The Flies That Pointed to the Killer / 蝇验镰：宋慈让一群苍蝇指认凶手

Forensic Entomology Six Hundred Years Before Forensic Entomology

From The Coroner's Notebook — a retelling of cases from Xi Yuan Ji Lu (洗冤集录), Volume II: Difficult and Doubtful Cases (Part Two), Case 6

By Song Ci (宋慈, 1186–1249) · Retold by Cathay Tales

Somewhere in the rice country south of the Yangtze, summer, around 1240. A farmer is found dead in a field path with more than ten sickle-cuts on his body. His silver and his clothes are still on him. The coroner takes one look, decides it is not a robbery, and asks every household in the surrounding hamlet to bring out their sickles. He lays seventy of them in a row in the dirt, in full noon sun, and waits. The flies do the rest.

The Case

The body was found in the morning, lying in a path between rice paddies. A man in his thirties. A peasant. Locally known. He had not been gone long — the family had reported him missing only the night before.

The first guess, the obvious guess, was robbery. The roads outside the county walls were not safe in the summer of a poor harvest, and the sickle was the standard farmhand's sidearm: cheap, sharp, easy to swing, easy to hide in a sleeve. Travelers were knifed for their copper. The coroner was prepared, walking out to the scene, to record one more highway killing.

Then he counted the wounds.

There were more than ten of them — slashes, not stabs — distributed across the dead man's chest, shoulders, and forearms. The forearms were what caught the coroner's eye. They were defensive wounds: the kind a man takes on his crossed forearms while trying to ward off a blade swung at his face. He had seen the killer coming. He had fought back. He had lost.

The other thing the coroner noticed was the dead man's clothing. The waistband was untouched. The sleeve where copper would normally be tucked was untouched. The little hempen pouch at his hip — checked carefully — still held its few small coins.

The coroner stood over the body and did the arithmetic out loud, for the constables and the family and the small crowd that had gathered along the dyke. A robber wants the man dead and the goods gone. This man is dead and his goods are still here. Whoever swung the sickle was not after copper. Whoever swung the sickle was after this particular man.^{[1]Song Ci (宋慈, 1186–1249): Author of Xi Yuan Ji Lu (洗冤集录, "The Washing Away of Wrongs"), the world's first systematic treatise on forensic medicine, completed in 1247. Born in Jianyang, Fujian — the same county that produced the Neo-Confucian master Zhu Xi a century earlier — Song Ci passed the jinshi examination in 1217 and spent his career as a judicial official in the southern provinces, rising to tí xíng àn cá shǐ (提刑按察使, judicial inspector) for Guangdong, Jiangxi, Guangxi, and Hunan. The manual was distilled from his own field experience and from a half-dozen earlier (now mostly lost) Song-dynasty coroner handbooks, including the Nèi Shù Lù (内恕录) cited in his preface.}

He turned, then, to the wife.

"Did your husband have an enemy?" he asked her. "Anyone he had quarreled with — recently — to the point of bad blood?"

She thought about it. She started to say no, my husband was a peaceful man, and then she stopped and corrected herself. There was one thing. A neighbor — she gave a name — had come to the house a few weeks earlier asking to borrow money. Her husband had refused. The neighbor had not taken it well. There had been words. The neighbor had said something to the effect of I will get back what is owed me, one way or another.

"Was the bad blood deep enough to kill over?" the coroner asked.

The wife shook her head. "No. Not that deep. But that is the only quarrel I can think of."

The coroner thanked her, told the constables to keep her at the scene for the moment, and quietly noted the name of the neighbor and the location of his house. Then he stepped out onto the dyke and gave a different order — one that, for a man standing over a fresh corpse on a hot afternoon, must have sounded slightly unhinged.

"Go to every household within a li of this field," he told the constables. "Tell them the magistrate's office is collecting all of the sickles in the hamlet for inspection. Every household must bring every sickle they own. Anyone who hides a sickle, or claims to own none when the neighbors say they do, will be presumed the killer. Make sure they understand that part."^{[2]The "compulsory production" technique: Song Ci's instruction that every household in the area must bring every sickle, and any household concealing a tool will be presumed the killer is a deliberately designed game-theoretic trap. Once the order is issued, the killer's only options are (a) hand over the murder weapon and trust that wiping has rendered it indistinguishable from the others, or (b) refuse to hand it over and be presumed guilty by default. Both options, Song Ci's procedure ensures, lead to detection. This is one of several places in Xi Yuan Ji Lu where Song Ci shows an explicit awareness of suspect psychology and incentive design — a feature of his work that distinguishes it from earlier coroner manuals, which tend to focus narrowly on what to do with the body.}

The constables looked at each other. They went.

