Does the Classification rank the soil?

In 1767 a man named Saint-Pierre bought an estate in Saint-Julien called Serançan and gave it his own name. That is how the Bordeaux directory of Cocks and Féret tells it, in a footnote: a purchase and a new name. In 1855, 88 years later, his family's growth went into the official Classification of the Médoc as a Fourth.

In 1942, in an occupied country whose wine trade had almost nothing to sell, a cooper's son from Saint-Julien registered a name for an estate he had not finished buying. Henri Martin had been born in 1903 at Château Gruaud-Larose, where his grandfather ran the cellar and his father made the barrels. The name was Gloria, chosen, by the estate's own account, for the choir of classified vineyard voices he meant to gather into one song. The same year he made his first serious purchase, 6 hectares from Duhart-Milon, a classified Fourth Growth of Pauillac.

He kept buying for 40 years. Parcel by parcel he took vineyard out of 9 classified growths of Saint-Julien and Pauillac, pieces of Beychevelle, Saint-Pierre, both Léovilles, Gruaud-Larose, Ducru-Beaucaillou, until he had about 50 hectares in the middle of the best commune in the Médoc. Almost every row of it had been classified ground. None of it was classified once it was his, and none of it is classified now.

So one buyer gave Serançan a new name, and 88 years later that name was a Fourth Growth. Another gathered about 50 hectares, almost all of it classified ground, and came away with no rank at all. Between them sits the question of this case. When the brokers ranked the Médoc in 1855, what exactly did they rank?

The answer everybody reaches for is that the 1855 Classification is a survey of the Médoc's sacred earth, that the brokers walked the slopes and graded the gravel, and that Martin's parcels lost their standing the moment they were cut out of the estates the survey had blessed. It is the most natural belief in wine. The trade's own books test it in the driest way available to them, which is by bookkeeping, and you can follow the rank and the ground through their tables.

One Léoville becomes three

Start with a division. In 1824 William Franck, the Bordeaux merchant whose Traité sur les vins du Médoc sat on every dealer's desk in the city, printed a ranking of the Médoc at the back of his first edition. His Second Growths were 4 names: Brane-Mouton, Rauzan, Léoville, and Gruaud or Larose.

By his third edition of 1853, Léoville alone takes up 3 lines, and Franck writes in the table exactly why. Beside the name he prints Léoville, divisé en trois, Léoville divided in three. Then come the 3 proprietors: the Marquis de Las Cases at 80 to 100 tonneaux, the Baron Poyféré de Cères at 40 to 50, Barton at 50 to 70.

A few lines above, Gruaud-Larose, the estate where Martin would be born, is divisé entre Messieurs Bethmann, Boisgérard, and the heirs of the Baron Sarget. The brokers of Bordeaux drew up their list for Napoleon III's Exposition Universelle in Paris 2 years later, the list that became the 1855 Classification, and carried every one of those divisions into the official table. The Médoc got 3 Léovilles where it had one, 2 Gruaud-Laroses, 2 Pichon-Longuevilles, and 2 Saint-Pierres.

Saint-Pierre, the old Serançan, was one of the split ones. Franck printed it in 1853 as St-Pierre divisé entre Bontemps du Barry at 25 to 35 tonneaux and Roullet and Galoupeau at 25 to 35. Armand d'Armailhacq, whose treatise on the vines and wines of the Médoc appeared in 1855, found it split 3 ways on the ground.

When the directory reproduced the official table in 1874, the Saint-Pierre line still read like a clerical stutter: Saint-Pierre, Saint-Julien, Mme Bontemps du Barry, and then Id. Id. O. de Luetkens. The same growth in the same commune stood under 2 proprietors, and both of them were Fourth Growths.

Split a classified vineyard between 2 heirs and you do not get 2 half-ranks. You get 2 classified growths, each entitled to the full price of its class, and no authority is asked about any of it. The hectares are the same hectares they were the week before the will was read.

Martin would later buy from both Léovilles, from Gruaud-Larose and from Saint-Pierre. By then each was a classified growth with a rank of its own, and every parcel he bought out of them came to Gloria without it.

A growth that left its rank only because of its division

Run the machine the other way and you get Angludet, at Cantenac. The directory's own history of the estate is brisk. In 1150 it belonged to a knight called Bertrand d'Angludet; in the 18th century it was the property of Pierre Legras; in 1791 it was divided into 4 parts. Then comes the sentence, and Féret printed it in the sixth edition of 1893, in the seventh of 1898, and again in the ninth of 1922:

This growth, classed a Fourth before the Revolution, left that high rank only because of its division.

« Ce cru, classé 4e avant la Révolution, ne sortit de ce rang élevé qu'à cause de sa division. »

Cocks and Féret, Bordeaux et ses vins classés par ordre de mérite, 6th edition, Bordeaux, 1893, Cantenac; repeated in the 7th edition of 1898 and the 9th of 1922

Only because of its division. Not the soil, and not the winemaking. The estate was cut into 4 pieces for 4 sons, the rank had nothing left to stand on, and the people who kept the register wrote that down as the reason, 3 times, across 29 years. In 1855, the year of the list, d'Armailhacq records Angludet in 3 hands.

Look at the ground itself. The directory says the vineyard runs to 55 hectares on fine sandy-gravel rises over an iron subsoil, and shares its boundaries with Giscours, Cantenac-Brown, Desmirail, and Kirwan. That is a classified Third, a classified Third, a classified Third, and a classified Third.

The English edition of 1883 says the wines are especially appreciated in England for their delicacy, their bouquet, their color and their development in bottle. Then it prints the sentence that matters: "They have often been preferred to 3rd growths, which is explained by the fine position of the vineyard, which besides is completely surrounded with 2nd and 3d classed growths." There are whole parcels of carmenère there more than a century old.

