Chapter 51

There are strings in the human heart that had better not be vibrated.

Charles Dickens, Barnaby Rudge

Louisa, the stout-hearted, the defiantly inscrutable, refused to appease society’s concerns by confessing her guilt or publicly professing her innocence. It falls to history’s detectives to reassess the evidence in order to determine the truth about this celebrated historical case.

The question of causation is pivotal: how did Andrews die? While the chemical evidence conclusively shows that Collins died of arsenic poisoning, it is inconclusive in terms of Andrews’ death. Nonetheless, the similarities between their illnesses are glaring: the vomiting and diarrhoea, suggesting that their bodies were attempting to expel substances that acted as irritants; the failure of the prescribed medicines to ameliorate their symptoms or slow down their deterioration; the evidence of arsenic, an irritant, in their remains.

There are other less obvious but no less significant similarities, including the fact that both men vomited green matter; industry once used arsenic to produce green dyes. Both men included ‘shoulder pain’ in their short list of symptoms—not foot pain or knee pain or wrist pain but shoulder pain—and arsenic poisoning can cause joint pain. Also, Andrews complained of a severe burning in his stomach and bowel, which resembles the ‘bowels of fire’ mentioned by other victims of arsenic poisoning. And the region of Andrews’ torso covering his stomach and gut—that is, the region that would have suffered most from short-term arsenic poisoning—was surprisingly well preserved, which is worth noting because the exhumed remains of Styrian arsenic-eaters showed considerably less decomposition than other corpses from the same period. Thus, while it is impossible to state with one hundred per cent certainty that Andrews died of arsenic poisoning, the body of evidence supports the conclusion ‘beyond reasonable doubt’. Moreover, this was the verdict of the treating doctors and other expert medical practitioners once they had assessed all the evidence for themselves.

So how did arsenic enter Andrews’ body? Although Andrews might have been exposed to tiny amounts of arsenic while working as a wool-washer, evidence now shows that this type of exposure does not cause fatal arsenic poisoning. The authors of an article published in The Lancet in 1899 could find no cases of arsenic poisoning resulting from contact with items made from the wool of arsenic-dipped sheep. The arsenic bonded so tightly to the wool that even carding and combing failed to release it.⁹

Importantly, though, Andrews was not working in the month before his death because his employer’s business had closed. Therefore, workplace absorption could not have been responsible for the arsenic traces found in his stomach remains. His continued vomiting and purging throughout his illness suggests that he received a moderate dose of arsenic around ten days before his death and another dose or doses in the days following. As his body had expelled most of the arsenic prior to his death, the last dose was probably administered a few days beforehand.

Since there is no evidence that Andrews accidentally ingested or absorbed the arsenic and no evidence of suicidal intent, it is clear that he was murdered. This brings us to the question of culpability, not only in terms of Andrews’ death but in relation to Collins’ as well. Putting Louisa aside for the moment, there are others who need to be considered as possible perpetrators.

A suggestion at the time was that Collins was responsible for both Andrews’ death and his own; however, this is highly unlikely. Andrews’ unemployment meant that Collins no longer had ready access to the house, let alone to the family’s food and beverages, during the critical ingestion period. And Collins’ ‘suicide’ seems improbable; humans do not put themselves through an extended period of torture when they are intent on killing themselves. Indeed, James Whorton, Professor Emeritus of the History of Medicine at the University of Washington, Seattle, came across no instances of suicide by slow arsenic poisoning during his extensive research for his book, The Arsenic Century.¹⁰

Another suggestion was that Collins was responsible for Andrews’ death and Louisa for Collins’ death. To posit that there were two murderers, both working independently of each other using the same modus operandi without recognising what the other was doing, beggars belief. Moreover, if Collins and Louisa had acted together to kill Andrews, Collins would have had to be remarkably obtuse not to recognise that he was suffering from the same symptoms. Under the circumstances, the only logical conclusion is that Collins was not responsible for either death.

Could someone else in the house have been responsible? Louisa’s son Arthur, for instance? While there was no love lost between him and Collins, he had no reason to kill his own father. Nor did his younger siblings, who were barely into double digits at that time, or his older siblings, who were not even in Sydney when Collins died.