The Sickles

By the middle of the afternoon, the dyke was lined with sickles.

Seventy or eighty of them — the old text says qī bā shí zhāng (七八十张), seventy or eighty pieces — laid out side by side along the dirt path in the order the constables had collected them. Each one was tagged with the name of the household that had brought it. The coroner walked the row slowly, the way a man inspecting cavalry mounts might walk down the line of horses, looking at each blade in turn.

There was no visible blood on any of them. Whoever had swung a sickle that morning had wiped it clean. Sickles are easy to wipe clean. A handful of straw, a stream, two minutes — and the iron looks the same as it did the day before.

The coroner did not pick up any of them. He did not test them with his thumb. He did not hold them up to the light. He simply walked back to the dyke, sat down in what shade was available, and waited.

The day was hot. The text says shí fāng shèng shǔ (时方盛暑) — the season was at the height of the summer heat. The flies were everywhere. They had been around the body all morning. Now they were drifting up and down the row of sickles, settling, lifting off, settling again.

After perhaps half an hour — perhaps an hour, the original is not specific — something happened that everyone watching could see.

The flies began to gather, in numbers, on a single sickle.

It was three or four sickles in from one end of the row. To the eye, it looked the same as the seventy-something others. The blade was clean. The handle was clean. There was no visible blood, no smear, no rust-brown discoloration on the metal. But the flies, undeceived by anyone's wiping, had found something on it that the rest of the row did not have. They were settling on the blade and the joint and the wooden handle in a slow, attentive cluster — the same way they had been settling on the dead man.^{[3]The fly-on-blade observation: The original text says only nèi liǎn dāo yī zhāng, yíng zǐ fēi jí (内镰刀一张, 蝇子飞集) — "among them, on one sickle, the flies gathered and flew in." The English version expands the sentence to make the comparative-test logic visible to the reader, but the core observation in the Chinese is exactly that: one blade out of seventy or eighty drew the flies. The compactness of the original (the whole case occupies fewer than 200 characters in classical Chinese) is a feature of the manual's prose style — Song Ci writes for working magistrates who do not have time to read paragraphs.}

The coroner walked over. He did not touch the sickle. He pointed to it and asked, calmly:

"Whose is this one?"

There was a long pause. Then a man stepped forward out of the crowd of householders standing along the dyke, and acknowledged that the sickle was his. It was the neighbor whose name the wife had given an hour before — the man who had asked to borrow money and been refused.

The coroner had him brought up to the row.

"Look at the others," he said. "None of the other men's sickles have flies on them. Yours has flies. The blood-smell is still on it. The flies smell what we cannot. You have killed a man with this sickle. The blood is on it. How can you deny it?"

The neighbor, the original says, kòu shǒu fú zuì (叩首服罪) — kowtowed and confessed his crime.

The crowd along the dyke, the original adds, shī shēng tàn fú (失声叹服) — lost their voices in astonishment and admiration.

The Detective Behind the Case

The coroner in this case is not named in the original. Xi Yuan Ji Lu is not a casebook of Song Ci's personal investigations; it is a manual of procedures, illustrated with anonymized cases drawn from his own service and from earlier authorities. The narrative voice is third-person, and most of the detective-figures in the book are referred to simply as jiǎn guān (检官) — "the inspecting official" — or tǐ jiū guān (体究官), "the investigating official."

Song Ci himself was, by 1247, the senior judicial inspector for four southern provinces — Guangdong, Jiangxi, Guangxi, and Hunan. He had served as a county-level coroner in his thirties and forties, then risen through the system on the strength of a single quality his preface emphasizes again and again: a refusal to delegate the inspection of corpses to subordinates. Most of his peers, Song Ci notes with controlled disgust, yáo wàng ér fú qīn, yǎn bí ér bù xiè (遥望而弗亲, 掩鼻而不屑) — they look from a distance without going near, they cover their nose and refuse to stoop. The result, he says, is bad verdicts and unavenged dead.

The seventy-sickle case appears in Volume II of his manual, in a section titled Yí Nán Zá Shuō Xià (疑难杂说下) — "Difficult and Doubtful Matters, Part Two." It is sandwiched between an unrelated procedure for distinguishing a drowning victim from a body dumped into water after death (which involves pouring hot water through the nostrils of the skull and watching for sand grains to come out), and an even more elaborate routine for examining ancient skeletal remains. Song Ci does not single it out as exceptional. To him it is simply Case 6 in a chapter of difficult cases — one example of how a careful magistrate, refusing to settle for the obvious robbery verdict, can read both the wounds on a body and the behavior of insects in the air to identify a killer.

