripper could not have been a Jew. There is more than a hint of personal and professional rivalry in Smith's account but his views cannot be dismissed lightly. We know from Swanson that the City CID actively partic.i.p.ated in the Kosminski inquiry and by then Smith had already succeeded Sir James Fraser as Commissioner of the City Police.

Macnaghten, Abberline and Smith. These men must have known the truth about Kosminski. Had the Ripper case been solved they would presumably have been only too glad to say so. So by disa.s.sociating themselves from Anderson on this point they demonstrated that his claim to have definitively identified the murderer was simply addle-headed nonsense. They were not alone. Thomas Arnold, interviewed upon his retirement as Superintendent of H Division in 1893, spoke of the Whitechapel murders as unsolved. Edmund Reid, who served in H Division as Head of CID between 1887 and 1896, clearly thought that Frances Coles, killed after Kosminski's committal to Colney Hatch, was the last victim of Jack the Ripper. In 1903 he dismissed Macnaghten's draft account of the three suspects, as served up by Griffiths, as 'full of inaccuracies'. And, after mooting a different theory of his own, John Littlechild, Head of the Special (Irish) Branch at the time of the murders, pointedly told George R. Sims in a personal letter of 1913 that Anderson 'only "thought he knew".'19 Chief Inspector Swanson, a strong authority on the case, did endorse Anderson. Swanson enjoyed a particularly close friendship with Sir Robert, transmitting greetings to his 'dear former master' every Christmas until Anderson's death in 1918. Loyalty and a deep sense of personal obligation may have coloured his judgement. It is also possible that the Kosminski theory originated with Swanson rather than Anderson. If so Swanson's known comments suggest that he would have argued it with greater caution.20 No matter, Anderson's claim that the case was solved by the unmasking of Kosminski cannot possibly be reconciled with the fact that inquiries continued until at least as late as 1895, with Macnaghten's a.s.sertion, in his official report of 1894, that 'many homicidal maniacs were suspected, but no shadow of proof could be thrown on any one', or with the undoubted fact that on the ident.i.ty of the Whitechapel killer senior police officers continued to contradict each other in print well into the next century.

Anderson's contention that the ident.i.ty of Jack the penman was positively established as that of a London journalist is equally untenable. I have dealt with this matter elsewhere21 and need not reiterate the arguments but doc.u.mentary evidence from the Scotland Yard files is there adduced which proves that as late as 1896, eight years after the murders, the police still did not know who had written the original letter and post card.

Anderson's memoirs do not seem to have enjoyed a high contemporary reputation. Winston Churchill, then Home Secretary, read them while they were being serialized in Blackwood's to determine whether Anderson should forfeit his police pension because of his disclosures of confidential information. He decided that to deprive Sir Robert of his pension would be to attach 'far too much importance to the articles and to their author' but noted that the articles did Anderson little credit. In particular he hit the nail right on the head when he told the Commons in April 1910 that the memoirs seemed 'to be written in a spirit of gross boastfulness . . . in the style of "How Bill Adams Won the Battle of Waterloo." The writer has been so anxious to show how important he was, how invariably he was right, and how much more he could tell if only his mouth was not what he was pleased to call closed.'22 There is no doubt that the section on the Ripper crimes was very misleading indeed and Anderson may have made at least one other blunder when he implied that James Monro had sanctioned the 'Parnellism and Crime' articles he had written for the Times in 1887. Monro denied it: 'No such authority was asked by Mr Anderson, and none was given to him by me . . . A long time afterwards Mr Anderson informed me that he had written one or more of the articles, and I felt much annoyed.'23 None of this ent.i.tles us to dismiss Sir Robert as an arrant liar. A competent police chief, he was valued and respected by many of his colleagues, and he did not invent Kosminski. Why, then, did he write so misleadingly about him? We can but speculate. That irritating sense of self-importance detected by Churchill suggests part, but only part, of the answer. I incline to the belief that Anderson's errors of interpretation stemmed not from a wilful intent to deceive but from wishful thinking, that what he was doing was interpreting his memories of Kosminski in exactly the same way that Warren and Abberline had interpreted their clues on Jacob Isenschmid in 1888. They had found themselves propelling the mad pork butcher in the direction of the gallows because of the public clamour for a conviction. The pressures upon Anderson, though different, were productive of similar results.