If we apply Ockham’s Razor to the question of culpability—that is, ‘Of two (or more) competing theories, the simplest explanation is to be preferred’—the ‘simplest explanation’ is that Louisa caused the deaths of her two husbands as she alone had means and opportunity for both deaths and a recognisable motive for one. If we accept this ‘simplest explanation’ for the sake of argument and attempt to find supportive evidence by focusing on Louisa’s remarks and behaviour, the case against her becomes much stronger.

Louisa showed an extraordinary callousness in her attitude and behaviour towards Andrews during his illness and afterwards. Most notably, this is seen in her failure to follow the basic mourning tradition of straightening his body before racing off to obtain his insurance money, and in permitting Collins to move into her marriage bed less than a week after Andrews’ death. Moreover, she showed guile when she attempted to justify her abandonment of Andrews’ body by saying that the insurance doctor might have needed to see it.

Her callousness can also be seen in her behaviour towards Collins. She left his body unattended for hours with the front door wide open and showed no grief or concern when she discovered that the police had removed his body during that period. She behaved in a conniving manner when she attempted to obtain his death certificate by telling the doctor that the police were meeting her at his surgery. Additionally, she showed overt displays of affection towards Collins on the morning of his death after she had sent a message requesting Senior Constable Sherwood to visit her—an obvious move to divert suspicion.

And what about ‘Mick’s milk’: the jug of condensed milk that sat on the kitchen bench? She stopped her children from pouring the milk into their tea, and even refused to let her son drink his tea after he’d already added the milk to his mug. Significantly, the deadly beverage on Mick’s bedside table comprised condensed milk laced with arsenic.

The most eloquent evidence of her guilt, though, is her silence. She failed to protest her innocence on the occasions when she could and should have done so. After Senior Constable Sherwood told her she was to come with him to the Darlinghurst police station, she did not ask why or declare her innocence, but assumed she wouldn’t be returning. When he charged her with Collins’ murder, she didn’t say that she hadn’t done it as most people would do, innocent or guilty alike. Most significantly, she did not make a ‘dock statement’ in any of her four trials. Surely, anyone who was unjustly accused would vociferously protest his or her innocence to the court, particularly when facing a death sentence. Her failure to do so would have been noted by the judges and, no doubt, contributed to their certainty, as mentioned in the appeal, that she was guilty of both deaths. She didn’t even cry, ‘I am innocent!’ after being sentenced to death or when she begged the governor for mercy for the sake of her children.

When the body of evidence is examined in its entirety, it is clear that Louisa was responsible for dosing her two husbands with arsenic and watching them die the most agonising of deaths. If she hadn’t been caught, it is feasible that she would have killed again—perhaps not even intentionally. Keeping arsenic-laced milk in an unmonitored jug on the kitchen bench was a recipe for disaster if ever there was one.

But why did she murder her husbands?

Her motive for killing Andrews is obvious: lust and greed. Needless to say, most greedy, lustful people do not commit murder. Something has to happen to trigger a murderous response, something that creates a problem for which killing seems the optimal solution. The trigger in the Andrews murder was probably his sudden unemployment. With limited opportunities for romance, Louisa found herself at a crossroad. She must have known that her Romeo would soon tire of her if the situation didn’t change. One option was to abandon her family and run away with him; however, Collins was too lazy to be a good provider and her situation would have been dire if he’d later abandoned her. There was another option: Andrews’ insurance money, the ropes of gold that might tie Collins permanently to her side.

Killing Andrews rewarded her with love and money; however, there was no such payout for killing Collins. The lack of an obvious motive for killing her second husband handicapped the prosecution in the first two trials, although the issue was largely ignored in the third Collins trial because of the ‘double murder’ scenario presented. Nonetheless, in order to understand her actions, it is essential to determine her motive for killing Collins.

Studies reveal that women who kill their husbands are usually reacting in self-defence to ongoing violence. Louisa, however, was not the victim of physical assaults in either of her marriages. She wasn’t revenging herself on an unfaithful husband or, in the case of Collins, killing him to make room for another. Moreover, unless there was an insurance policy that never surfaced, she did not gain any financial benefit from the second killing. Perhaps Besomo and Canon Rich were correct in saying that she was suffering from a neurological or psychological disturbance, although the rewards for killing Andrews would suggest that a more pragmatic reason acted as a trigger in the second killing as well.

One type of husband-killer is the ‘black widow’: a female killer—sometimes a serial killer—who murders family members, usually by slow doses of poison and primarily for financial gain. All the ‘black widow’ boxes are ticked for Andrews’ death, but few for Collins’ death. Still, the driving force appears to be similar. Louisa’s drama began with the illicit thrill of an affair. Then came Andrews’ unemployment, followed by the killing, which empowered her both financially and emotionally, probably for the first time in her life. In the aftermath, she faced a new set of problems. Collins’ laziness created difficulties when the insurance money began to run out. Another trigger seemingly followed: the baby’s death in April 1888. Louisa began to drink heavily. Collins snapped at her and tried to stop her drinking. And Louisa decided to make use of those deadly crystals again.

The prosecution stumbled when attempting to provide a motive for the Collins killing because the only motive it could establish—Collins’ desire to stop Louisa drinking—seemed too weak a trigger in a relationship that otherwise seemed loving. Since Louisa’s day, the guilt/ innocence pendulum has swung both ways. Some writers have concluded that she was innocent because of the problems with the evidence and the lack of a motive for killing Collins. Others have assumed guilt without attempting to explain motivation. A few have proclaimed guilt while providing a motive unsupported by the evidence—for example, a reaction to marital abuse. With the benefit of today’s medical wisdom, we could also propose that postnatal depression might have triggered her actions; however, as this trigger would relate only to the Collins killing, and as serial killers are usually driven by similar underlying motivations, this seems an unlikely explanation.

As it turns out, when the evidence is exhaustively examined, a similar underlying motivation can be glimpsed in the Collins killing. It is clear from Louisa’s actions, conversations and communications that she loathed Andrews yet loved Collins. In fact, she loved Collins so much that she killed to get him. So how would she react if Collins threatened to leave her? Collins was much younger than most contemporary records suggested. Louisa reported that he was aged twenty-nine at the time of his death; however, Collins, by his own more reliable admission, said that he was only twenty-three when they married. Clearly, Collins was little more than a boy—and a lazy one at that. When his beloved son died, he found himself with a lush of a wife who was nearly twice his age, an empty purse, and a bevy of disliked stepchildren to support. What did the relationship hold for him?

That Collins was feeling unhappy about life in general is suggested by his taciturn manner even in the early days of his illness, which contrasted strongly with his garrulity at the dance held shortly after Andrews’ death. That he was less enamoured of Louisa is also suggested by his irritated attitude towards her while on his deathbed.

So, when Collins told Louisa that he would stop her drinking—as the children testified—it seems likely that an ‘or else’ was sooner or later attached. What form could that ‘or else’ have taken? He could have threatened to beat her into submission but, according to the children’s testimonies, that wasn’t the nature of their relationship. Otherwise, Collins had few sources of leverage. He had no money, no assets, no skills; in fact, he had nothing to offer except himself. So, the only ‘or else’ he could have threatened her with was to take himself away—to abandon her.

For Louisa, this ‘or else’ would have transformed an impotent demand for her to stop drinking into a life-changing threat to their relationship—one she had killed to gain. Additionally, it would have served as a challenge to her authority.

In many abusive marriages, the husbands see their wives as ‘possessions’. Some kill fleeing wives partly to stop anyone else having them but, primarily, as the ultimate expression of their power and control. In the Collins household, Louisa ‘wore the pants’. After their marriage, she not only acted as the couple’s voice, she retained control of all of their money, which was at odds with custom in her patriarchal society (and could even be seen as a ‘castrating’ move). Seemingly, she wanted to maintain the power and authority that her first killing had bequeathed her, particularly over her young, lazy, perhaps not-too-smart but much-loved husband. A threat that he would leave, with everything such a threat encompassed, would have served as a powerful trigger for one who had already killed to gain him.

