Halkides: Bad Forensics in the 1970s British Bombing Cases

Ed. Note: Chris Halkides has been kind enough to try to make us lawyers smarter by dumbing down science enough that we have a small chance of understanding how it’s being used to wrongfully convict and, in some cases, execute defendants. Chris graduated from the University of Wisconsin-Madison with a Ph.D. in biochemistry, and teaches biochemistry, organic chemistry, and forensic chemistry at the University of North Carolina, Wilmington.

In the mid 1970s Britain experienced a wave of bombings that some attributed to the Irish Republican Army (IRA) or related organizations. However, only after multiple appeals were a number of wrongfully convicted individuals exonerated. The miscarriages of justice were the result of coerced confessions, faulty forensics, failure to disclose various kinds of exculpatory evidence (including alibis and evidence of Ms. Ward’s mental instability), and other official misconduct.

The Birmingham Six were young Irish men picked up as they traveled from Birmingham to Ireland shortly after two bombs detonated in pubs. The Maguire Seven were arrested for allegedly making bombs in a home. Judith Ward was arrested in Liverpool days after the M62 coach, Latimer, and Euston bombings. Frank Skuse gave evidence against the Birmingham Six and Judith Ward, and scientists from the Royal Armament Research and Development Establishment (RARDE) gave evidence against the Maguire Seven and Judith Ward.

The Griess test detects nitrite ions via a reaction that produces a pink color. The modified Griess test uses sodium hydroxide to produce nitrite ions from compounds such as nitroglycerin (NG), PETN (an explosive not used by the IRA and not listed in the indictments), and nitrocellulose, a mostly benign substance. Any compound having a nitrate ester as a functional group should produce a positive result in the modified Griess test. Even at the time of these cases, this limitation was appreciated. Later authors leave no doubt that the Griess test is a presumptive (screening) test only, meaning that there are false positives and that a positive result must be followed up with a confirmatory test. The paradigm of using a presumptive test then a confirmatory test is commonly encountered in forensic and clinical chemistry, but the need to obtain a positive result from the confirmatory test is not always appreciated.

Thin-layer chromatography (TLC) separates chemicals on the basis of their polarities. Compounds are visualized as spots on a white plate. The faster a compound moves, the larger is its Rf value. If a standard and an unknown have the same Rf value, they may be the same substance; if the Rf values are substantially different, the standard and unknown compounds are not identical. The lab in question had apparently established a that Rf values could differ by no more than 0.05 and still allow a conclusion of probable identity. Nitrate esters were visualized with Griess test, a more discriminating technique than the Griess test alone is, but even so NG and PETN could not be told apart by the methods employed. TLC was also used to detect nitrotoluenes, which are sometimes present in gelignite explosives, along with NG.

Gas chromatography (GC) separates molecules but gives only one piece of information that helps to identify it, the retention time, how long it takes for a molecule to traverse the column. Mass spectrometry (MS) gives the mass of the whole molecule and the masses of fragments of this molecule. A signal at a mass of 46 may arise from a nitro group (one nitrogen plus two oxygen atoms), a group that can be found in NG and other chemicals. The same combination of two techniques into one (GC/MS) used in the Birmingham Six case was used also used to unravel the Patricia Stallings case (https://blog.simplejustice.us/2020/12/28/halkides-expectation-bias-and-the-patricia-stallings-case/).

With respect to the Maguire Seven case, there were some positive TLC/Griess results from hand swabs, but no explosives or traces were found in the home itself. Not disclosed were the negative results of TLC tests for nitrotoluenes. Some of the test kits produced by the Royal Armament Research and Development Establishment (RARDE) were contaminated with NG; given that RARDE also produced NG at the same site, this is not altogether surprising. It is not known whether or not these particular kits were contaminated. Certain areas in the RARDE facility were also contaminated, which might have prompted the criminal justice system to ask, “If professionals cannot avoid contaminating their surroundings, how could the Maguire Seven have done so?”

