英语小说网

Apparently Thorndyke had seen me from the window as I crossed the Walk, for, when I reached the landing, I found him standing in the open doorway of his chambers; and at the sight of him, whatever traces of unreasonable resentment may have lingered in my mind, melted away instantly. He grasped my hand with almost affectionate warmth, and looking at me earnestly and with the most kindly solicitude, said:

“I am glad you have come, Mayfield. I couldn’t bear to think of you alone in your chambers, haunted by this horrible tragedy.”

“You have heard, then—about Barbara, I mean?”

“Yes. Miller called and told me. Of course, he is righteously angry that she has escaped, and I sympathize with him. But for us—for you and me—it is a great deliverance. I was profoundly relieved when I heard that she was gone; that the axe had fallen once for all.”

“Yes,” I admitted, “it was better than the frightful alternative of a trial and what would have followed. But still, it was terrible to see her, lying dead, and to know that it was my hand—the hand of her oldest and dearest friend—that had struck the blow.”

“It was my hand, Mayfield, not yours that actually struck the blow. But even if it had been yours instead of your agent’s, what could have been more just and proper than that retribution should have come through the hand of the friend and guardian of that poor murdered girl?”

I assented with a shudder to the truth of what he had said, but still my mind was too confused to allow me to see things in their true perspective. Barbara, my friend, was still more real to me than Barbara the murdress. He nodded sympathetically enough when I explained this, but rejoined, firmly:

“You must try, my dear fellow, to see things as they really are. Shocking as this tragedy is, it would have been immeasurably worse if that terrible woman had not received timely warning. As it is, the horrible affair has run its course swiftly and is at an end. And do not forget that if the axe has fallen on the guilty its menace has been lifted from the innocent. Madeline Norris and Anthony Wallingford will sleep in peace to-night, free from the spectre of suspicion that has haunted them ever since Harold Monkhouse died. As to the woman whose body you found this morning, she was a monster. She could not have been permitted to live. Her very existence was a menace to the lives of all who came into contact with her.”

Again, I could not but assent to his stern indictment and his impartial statement of the facts.

“Very well, Mayfield,” said he. “Then try to put it to yourself that, for you, the worst has happened and is done with. Try to put it away as a thing that now belongs to the past and is, in so far as it is possible, to be forgotten.”

“As far as is possible,” I repeated. “Yes, of course, you are quite right, Thorndyke. But forgetfulness is not a thing which we can command at will.”

“Very true,” he replied. “But yet we can control to a large extent the direction of our thoughts. We can find interests and occupations. And, speaking of occupations, let me show you some of Polton’s productions.”

He rose, and putting a small table by the side of my chair, placed on it one or two small copper plaques and a silver medallion which he had taken from a drawer. The medallion was the self-portrait of Stella which had lain dormant in the wax mould through all the years which had passed since her death, and as I took it in my hand and gazed at the beloved face, I found it beautiful beyond my expectations.

“It is a most charming little work,” I said, holding it so that the lamp light fell most favourably on the relic, “I am infinitely obliged to you, Thorndyke.”

“Don’t thank me,” said he. “The whole credit is due to Polton. Not that he wants any thanks, for the work has yielded him hours of perfect happiness. But here he is with the products of another kind of work.”

As he spoke, Polton entered with a tray and began in his neat, noiseless way, to lay the table. I don’t know how much he knew, but when I caught his eye and his smile of greeting, it seemed to me that friendliness and kindly sympathy exuded from every line of his quaint, crinkly face. I thanked him for his skilful treatment of my treasures and then, observing that he was apparently laying the table for supper, would have excused myself. But Thorndyke would hear of no excuses.

“My dear fellow,” said he, “you are the very picture of physical exhaustion. I suspect that you have had practically no food to-day. A meal will help you to begin to get back to the normal. And, in any case, you mustn’t disappoint Polton, who has been expecting you to supper and has probably made a special effort to do credit to the establishment.”

I could only repeat my acknowledgments of Polton’s goodness (noting that he certainly must have made a special effort, to judge by the results which began to make themselves evident) and, conquering my repugnance to the idea of eating, take my place at the table.

