A great site

Archive for the tag “probability”

Has mtDNA identified Jack the Ripper?

“Ripperology” is quite a confused subject and at least a dozen suspects have been conclusively “identified as the Whitechapel fiend. Nevertheless, this article and the book detailed within, if taken at face value, uses the scientific techniques that identified Richard III, Jesse James, Nicholas II and others to claim to solve the East London riddle once and for all.

The critical case is that of the fourth definite, or “canonical”, victim: Catherine Eddowes, who was slain in the early hours of Sunday 30 September 1988 and found in Mitre Square, about an hour after the discovery of Elizabeth Stride. Eddowes’ apron, found nearby, is the first exhibit here and the second is a silk shawl (below left), apparently found by her body and heavily bloodstained. Russell Edwards, who wrote the article in question, located the shawl at an auction in Bury St. Edmunds in 2001 and contacted Jari Louhelainen, a senior biology lecturer at Liverpool John Moores University and crime buff.

Jack the Ripper suspect Aaron Kosminski

First, they traced Eddowes’ great-great-great granddaughter and the multiple-great-niece of leading suspect Aaron Kosminsky (right), who was, in September 1888, a 23 year-old Polish Jewish hairdresser in East London that month, later to die in a lunatic asylum in Colney Hatch, where he was admitted in 1891. Importantly, both of these living relatives were related strictly through the female line and the mtDNA of both matched the blood on the apron and semen on parts of the shawl, thus seemingly demonstrating beyond peradventure that their respective Victorian relatives had both left bodily fluids on them.

Sadly, Edwards and Louhelainen’s research is yet, five years after publication, to be peer reviewed. If it were, we would need look no further for “Jack” as Kosminski briefly entered a workhouse in 1890 and was committed to Colney Hatch (left) the following February.

Here, however, is the counter-argument, quoting Paul Begg and Dr. Turi King, disputing the connection between Eddowes and the shawl, the genealogy and the significance of the DNA match. In summary, Kosminski cannot be eliminated by Edwards and Louhelainen’s research, nor can he yet be identified as “Jack” with forensic certainty. This BBC documentary points towards him in other ways.

Richard III, snooker and probability

One thing of which we can be certain is that Richard III never played snooker. It was not invesnookernted until 1875 in Jabalpur by a Colonel Chamberlain (1). Nevertheless, it is an excellent vehicle for demonstrating the laws of probability with particular reference to the descent of the Plantagenet Y-chromosome from Edward III.

Imagine that you have walked into a snooker club where a member lends you four white balls and fifteen reds, the white balls obviously from more than one set, but in a drawstring bag. The cue balls represent the paternal links from Edward III to Richard III and the reds represent the descent from Edward III to Henry, 5th Duke of Beaufort (2). We already know that the 5th Duke’s living putative descendants have a different Y-haplogroup to Richard III, indicating that there is at least one “false paternity event” in one or both lines, but “Somerset 3” has a different Y-chromosome to his putative cousins, showing that another such has occurred at some time since 1760.

The bag is now held towards you and you are invited to insert your hand and withdraw a ball but you cannot discern its colour until you are holding it outside the bag – we are assuming randomness a priori. The probability of one random ball being red is 15/19 or approximately 79%. If you withdrew two balls, the probability of both being red is 15×14/19×18 or about 61%. The probability of three balls all being red is 15x14x13/19x18x17 or about 47%.

The probability of any paternal link in these chains being false is the same as stated above. We only know that there is one such event and it is 79% likely to have been in the descent to the 5th Duke but 21% to Richard. We cannot yet assume there to be more than one broken link in either chain and it would take three “milkmen” for the red ball (Beaufort) probability to fall below 50% and for a York false paternity event to be probable.



Post Navigation

%d bloggers like this: