Twice Told
Tale
Period
13th century to 20th century

Saint Guinefort, the greyhound in the well

A greyhound killed for saving a child was honoured as a martyr near Lyon. The friar who recorded the cult says he had the dog dug up and its bones burned.

Contents
  1. Paragraph 370 of a book for preachers
  2. Novile, Villars or Sandrans
  3. After the fire
  4. A dog with a saint’s name
  5. What we could not verify
  6. Sources

Published 20 Oct 2026. Sources checked 17 Oct 2026.

Before he died around 1261, Stephen of Bourbon, a Dominican inquisitor, wrote that in the diocese of Lyon he had had a dog’s grave dug up and the bones burned.61 The dog was a greyhound, and mothers had been carrying their children to it as to a saint.7 Stephen says he heard of it in the confessional, took Guinefort for some saint, and asked who he was.7 His page is the earliest account of the cult we found, and Jorion calls him its first learned reporter.2

The story, as Stephen of Bourbon wrote it down, goes like this.

Near a village of nuns called Novile, on the land of the lord of Villars, stood a castle whose lord had a baby son.8 One day the lord, his wife and the nurse all went out and left the child alone in his cradle.8 A huge snake came into the house and made for the cradle.8 The greyhound that had stayed behind went after it, and in the fight the cradle was knocked over.8 The dog killed the snake and threw it far from the child, then stood by the cradle, mauled and smeared with the snake’s blood.8

The nurse came in, saw the blood, thought the dog had eaten the child and screamed.8 The mother ran in and thought the same.8 The knight came up, drew his sword and killed the dog.8 Then they found the child unhurt and asleep, and the snake torn by the dog’s bites.8 They threw the dog into a well in front of the castle gate, heaped a great pile of stones on it and planted trees beside it.9

By God’s will the castle was destroyed and the land left a wilderness.10 Country people heard of the dog’s noble deed, visited the place, honoured the dog as a martyr and prayed there for their illnesses.10

Paragraph 370 of a book for preachers

Stephen was born at Belleville-sur-Saône, belonged to the Dominicans and held the office of inquisitor by mandate of the Holy See.6 In his last years, at the Dominican house in Lyon, he wrote a collection of short stories for preachers to use in sermons.11 The Guinefort story is paragraph 370 of it, in Lecoy de La Marche’s edition of 1877.7 Stephen files it under superstitions that insult God, next to women who ask for health by worshipping elder trees and carry their children to anthills.12

He gives no year. The practice was recent, he says (“nuper”), and nothing on the four printed pages dates the confessions, his visit or the killing of the dog.13 His frame is plain: the people who honoured the dog as a martyr were deceived by the devil.10

What the women did at the place, in his account, was elaborate. A mother took a sick child to an old woman in a castle a league away, who taught her the ritual and led her to the spot.14 They offered salt, hung the child’s rags on the bushes and stuck a needle into the trees.14 Then the naked child was passed nine times through the gap between two tree trunks, the mother on one side and the old woman on the other, while they called on the fauns of the forest of Rimite to take the sick child and bring back their own, fat and well.14

After that the child was laid on the straw of a cradle at the foot of a tree, two candles a thumb’s measure long were lit and fixed in a log, and the mothers went away and stayed away until the candles had burned out.15 The candles burned very many children to death, Stephen writes, “sicut ibidem de aliquibus reperimus”, as we found in some cases there.15 One woman told him she had seen a wolf come out of the wood toward her child, and that it would have eaten it, or the devil in the wolf’s shape, if she had not run back.15 A child found alive was carried to the river Chalarone and dipped nine times.16

Then comes the burning.

Canem mortuum fecimus exhumari et lucum succidi, et cum eo ossa dicti canis pariter concremari1

We had the dead dog dug up and the grove cut down, and the bones of the dog burned together with it.

Our translation

The “it” is the grove, as we read the Latin, since the bones are already the dog’s. Stephen writes “we” and does not say who held the spade.1 He adds that he had the lords of the land post an edict about the spoliation and ransom of anyone who gathered there for such a cause.1 Reviewers of Schmitt’s book report the act as his: he had the dog exhumed and burned (Dubois), had the wood cut (Velay-Vallantin), destroyed the cult (Jorion).17

The passage reached print in 1719, in Quétif and Échard’s Scriptores ordinis praedicatorum, 158 years before Lecoy’s edition.18 Échard says that whether the dog really fought the snake is nothing to Stephen, who only reports what people said, but that nobody can doubt the horrible superstition was in use.19

Novile, Villars or Sandrans

Stephen names no wood and no parish, only a nuns’ village called Novile in the land of the lord of “Vilarius”.20 Échard, in 1719, glosses Novile as possibly Villeneuve in the Dombes and the lord’s seat as Villars on the Chalaronne.20 Lecoy takes Villars as Villars-en-Dombes.20

The modern location is different. Reviewers report that Schmitt puts the legend and the rite in the bois de Saint-Guinefort, in what is now the commune of Sandrans, in the Ain.21 Isambert’s review of the book has weak children laid in a wood of Saint Guinefort at Neuville-les-Dames, also in the Dombes.22 We could not read Schmitt and cannot say how he joins these names.

Geary reports that digs made since the book appeared found a fortified dwelling on the site, built and abandoned soon after, probably in the twelfth century, which he takes to support Schmitt’s link between the cult and the twelfth-century clearing of the land.23 Chanaud says Schmitt reads in the legend a quarrel between peasants and lords who were tempted to dig fish ponds.24 Stephen’s text does not name the wood, and the identification at Sandrans reaches us only through reviews of a book we could not open.

After the fire

The burning and the edict did not end the cult. The reports we found run like this.

DateWhat is reportedWhere
Before about 1261The dog dug up, the grove cut, an edict posted16Near Novile, diocese of Lyon8
1632A notarial deed, the only evidence for an old chapel25The place of the cult
1719Échard prints Stephen’s passage18Paris
1826Legend and rite still attested, much changed3Sandrans
1873Guigue: the devotion would still exist, women invoke him for husbands or children3Romans
After the First World WarThe cult finally vanishes, in George’s reading of Schmitt4The Dombes

Guigue’s book of 1873 reaches us through Lecoy’s footnote, in the conditional (“existerait encore”).3 Chanaud, reporting Schmitt, says the legend and rite at Sandrans were still attested in 1826, with notable changes.3 Jorion says the cult, destroyed in the thirteenth century by an inquisitor, was found again in the nineteenth by folklorists in a barely changed form.26 George writes that, for Schmitt, it did not truly disappear until the aftermath of the First World War.4 Dubois says a popular cult was kept up “until today” in the wood.27 Isambert says the belief lasted at least six centuries despite the efforts to uproot it.22

So the reviews give “destroyed” and “persisted” in the same breath. Nobody we read says the burning worked.

A dog with a saint’s name

Pen and ink drawing of a greyhound standing in profile, facing left, with a long narrow head, deep chest and tucked waist

A greyhound in profile, drawn in Germany in 1549, nearly three centuries after Stephen of Bourbon wrote of the dog near Lyon. It is not an image of Guinefort or of the legend. Virgil Solis, A Greyhound in Profile Facing Left, pen and brown ink over graphite, 1549. The Metropolitan Museum of Art, 2001.497, The Elisha Whittelsey Collection, The Elisha Whittelsey Fund, 2001, public domain (CC0).

Schmitt’s Le saint lévrier came out in 1979, and in English in 1983 as The Holy Greyhound.28 The story of a dog killed by its master after saving a child from a snake is a widespread type. Geary reports twelve medieval Western versions, and Lecoy’s footnote names an analogous legend in Wales and one in the Indian book of Sindibad, from which the romance of the Seven Sages came.29 Échard adds that Johannes Junior retold the story with details and an allegory in his Scala coeli.19

The rite is a separate thing. Schmitt explains it by the belief in changelings, in which vaguely defined spirits carry off children and leave their own in their place.30 Stephen’s women called on fauns for the same exchange.14

The name is the hardest part. Isambert says the hagiography knows a Saint Guinefort of Pavia, whose legend, written in the eighth century, is a kind of doublet of that of Saint Sebastian, and whose cult appears in several places in Italy and France, always for healing children.31 Chanaud reports Schmitt’s answer to how a dog got the name: Cluniac monks carried the martyr’s cult toward the Saône, the monks of Gigny passed with relics of Saint Taurin in 1158, and the martyr’s feast on 22 August falls in the dog days.5 Geary calls the Cluny hypothesis reasonable, though not certain.32 Isambert says nothing in the archaeology of the place supports a pre-Christian animal cult.33

Nothing we read says the Church declared the greyhound a saint. Stephen’s answer was the fire, and we found no text in which the Church recognised the dog or turned him down.

What we could not verify

Sources we have not seen

  • Schmitt’s Le saint lévrier (1979) and The Holy Greyhound (1983). It would give his survey of the site, the sources behind the dates of the cult and his twelve versions of the legend. The lending copy on the Internet Archive opens no text, and we found no other copy online.
  • Guigue, Topographie historique de l’Ain (1873), Lecoy’s source for the devotion at Romans. It would show how the cult was described a century before Schmitt. Gallica refused our request on 17 October 2026 and the Internet Archive has no copy.
  • Rebecca Rist’s article on the papacy, the Inquisition and Saint Guinefort (Reinardus, 2019). It would show how a recent historian reads Stephen’s action. The repository refused access, and the Project MUSE page gave only an excerpt.
  • Johannes Junior’s Scala coeli, which Échard says retells the dog and the snake. It would date a preacher’s use of the story outside Stephen’s circle. We did not locate a text.

Seen only where others quote them

  • What Schmitt argues about the Pavia martyr, the changelings, the fish ponds and the end of the cult, which we read only where seven reviewers summarise him (Dubois, George, Chanaud, Isambert, Geary, Jorion, Velay-Vallantin). They are not independent of each other on what Schmitt says.

Questions we could not settle

  • Whether the Church ever ruled on Guinefort. No text we read records a decision, and Stephen’s own answer was to destroy the place.
  • Which wood Stephen meant. Novile and Vilarius are glossed as Villeneuve and Villars, while the reviews put the cult at Sandrans.
  • When the cult ended. Reviewers give 1826, the nineteenth century, the end of the First World War and “today”.

Sources

Documents quoted