It is perhaps somewhat humiliating to reflect that our emotional states, which we are apt to consider on a lofty spiritual plane, are controlled by matters so grossly material as the mere contents of our stomachs. But such is the degrading truth, as I now realized. For no sooner had I commenced a reluctant attack on the products of Polton’s efforts and drunk a glass of Burgundy—delicately warmed by that versatile artist to the exact optimum temperature—than my mental and physical unrest began to subside and allow a reasonable, normal outlook to develop, with a corresponding bodily state. In effect, I made quite a good meal and found myself listening with lively interest to Thorndyke’s account of the technical processes involved in converting my little plaster plaques and the wax mould into their final states in copper and silver.

Nevertheless, in the intervals of conversation the unforgettable events of the morning and the preceding night tended to creep back into my consciousness; and now a question which I had hitherto hardly considered began to clamour for an answer. Towards the end of the meal, I put it into words. Apropos of nothing in our previous conversation, I asked:

“How did you know, Thorndyke?” and as he looked up inquiringly, I added: “I mean, how were you able to make so confident a guess, for, of course, you couldn’t actually know?”

“When do you mean?” he asked.

“I mean that when you applied for a Home Office authority you must have had something to go on beyond a mere guess.”

“Certainly I had,” he replied. “It was not a guess at all. It was a certainty. When I made the application I was able to say that I had positive knowledge that Stella Keene had been poisoned with arsenic. The examination of the poor child’s body was not for my information. I would have avoided it if that had been possible. But it was not. As soon as my declaration was made, the exhumation became inevitable. The Crown could not have prosecuted on a charge of poisoning without an examination of the victim’s body.”

“But, Thorndyke,” I expostulated, “how could you have been certain—I mean certain in a legal sense? Surely it could have been no more than a matter of inference.”

“It was not,” he replied. “It was a matter of demonstrated fact. I could have taken the case into court and proved the fact of arsenical poisoning. But, of course, the jury would have demanded evidence from an examination of the body, and quite properly, too. Every possible corroboration should be obtained in a criminal trial.”

“Certainly,” I agreed. “But still I find your statement incomprehensible. You speak of demonstrated fact. But what means of demonstration had you? There was my diary. I take it that that was the principal source of your information; in fact I can’t think of any other. But the diary could only have yielded documentary evidence, which is quite a different thing from demonstrated fact.”

“Quite,” he agreed. “The diary contributed handsomely to the train of circumstantial evidence that I had constructed. But the demonstration—the final, positive proof—came from another source. A very curious and unexpected source.”

“I suppose,” said I, “as the case is finished and dealt with, there would be no harm in my asking how you arrived at your conclusion?”

“Not at all,” he replied. “The whole investigation is a rather long story, but I will give you a summary of it if you like.”

“Why a summary?” I objected. “I would rather have it in extenso if it will not weary you to relate it.”

“It will be more likely to weary you,” he replied. “But if you are equal to a lengthy exposition, let us take to our easy chairs and combine bodily comfort with forensic discourse.”

We drew up the two arm-chairs before the hearth, and when Polton had made up the fire and placed between us a small table furnished with a decanter and glasses, Thorndyke began his exposition.

“This case is in some respects one of the most curious and interesting that I have met with in the whole of my experience of medico-legal practice. At the first glance, as I told you at the time, the problem that it presented seemed hopelessly beyond solution. All the evidence appeared to be in the past and utterly irrecoverable. The vital questions were concerned with events that had passed unrecorded and of which there seemed to be no possibility that they could ever be disinterred from the oblivion in which they were buried. Looking back now on the body of evidence that has gradually accumulated, I am astonished at the way in which the apparently forgotten past has given up its secrets, one after another, until it has carried its revelation from surmise to probability and from probability at last to incontestible proof.

“The inquiry divides itself into certain definite stages, each of which added new matter to that which had gone before. We begin, naturally, with the inquest on Harold Monkhouse, and we may consider this in three aspects: the ascertained condition of the body; the evidence of the witnesses; and the state of affairs disclosed by the proceedings viewed as a whole.

“First, as to the body: there appeared to be no doubt that Monkhouse died from arsenical poisoning, but there was no clear evidence as to how the poison had been administered. It was assumed that it had been taken in food or in medicine—that it had been swallowed—and no alternative method of administration was suggested or considered. But on studying the medical witnesses’ evidence, and comparing it with the descriptions of the patient’s symptoms, I was disposed to doubt whether the poison had actually been taken by the mouth at all.”