How the Trick Actually Works

The biology Song Ci was relying on is not, in fact, mysterious. It just had no name yet in 1247.

Common blowflies (Calliphoridae — the green and blue iridescent flies anyone who has been around a corpse on a hot day will recognize) navigate by olfaction. Their antennae are studded with chemoreceptors specifically tuned to the volatile compounds released by decomposing protein and oxidized iron — chiefly dimethyl trisulfide, putrescine, cadaverine, and the iron-bound aromatics released as hemoglobin breaks down. Modern entomological studies have measured the threshold of detection at around 0.04 milligrams of blood per liter of air — roughly four parts per hundred million. A blowfly can locate a fresh wound from forty meters downwind on a still day. A sickle that has been used to kill a man, no matter how thoroughly it has been wiped, will retain blood residue inside the handle joint, in the crack between the iron blade and the wooden socket, and in any micro-pitting on the iron surface. The cleaned-looking blade gives off a chemical signature that is invisible to the human eye and overwhelming to the fly.^{[4]Modern blowfly olfaction: The 0.04 mg/L blood-detection threshold cited in modern Chinese forensic-history retrospectives appears to derive from olfactometer studies on Lucilia sericata (the common green bottle fly) and Calliphora vomitoria (the bluebottle), conducted at agricultural-entomology laboratories in the 1970s and 1980s. The exact threshold varies with humidity, temperature, and individual fly age, but the order of magnitude is robust. For a comparison, the human nose's threshold for the same volatile sulfur compounds is roughly 1,000 times higher.}

What Song Ci's coroner did, in effect, was build a comparative test with seventy controls and one suspect, all subjected to identical environmental conditions — same sun, same air, same fly population, same dirt surface, same time of exposure. The flies, acting as living chemical sensors, sorted the suspect from the controls without any human intervention. The coroner did not have to claim to see anything special. He had only to point at the place the flies pointed.

This is, in modern terms, a single-blind comparative chemical assay with biological detectors. The Western forensic literature does not formalize this approach until Pierre Mégnin publishes La Faune des Cadavres in 1894, founding modern forensic entomology — six hundred and forty-seven years after Song Ci's text.

Translator's Notebook

What gets me, every time I re-read this case, is how theatrical the procedure is.

Song Ci could have asked his constables to canvas the village, find the man with motive, search his property, and confront him with the missing money story. That is what a modern detective would do. It would have produced, in this case, exactly the same suspect — the wife had already named him.

But Song Ci did not want the same suspect. He wanted a confession in front of the entire village. He wanted seventy households to stand on a dyke in the summer heat and watch the flies pick out one sickle from a row of seventy. He wanted the killer to be unmasked not by a magistrate's authority — I have decided you are guilty — but by something the killer himself could not argue with: a swarm of flies that had no opinion about money lent or refused, and no reason to lie.

I think this is the part of the Xi Yuan Ji Lu that is genuinely modern. Song Ci is not just teaching his magistrates how to find the truth. He is teaching them how to find the truth in a way that the village will accept as truth. A confession extracted by interrogation in a back room, in 1240 China, was not believable to anyone outside the room. A confession produced by seventy sickles and a column of flies, witnessed by the entire neighborhood, was unanswerable. The killer kowtowed and confessed because he had nowhere left to stand.

I had to look up, while writing this, whether modern forensic entomologists have actually tested the seventy-sickles trick. The answer, as far as I can find, is no — nobody has run the controlled experiment with seventy real sickles on a hot day. But the underlying biology is well established, and the threshold numbers (0.04 mg/L for blowfly detection of blood) come from independent modern studies. The trick should work. It works, on paper, the way Song Ci said it works in 1247.

The other thing I keep thinking about is the wife. She gives the coroner the killer's name in the first ten minutes of the investigation. He has it. He has the motive — refused loan — and the location of the suspect's house. He could have ridden over there with two constables and made an arrest before lunch.

He chose, instead, the long way around. He chose to spend the whole afternoon laying sickles in the sun. He chose to make the village watch. There is a reading of this case that says Song Ci did not trust himself to make the call alone — that he wanted the flies to make it for him. There is another reading that says he did not trust the village to accept the call if he made it alone — that he wanted the flies to make it for them.

I think both readings are right. He did not trust anyone with that kind of decision, including himself, and so he handed it to the only witnesses in the case who had no skin in the game.

The flies were the only honest people in the county that day.

Next case: The Hairpin That Caught the Drowning — how a Song coroner distinguished between a man who fell into the river and a man who was thrown in already dead. → [Coming soon]