Thus, the only motive that accounts for all the evidence is that Louisa killed Collins because she ‘loved too much’—if such a possessive, controlling attitude could be considered ‘love’. She killed to get him and she killed to ‘keep’ him. As she said in one of her letters, ‘It was not money I wanted from Collins, it was himself.’ Lust and greed had metamorphosed into wrath with devastating consequences for those around her.

Louisa seems to have given Collins low doses of arsenic until the Friday before his death, when the doctor said that he would recover if carefully looked after. Collins’ rapid decline over the next twenty-four hours, along with the large quantity of arsenic found in his body, suggests that she upped the dose after the doctor’s visit. Ironically, it was Collins’ rapid deterioration that most alarmed the doctors when they visited on the Saturday evening and led them to report the case to the police.

It is also possible that Louisa dosed him in front of her neighbours shortly before the doctors arrived on the Saturday night before his death, as he convulsed soon after she gave him a newly concocted drink. Such brazenness would suggest that this type of empowerment excited her, evoking the same type of illicit thrill that her affair had once generated.

The police found no evidence of Rough on Rats in the house despite an extensive search. It seems likely that the box was in Louisa’s purse during her many attempts to leave the house after Collins’ death, and that she eventually disposed of it when she visited Dr Marshall the following day.

A critical mistake, however, was Louisa’s failure to discard the contents of the arsenic-filled tumbler. It seems a strange error for such a clever woman, particularly when she knew the police were involved and that tests had already been undertaken. Or was this not her failing?

The arsenic in the tumbler can only have been introduced in one of two ways. Either, the person responsible for poisoning Collins put it there or someone tampered with the tumbler’s contents after it was collected by Constable Jeffes. While the tumbler evidence is beset with problems, to argue deliberate tampering wanders into conspiracy theory territory. Using Ockham’s Razor again, the simplest or most likely explanation is that Louisa was responsible for the arsenic in the tumbler but, in the stress of the moment, she forgot to get rid of its contents, only realising her omission when Jeffes picked up the tumbler with a view to taking it away.

As for the discrepancy between Jeffes’ and Hamlet’s testimonies regarding the quantity of the liquid, the simplest and most likely explanation is that some of the contents of the unsealed tumbler spilt in the time between collection and testing. It must be remembered that Jeffes carried the unsealed container when he walked the few hundred yards from the Collins house to the police station, and that he later took the container all the way into the city. Nonetheless, the fact that no one could account for the spill raises concern about the use of this evidence to convict her.

If Louisa was tried today for Collins’ death, the tumbler evidence would not be admissible because the chain of custody was not intact. Even if the constable and government analyst massaged their testimonies so they both testified to, say, a ‘half-filled’ glass of milky liquid, the tumbler itself had been left unsealed and unsecured for two days. By contrast, the autopsy doctors carefully sealed and secured their own specimen jars, indicating that the court of the day expected such a level of care with potentially critical evidence.

Other pieces of evidence would also be inadmissible. For example, May’s statement that she had seen Rough on Rats in the house prior to Andrews’ death would be problematic, partly because she revealed a faulty memory during the first trial about events that had happened only a month previously, which should have raised doubts about her likely accuracy when recollecting events that had happened nearly two years previously.

Finally, the evidence regarding Andrews’ death would be inadmissible because of the inability to establish Andrews’ cause of death with certainty, making this ‘similar facts’ evidence more prejudicial than probative. That being the case, if Louisa was brought to trial today for Collins’ murder, it is unlikely that she would have been convicted. Still, current procedure is of little relevance when exploring the reliability of a historical verdict. The question needing to be asked is whether she should have been convicted under the rules in force at the time of her own trial.

It seems unlikely that the Crown would have broken the usual two-trial convention and proceeded with a third Collins trial if the prosecutor hadn’t made the decision to include the critical Andrews ‘similar facts’ evidence. The ruling of the New Zealand Supreme Court suggests that the chief justice should not have allowed its admission; however, Lusk’s failings as Louisa’s counsel meant that the New South Wales courts were unable to answer the admissibility question.