“Why,” I exclaimed, “how else could it have been taken?”

“There are quite a number of different ways in which poisonous doses of arsenic can be taken. Finely powdered arsenic is readily absorbed by the skin. There have been several deaths from the use of ‘violet powder’ contaminated with arsenic, and clothing containing powdered arsenic would produce poisonous effects. Then there are certain arsenical gases—notably arsine, or arseniureted hydrogen—which are intensely poisonous and which possibly account for a part of the symptoms in poisoning from arsenical wall-papers. There seemed to me to be some suggestion of arsenical gas in Monkhouse’s case, but it was obviously not pure gas-poisoning. The impression conveyed to me was that of a mixed poisoning; that the arsenic had been partly inhaled and partly applied to the skin, but very little, if any, taken by the mouth.”

“You are not forgetting that arsenic was actually found in the stomach?”

“No. But the quantity was very minute; and a minute quantity is of no significance. One of the many odd and misleading facts about arsenic poisoning is that, in whatever way the drug is taken, a small quantity is always found in the stomach and there are always some signs of gastric irritation. The explanation seems to be that arsenic which has got into the blood in any way—through the skin, the lungs or otherwise—tends to be eliminated in part through the stomach. At any rate, the fact is that the presence of minute quantities of arsenic in the stomach affords no evidence that the poison was swallowed.”

“But,” I objected, “what of the Fowler’s Solution which was found in the medicine?”

“Exactly,” said he. “That was the discrepancy that attracted my attention. The assumption was that deceased had taken in his medicine a quantity of Fowler’s Solution representing about a grain and a half of arsenious acid. If that had been so we should have expected to find a very appreciable quantity in the stomach: much more than was actually found. The condition of the body did not agree with the dose that was assumed to have been taken; and when one came to examine the evidence of the various witnesses there was further room for doubt. Two of them had noticed the medicine at the time when the Fowler’s Solution had not been added; but no witness had noticed it after the alleged change and before the death of deceased. The presence of the Fowler’s Solution was not observed until several days after his death. Taking all the facts together, there was a distinct suggestion that the solution had been added to the medicine at some time after Monkhouse’s death. But this suggestion tended to confirm my suspicion that the poison had not been swallowed. For the discovery of the Fowler’s Solution in the medicine would tend to divert inquiry—and did, in fact, divert it—from any other method of administering the poison.

“To finish with the depositions: not only was there a complete lack of evidence even suggesting any one person as the probable delinquent; there was not the faintest suggestion of any motive that one could consider seriously. The paltry pecuniary motive applied to all the parties and could not be entertained in respect of any of them. The only person who could have had a motive was Barbara. She was a young, attractive woman, married to an elderly, unattractive husband. If she had been attached to another man, she would have had the strongest and commonest of all motives. But there was nothing in the depositions to hint at any other man; and since she was absent from home when the poisoning occurred, she appeared to be outside the area of possible suspicion.

“And now to look at the evidence as a whole: you remember Miller’s comment. There was something queer about the case; something very oddly elusive. At the first glance it seemed to bristle with suspicious facts. But when those facts were scrutinized they meant nothing. There were plenty of clues but they led nowhere. There was Madeline Norris who prepared the victim’s food—an obvious suspect. But then it appeared that the poison was in the medicine, not in the food. There was Wallingford who actually had poison in his possession. But it was the wrong poison. There was the bottle that had undoubtedly contained arsenic. But it was nobody’s bottle. There was the bottle that smelled of lavender and had red stains in it and was found in Miss Norris’s possession; but it contained no arsenic. And so on.

“Now all this was very strange. The strongest suspicion was thrown on a number of people collectively. But it failed every time to connect itself with any one individually. I don’t know precisely what Miller thought of it, but to me it conveyed the strong impression of a scheme—of something arranged, and arranged with extraordinary skill and ingenuity. I had the feeling that, behind all these confusing and inconsistent appearances, was a something quite different, with which they had no real connection; that all these apparent clues were a sort of smoke-screen thrown up to conceal the actual mechanism of the murder.

“What could the mechanism of the murder have been? That was what I asked myself. And by whom could the arrangements have been made and carried out? Here the question of motive became paramount. What motive could be imagined? And who could have been affected by it? That seemed to be the essential part of the problem, and the only one that offered the possibility of investigation.

