Test of an expert witness

When in 1836 James Marsh announced his test for the presence of arsenic, the news was greeted with huge relief. Unknown numbers of deaths, it was feared, were being attributed to diseases such as dysentery or food poisoning when the real cause was murder by arsenic poisoning.

Three years later the German chemist Hugo Reinsch developed another test. Simpler than Marsh’s to run, the results were not so comprehensive, so analysts tended to run both. The Pharmaceutical Journal summed up the prevailing view: arsenic poisoning, that “most execrable” crime, was now “happily banished from the world”.

Unfortunately, the journal’s confidence proved misplaced. Not only did arsenic continue to be used as a murder weapon, but the new tests were fraught with problems, as the Smethurst case in 1859 demonstrated only too well.

Thomas Smethurst – “a man of small and insignificant mien with a reddish-brown moustache” – was a 48-year-old doctor who was accused of poisoning Isabella Bankes, a woman with a private income and a large life insurance policy. Bankes died after prolonged bouts of vomiting and diarrhoea that began soon after she went through a bigamous wedding with Smethurst. An 1883 account of the case is available in Report of Trials for Murder by Poisoning, co-written by the senior laboratory assistant at St. Thomas’s Hospital, C.G. Stewart.

The man charged with carrying out the analyses for the court was Dr Alfred Swaine Taylor, Professor of Medical Jurisprudence at Guy’s Hospital and the foremost forensic toxicologist in the country.

Alfred Swaine Taylor. Photograph by Barraud & Jerrard, 1873. Wellcome Library reference: 14969i.

Taylor’s works were key texts for me when I was researching my book The Inheritor’s Powder, which looks at criminal arsenic poisoning and toxicology in 19th century Britain. Taylor’s wrote several books, over the course of a 40-year career, including A Manual of Medical Jurisprudence and On Poisons in Relation to Medical Jurisprudence and Medicine, both of which have been digitised.

As well as providing comprehensive guides to a huge number of poisons, their nature and mode of action, Taylor also advises on how to preserve a crime scene, collect samples for the court and give evidence at a trial.

When I first encountered his work, I was intrigued to discover what were, in effect, instruction manuals on how to be a detective at a time before the role of detective existed. I returned again and again to Taylor during my research, not only to understand the contemporary thinking on toxicology, but also for a fascinating insight into the role of the doctor-turned-criminal investigator. In particular, Taylor’s attention to practical matters such as how to retrieve flakes of dried vomit from a flagstone floor conjured up for me a vivid picture of a Victorian scene of crime that predated Sherlock Holmes by nearly 50 years.

Sadly Taylor’s expertise didn’t prevent hime making a mistake that schocked many people who had come to regard his opinions as fact. The trouble began when he ran the Reinsch test on Bankes’ medicines. Surprisingly, when he introduced into the contents of one bottle a copper gauze used to collect any traces of arsenic, the gauze dissolved. So he continued to add gauze after gauze until finally the liquid could dissolve no more. He then introduced a new gauze which “at once received the arsenic”.

The bottle contained chlorate of potash, an innocent diuretic, mixed with arsenic, Taylor told the magistrate. He had never before come across such a combination and there was no medical reason for prescribing it. Smethurst was duly charged with murder.

However, at the Old Bailey trial Taylor was forced to backtrack. He had run some more tests and come to an embarrassing conclusion. “From the copper I used in my experiment, I actually deposited myself arsenic in the liquid that remained,” he told the court. Arsenic was known to be an impurity in raw copper but thought to be eliminated during the refinement process. It seemed that this was not always so and that the chlorine in the chlorate of potassium had dissolved the copper, thereby freeing the arsenic.

How are 12 men notoriously deficient in specialist knowledge to decide when professional witnesses disagree?

The jury clearly didn’t take to Smethurst, however, and they found him guilty anyway. But without Taylor’s evidence, there was now nothing to prove that Isabella had even been poisoned, let alone to indicate who was responsible. Smethurst was reprieved four days before he was due to hang and the Home Secretary blamed the confusion on “the imperfection of medical science and the fallibility of judgement … even of skilful and experienced medical practitioners”.

“We must needs pin our faith on the conclusions of chymists and hang a fellow creature because a small crystal, so minute that it can only be recognised by the microscope, is exhibited on a scrap of copper wire,”

The Times commented. “What if the chymist should be mistaken? What if science is at fault…? How are 12 men notoriously deficient in specialist knowledge to decide when professional witnesses disagree?” The question remains pertinent.

Author: Sandra Hempel is a journalist and author.