They were, however, able to address the same type of question in an 1893 appeal. The judge presiding over the trial of baby-farmer murderers John and Sarah Makin had admitted ‘similar facts’ evidence regarding the deaths of other babies on the grounds of ‘commonsense’ rather than a legally binding authority. One of the appeal judges, Justice Innes—the same judge who allowed the Collins ‘similar facts’ evidence to be introduced into the Andrews trial (which was acceptable then and would be today)—argued that the ‘commonsense’ approach was unsafe and prejudicial except in cases where there was a large body of independent evidence to support the case; otherwise, an accused person might be tried and convicted of an offence not because of evidence connected with the offence, but because of evidence that the accused had committed other similar offences. However, Innes’ concerns—which reflected the type of concerns expressed by others about the admission of the Andrews evidence to secure Louisa’s conviction—were dismissed by the two other judges presiding over the Makin appeal, who happened to be two of the judges who presided over Louisa’s appeal. Consequently, what the three appeal judges—or any others from the then Supreme Court—would have decided if Louisa’s Queen’s Counsel had been able to raise the ‘similar facts’ question is still open to doubt.

Accordingly, we must turn to the opinions of those well versed in colonial law to determine how the courts should have treated Louisa’s case. For the many reasons mentioned above—the four trials, the problems with the tumbler evidence, the manipulation of the bench, and so on—along with the extremely prejudicial remarks made by the final prosecutor in his closing address, Judge Gregory D. Woods, who wrote A History of Criminal Law in New South Wales, declares that her case was ‘a legal disgrace’.¹¹ There can be no doubt whatsoever that a ‘spirit of vengeance’ drove both the Crown and the judiciary in its pursuit of her conviction and execution.

• • •

Should Louisa have been executed?

Ignoring the big-picture question of the morality and efficacy of capital punishment, we are still left with all the evidentiary and procedural problems identified by Louisa’s supporters and other members of the populace. Most serious of all was the psychological damage inflicted on a child by the knowledge that her words had killed her mother. If ever there was a case that demanded the hand of mercy, Louisa’s was it. Instead, the hand of wrath prevailed.

So why were these problems ignored? Why was she hanged?

Of the two dozen New South Wales women deposed for homicide between 1880 and 1900, nine were convicted; yet Louisa was the only woman to be sent to the gallows—and it wasn’t because she was a suspected double murderer. Baby farmer Sarah Makin, for example, had her 1893 death sentence commuted although she and her executed husband were considered responsible for killing a dozen infants for financial gain. Rather, the driving force in Louisa’s case seems to have been the social and political environment of her time.

Louisa broke not only the written criminal law but the unwritten social law. She breached society’s expectations of ‘womanly’ behaviour when it was particularly unwise to do so. Moreover, the face she presented to the world was cold and unemotional—‘unwomanly’. This was a time when society believed that external appearance denoted internal nature, that crime was most often a male behavioural trait, therefore, an unwomanly woman—that is, a more manly woman—was more likely to be criminal by nature.

This, also, was a time when women were becoming more assertive in demanding equality in employment, educational opportunities and voting rights, so the question of gender inequality was topical. One gender inequality that favoured women was the application of capital punishment. Men were executed; women were not. The Mount Rennie executions two years previously had shone a spotlight on this form of gender inequality. Four youths had been executed for a ‘single’ crime, one considered much less heinous than murder. As one newspaper argued, if men were to be hanged to protect women, then women should also be hanged to protect men.

It was time to set an example, time to show the public that society was not as unequal as the reformers might claim—and, conversely, that society was making efforts to increase women’s autonomy. By a bizarre piece of logical illogic, it was time to hang a woman. The cold-hearted killer of one husband (if not two), a killer who had failed to endear herself to the public and had offered no mitigating explanation, was the ideal choice. Louisa Collins was to be hanged, not only because of what she had done but because a female sacrifice was needed. She was the perfect scapegoat.