“Now, as I have said, the most obvious motive in cases of this kind is that of getting rid of a husband or wife to make room for another. And ignoring moral considerations, it is a perfectly rational motive; for the murder of the unwanted spouse is the only possible means of obtaining the desired release. The question was, could such a motive have existed in the present case; and the answer was that, on inspection, it appeared to be a possible motive, although there was no evidence that it actually existed. But, assuming its possibility for the sake of argument, who could have been affected by it? At once, one saw that Madeline Norris was excluded. The death of Harold Monkhouse did not affect her, in this respect, at all. There remained only Barbara and Wallingford. To take the latter first: He was a young man, and the wife was a young, attractive woman; he had lived in the same house with her, appeared to be her social equal and was apparently on terms of pleasant intimacy with her. If he had any warmer feelings towards her, her husband’s existence formed an insuperable obstacle to the realization of his wishes. There was no evidence that he had any such feelings, but the possibility had to be borne in mind. And there were the further facts that he evidently had some means of obtaining poisons and that he had ample opportunities for administering them to the deceased. All things considered, Wallingford appeared, prima facie, to be the most likely person to have committed the murder.

“Now to take the case of Barbara. In the first place, there was the possibility that she might have had some feeling towards Wallingford, in which case she would probably have been acting in collusion with him and her absence from home on each occasion when the poisoning took place would have been part of the arrangement. But, excluding Wallingford, and supposing her to be concerned with some other man, did her absence from home absolutely exclude the possibility of her being the poisoner? There were suggestions of skilful and ingenious arrangements to create false appearances. Was it possible that those arrangements included some method by which the poison could be administered during her absence without the connivance or knowledge of any other person?

“I pondered this question carefully by the light of all the details disclosed at the inquest; and the conclusion that I reached was that, given a certain amount of knowledge, skill and executive ability, the thing was possible. But as soon as I had admitted the possibility, I was impressed by the way in which the suggestion fitted in with the known facts and served to explain them. For all the arranged appearances pointed to the use of Fowler’s Solution, administered by the mouth. But this could not possibly have been the method if the poisoner were a hundred miles away. And as I have said, I was strongly inclined to infer, from the patient’s symptoms and the condition of the body, that the poison had not been administered by the mouth.

“But all this, as you will realize, was purely hypothetical. None of the assumptions was supported by a particle of positive evidence. They merely represented possibilities which I proposed to bear in mind in the interpretation of any new evidence that might come into view.

“This brings us to the end of the first stage; the conclusions arrived at by a careful study of the depositions. But following hard on the inquest was your visit to me when you gave me the particulars of your past life and your relations with Barbara and Monkhouse. Now your little autobiographical sketch was extremely enlightening, and, as it has turned out, of vital importance. In the first place, it made clear to me that your relations with Barbara were much more intimate than I had supposed. You were not merely friends of long standing; you were virtually in the relation of brother and sister. But with this very important difference: that you were not brother and sister. An adopted brother is a possible husband; an adopted sister is a possible wife. And when I considered your departure to Canada with the intention of remaining there for life, and your unexpected return, I found that the bare possibility that Barbara might wish to be released from her marriage had acquired a certain measure of probability.

“But further; your narrative brought into view another person who had died. And the death of that person presented a certain analogy with the death of Monkhouse. For if Barbara had wished to be your wife, both these persons stood immovably in the way of her wishes. Of course there was no evidence that she had any such wish, and the death of Stella was alleged to have been due to natural causes. Nevertheless, the faint, hypothetical suggestions offered by these new facts were strikingly similar to those offered by the previous facts.

“The next stage opened when I read your diary, especially the volume written during the last year of Stella’s life. But now one came out of the region of mere speculative hypothesis into that of very definite suspicion. I had not read very far when, from your chance references to the symptoms of Stella’s illness, I came to the decided conclusion that, possibly mingled with the symptoms of real disease, were those of more or less chronic arsenical poisoning. And what was even more impressive, those symptoms seemed to be closely comparable with Monkhouse’s symptoms, particularly in the suggestion of a mixed poisoning partly due to minute doses of arsine. I need not go into details, but you will remember that you make occasional references to slight attacks of jaundice (which is very characteristic of arsine poisoning) and to ‘eye-strain’ which the spectacles failed to relieve. But redness, smarting and watering of the eyes is an almost constant symptom of chronic arsenic poisoning. And there were various other symptoms of a decidedly suspicious character to which you refer and which I need not go into now.

“Then a careful study of the diary brought into view another very impressive fact. There were considerable fluctuations in Stella’s condition. Sometimes she appeared to be so far improving as to lead you to some hopes of her actual recovery. Then there would be a rather sudden change for the worse and she would lose more than she had gained. Now, at this time Barbara had already become connected with the political movement which periodically called her away from home for periods varying from one to four weeks; and when I drew up a table of the dates of her departures and returns, I found that the periods included between them—that is the periods during which she was absent from home—coincided most singularly with Stella’s relapses. The coincidence was so complete that, when I had set the data out in a pair of diagrams in the form of graphs, the resemblance of the two diagrams was most striking. I will show you the diagrams presently.

“But there was something else that I was on the lookout for in the diary, but it was only quite near the end that I found it. Quite early, I learned that Stella was accustomed to read and work at night by the light of a candle. But I could not discover what sort of candle she used; whether it was an ordinary household candle or one of some special kind. At last I came on the entry in which you describe the making of the wax mould; and then I had the information that I had been looking for. In that entry you mention that you began by lifting the reflector off the candle, by which I learned that the receptacle used was not an ordinary candlestick. Then you remark that the candle was of ‘good hard wax’; by which I learned that it was not an ordinary household candle—these being usually composed of a rather soft paraffin wax. Apparently, it was a stearine candle such as is made for use in candle-lamps.”

“But,” I expostulated, “how could it possibly matter what sort of candle she used? The point seems to be quite irrelevant.”

“The point,” he replied, “was not only relevant; it was of crucial importance. But I had better explain. When I was considering the circumstances surrounding the poisoning of Monkhouse, I decided that the probabilities pointed to Barbara as the poisoner. But she was a hundred miles away when the poisoning occurred; hence the question that I asked myself was this: Was there any method that was possible and practicable in the existing circumstances by which Barbara could have arranged that the poisoning could be effected during her absence? And the answer was that there was such a method, but only one. The food and the medicine were prepared and administered by those who were on the spot. But the candles were supplied by Barbara and by her put into the bedside candle-box before she went away. And they would operate during her absence.”

“But,” I exclaimed, “do I understand you to suggest that it is possible to administer poison by means of a candle?”

“Certainly,” he replied. “It is quite possible and quite practicable. If a candle is charged with finely powdered arsenious acid—‘white arsenic’—when that candle is burnt, the arsenious acid will be partly vaporized and partly converted into arsine, or arseniureted hydrogen. Most of the arsine will be burnt in the flame and reconverted into arsenious acid, which will float in the air, as it condenses, in the form of an almost invisible white cloud. The actual result will be that the air in the neighbourhood of the candle will contain small traces of arsine—which is an intensely poisonous gas—and considerable quantities of arsenious acid, floating about in the form of infinitely minute crystals. This impalpable dust will be breathed into the lungs of any person near the candle and will settle on the skin, from which it will be readily absorbed into the blood and produce all the poisonous effects of arsenic.

“Now, in the case of Harold Monkhouse, not only was there a special kind of candle, supplied by the suspected person, but, as I have told you, the symptoms during life and the appearances of the dead body, all seemed to me to point to some method of poisoning through the lungs and skin rather than by way of the stomach, and also suggested a mixed poisoning in which arsine played some part. So that the candle was not only a possible medium of the poisoning; it was by far the most probable.

“Hence, when I came to consider Stella’s illness and noted the strong suggestion of arsenic poisoning; and when I noted the parallelism of her illness with that of Monkhouse; I naturally kept a watchful eye for a possible parallelism in the method of administering the poison. And not only did I find that parallelism; but in that very entry, I found strong confirmation of my suspicion that the candle was poisoned. You will remember that you mention the circumstance that on the night following the making of the wax mould you were quite seriously unwell. Apparently you were suffering from a slight attack of acute arsenical poisoning, due to your having inhaled some of the fumes from the burning candle.”

“Yes, I remember that,” said I. “But what is puzzling me is how the candles could have been obtained. Surely it is not possible to buy arsenical candles?”

“No,” he replied, “it is not. But it is possible to buy a candle-mould, with which it is quite easy to make them. Remember that, not so very long ago, most country people used to make their own candles, and the hinged moulds that they used are still by no means rare. You will find specimens in most local museums and in curio shops in country towns and you can often pick them up in farm-house sales. And if you have a candle-mould, the making of arsenical candles is quite a simple affair. Barbara, as we know, used to buy a particular German brand of stearine candles. All that she had to do was to melt the candles, put the separated wicks into the mould, stir some finely-powdered white arsenic into the melted wax and pour it into the mould. When the wax was cool, the mould would be opened and the candles taken out—these hinged moulds usually made about six candles at a time. Then it would be necessary to scrape off the seam left by the mould and smooth the candles to make them look like those sold in the shops.”

“It was a most diabolically ingenious scheme,” said I.

“It was,” he agreed. “The whole villainous plan was very completely conceived and most efficiently carried out. But to return to our argument. The discovery that Stella had used a special form of candle left me in very little doubt that Barbara was the poisoner and that poisoned candles had been the medium used in both crimes. For we were now out of the region of mere hypothesis. We were dealing with genuine circumstantial evidence. But that evidence was still much too largely inferential to serve as the material for a prosecution. We still needed some facts of a definite and tangible kind; and as soon as you came back from your travels on the South-Eastern Circuit, fresh facts began to accumulate. Passing over the proceedings of Wallingford and his follower and the infernal machine—all of which were encouraging, as offering corroboration, but of no immediate assistance—the first really important accretion of evidence occurred in connection with our visit to the empty house in Hilborough Square.”

“Ha!” I exclaimed. “Then you did find something significant, in spite of your pessimistic tone at the time? I may say, Thorndyke, that I had a feeling that you went to that house with the definite expectation of finding some specific thing. Was I wrong?”

“No. You were quite right. I went there with the expectation of finding one thing and a faint hope of finding another; and both the expectation and the hope were justified by the event. My main purpose in that expedition was to obtain samples of the wall-paper from Monkhouse’s room, but I thought it just possible that the soot from the bedroom chimneys might yield some information. And it did.

“To begin with the wall-paper: the condition of the room made it easy to secure specimens. I tore off about a dozen pieces and wrote a number on each, to correspond with numbers that I marked on a rough sketch-plan of the room which I drew first. My expectation was that if—as I believed—arsenical candles had been burnt in that room, arsenic would have been deposited on all the walls, but in varying amounts, proportionate to the distance of the wall from the candle. The loose piece of paper on the wall by the bed was, of course, the real touchstone of the case, for if there were no arsenic in it, the theory of the arsenical candle would hardly be tenable. I therefore took the extra precaution of writing a full description of its position on the back of the piece and deposited it for greater safety in my letter-case.

“As soon as I reached home that day I spread out the torn fragment on the wide stage of a culture microscope and examined its outer surface with a strong top light. And the very first glance settled the question. The whole surface was spangled over with minute crystals, many of them hardly a ten-thousandth of an inch in diameter, sparkling in the strong light like diamonds and perfectly unmistakable; the characteristic octahedral crystals of arsenious acid.

“But distinctive as they were, I took nothing for granted. Snipping off a good-sized piece of the paper, I submitted it to the Marsh-Berzelius test and got a very pronounced ‘arsenical mirror,’ which put the matter beyond any possible doubt or question. I may add that I tested all the other pieces and got an arsenic reaction from them all, varying, roughly, according to their distance from the table on which the candle stood.

“Thus the existence of the arsenical candle was no longer a matter of hypothesis or even of mere probability; it was virtually a demonstrated fact. The next question was, who put the arsenic into the candle? All the evidence, such as it was, pointed to Barbara. But there was not enough of it. No single fact connected her quite definitely with the candles, and it had to be admitted that they had passed through other hands than hers and that the candle-box was accessible to several people, especially during her absence. Clear evidence, then, was required to associate her—or some one else—with those poisoned candles, and I had just a faint hope that such evidence might be forthcoming. This was how I reasoned:

“Here was a case of poisoning in which the poison was self-administered and the actual poisoner was absent. Consequently it was impossible to give a calculated dose on a given occasion, nor was it possible to estimate in advance the amount that would be necessary to produce the desired result. Since the poison was to be left within reach of the victim, to be taken from time to time, it would be necessary to leave a quantity considerably in excess of the amount actually required to produce death on any one occasion. It is probable that all the candles in the box were poisoned. In any case, most of them must have been; and as the box was filled to last for the whole intended time of Barbara’s absence, there would be a remainder of poisoned candles in the box when Monkhouse died. But the incident of the ‘faked’ medicine showed that the poisoner was fully alive to the possibility of an examination of the room. It was not likely that so cautious a criminal would leave such damning evidence as the arsenical candles in full view. For if, by chance, one of them had been lighted and the bearer had developed symptoms of poisoning, the murder would almost certainly have been out. In any case, we could assume that the poisoner would remove them and destroy them after putting ordinary candles in their place.