By 1910 the Ripper murders had slipped into history. In writing their reminiscences, however, public servants naturally have no wish to depict themselves as fools or failures. And a man of Anderson's self-conceit would have found it especially difficult to concede a blow to his personal and professional pride as ignominious as the CID's inability to detect the Whitechapel killer.

It is not unreasonable to suppose that Anderson was also deeply galled by public criticism over the Ripper affair. Certainly he fumed and not without reason at the jibes of amateur sleuths like Dr Winslow, who freely dispensed blame and advice and pretended that they knew more than the police. About one such, Edward K. Larkins, Anderson wrote in 1893: 'Mr Larkins is a troublesome busybody whose vagaries on the subject of the Whitechapel murders have cost this department, the Public Prosecutor and the Foreign Office a great deal of trouble . . . it is a mere waste of time attempting to deal with him on this subject.' And the same brand of derisive exasperation surfaced in his 1910 reminiscences. 'When the stolid English go in for a scare,' he observed tartly, 'they take leave of all moderation and common sense. If nonsense were solid, the nonsense that was talked and written about those murders would sink a Dreadnought.'24 Troubled by deafness and an increasing sense of isolation, his days occupied in quiet contemplation of the scriptures, his nights plagued by the attacks of 'blue devils', Sir Robert lived out his retirement at his home at 39 Linden Gardens, Hyde Park. He must sometimes have reflected there upon those hectic days at the Yard. And when he did it would doubtless have given him comfort to think, that whatever the world might say, he had laid the Ripper by the heels. Over the years, with the selective and faulty memory characteristic of advancing age, he came to believe it.

In supporting him, Swanson exhibited that same capacity for self-deception. 'After this [Kosminski's] identification which suspect knew,' he wrote, 'no other murder of this kind took place in London.' He had conveniently forgotten, of course, about the Ripper-type slaying of poor Frances Coles in February 1891, only six days after Kosminski had been 'caged' in his asylum. And if it be objected that Swanson was subscribing to the conventional view that Mary Kelly had been the Ripper's last victim, surely he should have made it clear that the crimes had ended, not with Kosminski's identification, but two years before it.

None of this mattered. Anderson and Swanson had come to inhabit a world of wish-dreams. And together they transformed a harmless imbecile, sheltering within the walls of Leavesden, into the most infamous murderer of modern times.

21.

The Mad Russian: Michael Ostrog.

THE THIRD MAN, named by Macnaghten as Michael Ostrog, was a thief and confidence trickster accustomed to living under numerous aliases.

In his draft report Macnaghten wrote: No: 3. Michael Ostrog, a mad Russian doctor & a convict & unquestionably a homicidal maniac. This man was said to have been habitually cruel to women, & for a long time was known to have carried about with him surgical knives & other instruments; his antecedents were of the very worst & his whereabouts at the time of the Whitechapel murders could never be satisfactorily accounted for. He is still alive.

The official version, preserved in the Scotland Yard files, is just two sentences long: (3) Michael Ostrog, a Russian doctor, and a convict, who was subsequently detained in a lunatic asylum as a homicidal maniac. This man's antecedents were of the worst possible type, and his whereabouts at the time of the murders could never be ascertained.1 Until recently nothing else was known about Ostrog, which is a mystery in itself because at the time of the Ripper murders the Metropolitan Police publicized their interest in him in The Police Gazette, an obvious source for any student of the Whitechapel crimes. Six years ago, when I came to investigate the double murder,