Yet even though Louisa breached society’s womanly rules, she was not completely unwomanly. She was an attractive, soberly dressed mother of seven children. Thus, while the ‘hang her’ advocates used womanhood as one of their supportive arguments, the ‘mercy’ advocates were able to do so as well. A tug-of-war began with Louisa in the middle: Louisa the stereotype rather than Louisa the person.

Capital punishment would always be a politically divisive subject; however, when Ninian Melville introduced Louisa’s case into parliament, he forced the members of this especially adversarial House to take sides along party lines. It became another contentious issue confronting parliament. Parkes reacted by, ostensibly, taking the decision-making power away from his cabinet and passing it to the one person who could rise above the maelstrom of New South Wales politics—the governor.

Lord Carrington was an intelligent man with progressive views. He would have seen the myriad problems in the material submitted with the mercy pleas. His words rejecting these mercy pleas suggest that his response would have been different if he had truly enjoyed unfettered authority, but Parkes’ government had made it clear that to pursue such a course would be political suicide. Carrington had little choice but to toe Parkes’ line. His shame at doing so, though, is evident in his refusal to see Louisa’s children.

Ultimately, Louisa was hanged because Parkes was determined to get his own way and was willing to push New South Wales towards a constitutional crisis to achieve it. Yet Parkes was a social progressive who deplored capital punishment and favoured women’s rights, so why was he determined to hang Louisa? He justified his government’s decision on the grounds that capital punishment was ‘the law’, that the law was gender-blind, and that women had the same rights and responsibilities as men so should be treated similarly when they broke the same law. On the surface, this seems a valid ‘progressive’ argument. Nonetheless, Parkes was soon reminded that the law was not gender-blind and that women had no vote so, clearly, they hadn’t the same rights and responsibilities as men. Moreover, as everyone knew, Parkes had himself breached the law six months previously when he responded to public concerns about Chinese immigration. Accordingly, his ‘it’s the law’ argument lacked any moral authority. Even Parkes must have been aware, at some level, of his hypocrisy, especially as the press was keen to point it out. That being the case, it seems surprising that he didn’t dismiss all the rhetoric and embrace the opportunity to grant her a reprieve, particularly in view of his statement to parliament that his government would render every conceivable assistance that would serve the ends of justice in her favour.

The explanation for Parkes’ wrathfulness seems to lie in the community’s long-held and widely embraced views about women. Parkes and his brethren still maintained a patriarchal and patronising attitude to women. While the conservatives were determined to control and silence women for fear that those who had money, knowledge and power would destroy the fabric of society, the progressives were at least willing to allow women a say in their own future so long as the decision was made by men in a timeframe that suited their own agenda.

These men had a primal attitude to power. They saw it as a dichotomy—either they had it or fought to gain it, or they chose to support someone else who had it. They did not share power or accept its imposition lightly. This type of attitude was carried into their strongest relationships with women. Marriage was never intended to be a power-sharing arrangement. The word ‘obey’ made that clear—to everyone.

Then along came a ‘black widow’ who dished out love and arsenic and nearly got away with killing two husbands.

While husband-killing was no longer ‘petit treason’ in the eyes of the law, the notion that this act reflected a profound betrayal of trust and a serious social threat still permeated the community consciousness, as the contemporary articles and correspondence about Louisa’s case reveal. Since husband-killing had long been considered nearly as heinous as killing a king, a woman who could kill two husbands, two ‘petit kings’, must therefore be the personification of evil, a destroyer not just of souls but of society and its core values.

Additionally, Louisa inflicted on her two husbands a humiliating death. Dying in battle was noble. Being poisoned by one’s wife? Effectively, she didn’t just kill them; she castrated them. And impotence for a man is one of the greatest fears of all.

Fear unbinds the ropes that cage superstition. To Parkes, Louisa was a reminder that, while everyman’s home was his castle, Eve still lurked in the dungeon. He stated his attitude clearly when he said, ‘At all times, and under all circumstances, when woman once forgets the character of her sex, there is no barrier to the lengths she will go to in crime.’ Clearly, a message needed to be sent to those lurking Eves.

Evil women were no longer burnt at the stake. The gallows would have to do.