“But a candle is not a very easy thing to destroy. You can’t throw it down a sink, or smash it up and cast it into the rubbish-bin. It must be burnt; and owing to its inflammability, it must be burnt carefully and rather slowly; and if it contains a big charge of arsenic, the operator must take considerable precautions. And finally, these particular candles had to be burnt secretly.

“Having regard to these considerations, I decided that the only safe and practicable way to get rid of them was to burn them in a fireplace with the window wide open. This would have to be done at night when all the household was asleep, so as to be safe from interruption and discovery; and a screen would have to be put before the fireplace to prevent the glare from being visible through the open window. If there were a fire in the grate, so much the better. The candles could be cut up into small pieces and thrown into the fire one at a time.

“Of course the whole matter was speculative. There might have been no surplus candles, or if there were, they might have been taken out of the house and disposed of in some other way. But one could only act on the obvious probabilities and examine the chimneys, remembering that whereas a negative result would prove nothing for or against any particular person, a positive result would furnish very weighty evidence. Accordingly I collected samples of soot from the various bedroom chimneys and from that of Barbara’s boudoir, labelling each of them with the aid of the cards which you had left in the respective rooms.

“The results were, I think, quite conclusive. When I submitted the samples to analysis I found them all practically free from arsenic—disregarding the minute traces that one expects to find in ordinary soot—with one exception. The soot from Barbara’s bedroom chimney yielded, not mere traces, but an easily measurable quantity—much too large to have been attributable to the coal burnt in the grate.

“Thus, you see, so far as the murder of Monkhouse was concerned, there was a fairly conclusive case against Barbara. It left not a shadow of doubt in my mind that she was the guilty person. But you will also see that it was not a satisfactory case to take into court. The whole of the evidence was scientific and might have appeared rather unconvincing to the ordinary juryman, though it would have been convincing enough to the judge. I debated with myself whether I should communicate my discoveries to the police and leave them to decide for or against a prosecution, or whether I should keep silence and seek for further evidence. And finally I decided, for the present, to keep my own counsel. You will understand why.”

“Yes,” said I. “You suspected that Stella, too, had been poisoned.”

“Exactly. I had very little doubt of it. And you notice that in this case there was available evidence of a kind that would be quite convincing to a jury—evidence obtainable from an examination of the victim’s body. But here again I was disposed to adopt a waiting policy for three reasons. First, I should have liked to avoid the exhumation if possible. Second, if the exhumation were unavoidable, I was unwilling to apply for it until I was certain that arsenic would be found in the body; and third, although the proof that Stella had been poisoned would have strengthened the case enormously against Barbara, it would yet have added nothing to the evidence that a poisoned candle had been used.

“But the proof of the poisoned candle was the kernel of the case against Barbara. If I could prove that Stella had been poisoned by means of a candle, that would render the evidence absolutely irresistible. This I was not at present able to do. But I had some slight hopes that the deficiency might be made up; that some new facts might come into view if I waited. And, as there was nothing that called for immediate action, I decided to wait, and in due course, the deficiency was made up and the new facts did come into view.”

As he paused, I picked up Stella’s medallion and looked at it with a new and sombre interest. Holding it up before him, I said:

“I am assuming, Thorndyke, that the new facts were in some way connected with this. Am I right?”

“Yes,” he replied, “you are entirely right. The connection between that charming little work and the evidence that sent that monster of wickedness to her death is one of the strangest and most impressive circumstances that has become known to me in the whole of my experience. It is no exaggeration to say that when you and Stella were working on that medallion, you were forging the last link in the chain of evidence that could have dragged the murdress to the gallows.”

He paused, and, having replenished my glass, took the medallion in his hand and looked at it thoughtfully. Then he knocked out and refilled his pipe and I waited expectantly for the completion of this singular story.