Frank Skuse applied his variation of the modified Griess test to the hands of the Birmingham Six, and his positive results were a key factor in their wrongful conviction. Dr. Skuse believed that he could adapt this test to be specific for NG (in other words, to exclude other nitroesters), but this is no better than half true. Even at the time, it was known that handling adhesive tape produced a false positive. Later workers showed that nitrocellulose-coated playing cards could also produce a false positive.

At the 1987 appeal, it was disclosed that there was one positive result from Paddy Joe Hill’s left hand, a weak signal at 46 with about the correct retention time for NG, and this proved fatal to the appeal. Most of the GC/MS results were negative or poorly documented, and the positive result is questionable on several grounds. First, Mr. Hill’s left hand gave a negative result in the Griess test. Second, one cause of spurious signals in GC/MS is carryover of a compound from one injection onto the column to the next. A laboratory’s standard operating procedures should include steps to minimize the chances of spurious signals, and a careful worker should document following such a protocol in his or her laboratory notebook. Some pages from the notebook of Janet Drayton, a scientist who performed the GC/MS testing, had been ripped out, and these pages apparently covered the running of a standard of NG and any subsequent cleaning of the GC/MS instrument. Third, a later test showed that an unidentified compound from a smoker’s hand had a similar retention time to NG and also produced a fragment with a mass of 46, and Paddy Hill smoked.

TLC evidence but no GC/MS evidence was offered against Judith Ward; the samples came from her fingers and from items in her possession. At least some of the TLC results using the Griess test were colorless or “faint” yet were considered positive for NG. The RARDE scientists knew that a dye found in shoe polish and elsewhere, Solvent Yellow 56, was a false positive in the TLC/Griess test, but they did not disclose this information. The Rf values between unknowns and standards differed by 0.05-0.10; this difference was probably larger than the tolerable range. (http://netk.net.au/UK/Ward1.asp).

Around the time of the acquittals and enquiries (circa 1990) RARDE stopped keeping standard laboratory notebooks in favor of loose-leaf notes.

For Further Reading

Mick Hamer, “Faulty forensic testing convicted Maguire Seven,” New Scientist, 18 May 1991.
Paddy Joe Hill, Forever Lost, Forever Gone, 1996.
Rebecca Nicholson, “A Great British Injustice: The Maguire Story review – a harrowing tale,” The Guardian, 25 November 2018.
Beverly Schurr, “Expert Witnesses and the duties of disclosure & impartiality: the lessons of the IRA cases in England.”

8 thoughts on “Halkides: Bad Forensics in the 1970s British Bombing Cases

  1. Howl

    “Around the time of the acquittals and enquiries (circa 1990) RARDE stopped keeping standard laboratory notebooks in favor of loose-leaf notes.”

    1. Chris Halkides

      Much of the record keeping in these cases fails to conform to what is taught to sophomore organic chemistry students: record all that you did; use a bound notebook; don’t remove pages, etc.

  2. angrychiatty

    I have no idea how knowledgeable the defense lawyers were, or how well done the cross examinations were. I wonder whether, given the climate of the time, an effective cross on the science would have significantly affected the outcome. Fascinating read, I really appreciate you taking the time to teach us this.

    1. Chris Halkides

      One commentator on the case thought that the defense could have been more effective. However, the lack of disclosure limited what the defense could have done, in my opinion. The judge in the first trial of the Birmingham Six expressed his preference for the Dr. Skuse’s testimony over the defense witness’s testimony.

  3. losingtrader

    So my conclusion, based on your explanation is that TLC evidence is not an acronym for evidence handled with “tender, loving care.”

    1. Chris Halkides

      We use thin-layer chromatography routinely in my lab. It is quite valuable, but it is subject to artifacts (the colors fade with time, for example). I have reservations about its suitability as a confirmatory test.

Comments are closed.