Nor was it small. In 1893, 1898, and 1922 the estate was making 150 tonneaux of wine a year, while in the same commune the classified Le Prieuré made 45 and the classified Pouget made 25.

The 4 parts did come back together. The directory says so in the same paragraph: divided in 1791 into 4, and exactly a century later reunited in the hands of Monsieur Jules Jadouin. The land was whole again, the neighbors were still Giscours and Kirwan, the English still preferred the wine to Third Growths, and the rank did not return, because by then the thing that had been classified in 1855 was a list of names and Angludet was not on it. In the years of its division the estate had been called, of all things, La République.

A name could be made, until 1855

Which brings the story back to Serançan, and to the door Martin ran into. Before 1855 a name could be made. The footnote that records the renaming adds that the wines of Saint-Pierre sometimes fetch the price of the Thirds.

It happened at Cantenac too. In 1814 an English officer bought Château de Gascq and called it Palmer. Franck's 1824 table records the change in 2 columns, old proprietor De Gasq, new proprietor Palmer, and by 1855 Palmer was a classified Third Growth. A private buyer renamed a piece of the Médoc after himself and the trade began printing the new name. In time the new name was worth money.

After April 1855 that stopped working. The list has been amended twice since, in September of its own year and by ministerial decree in 1973. On the account of Dewey Markham Jr., whose 1997 book is a history of the Classification, both amendments moved estates that were already inside the system. Serançan had become Saint-Pierre with nothing more than a deed and a printing habit. Martin had pieces of the same ground and a good deal more patience, and he was 175 years late.

Subtraction behaves the same way. When a classified estate sells off a piece of itself, the piece leaves without a rank, and the clearest case is Palmer's own. Part of the Palmer vineyard was detached and became Château Montbrun; Franck was printing it as its own property in 1853, 2 years before the brokers met, at 15 to 20 tonneaux. Nobody classified it.

What it took away instead was the word. Its wines, the English edition of 1883 says, "after its separation from chateau Palmer have borne for many years the brand of « Montbrun-Palmer »." The rank stayed with the estate that kept the whole name.

What the brokers were actually doing

They deserve fairness, because the system is not a swindle and the men who built it were not fools. They had about 2 weeks to rank the Médoc from sale records for a fair in Paris. A sale record contains a name and a price. It does not contain a cadastral boundary, and in 1855 a great many parcels had no boundary anybody had drawn. Names could be followed through the ledgers for 50 years, and land could not, so ranking marks was the only thing the available paper would support.

The ground is not irrelevant either. The classified growths began on the best gravel, which is why the old list still works as a rough guide to price, as the economists who keep testing it keep finding. Franck describes the floor of Saint-Julien in 1845 in a line a farmer would recognize: a black gravel, strong and barely sandy, lying on a bottom of clay and marl. The directory calls the ring of estates around Saint-Pierre "that glorious pleiad." Martin bought into that, and he was not buying a technicality.

The most classified estate in Saint-Julien

Martin understood the machinery long before anybody wrote it down for him. He liked to raise his glass to Gloria as the most classified estate in Saint-Julien, which is true of its soil and false of its standing.

The rest came anyway. In 1963 he was handed the keys to Château Latour and ran it, with Jean-Paul Gardère, for 24 years. He was mayor of his commune for about 40. He founded the Commanderie du Bontemps. In 1976 an American magazine called his wine a miracle and the American market came looking for it. None of that is a rank.

In 1982 he did the only thing that has ever worked. He bought Château Saint-Pierre, the estate a buyer had named after himself in 1767, and with it the Fourth Growth that name carries (the estate's own page gives 1981). Assembling classified ground for 40 years had produced no rank at all, and one purchase of a classified word produced one immediately. His family runs both houses today, and on the other side of that one transaction Angludet has been waiting since 1791.

The verdict

The claim on trial is narrower than the defense of the brokers, and it is the one people actually repeat: that the list ranks land, so classified land makes classified wine and unclassified land cannot. The books say no. What was ranked in 1855 was a set of marks that happened, that April, to be sitting on good gravel, and the marks have been sliding around on the gravel ever since.

False. The 1855 ranked prices, and prices tracked names. That the names sat on good gravel is the coincidence history forgets. Gloria farms 50 hectares of the proof.

The wine

Gloria is the Saint-Julien I put in front of people about to spend 3 times as much for a classified label from the same road, and it wins more of those arguments than it loses. The style is the commune's own, cedar and dark fruit with the tannins folded in rather than stacked up, more polished than a Pauillac and firmer than a Margaux, and it drinks well earlier than its neighbors without giving up the years at the other end. The vines are the point. This is old classified vineyard farmed as one estate, and it tastes like it. Give it 8 to 10 years, buy the strong vintages by the case, and pay no attention to anybody who calls it a bourgeois wine, which is a fact about a list rather than a description of anything in the glass.

Chronology

Sources

Period sources, read in the original.

• William Franck, Traité sur les vins du Médoc et les autres vins rouges du département de la Gironde, 1st edition, Bordeaux, 1824, the classification and the Cantenac table.
• William Franck, Traité sur les vins du Médoc, 2nd edition, Bordeaux, 1845, the description of Saint-Julien and the Cantenac table.
• William Franck, Traité sur les vins du Médoc, 3rd edition, Bordeaux, 1853, the classification and its divisions, and the Cantenac table.
• Charles Cocks, Bordeaux, ses environs et ses vins classés par ordre de mérite, Bordeaux, Féret fils, 1850, the Cantenac and Saint-Julien pages.
• A. d'Armailhacq, La culture des vignes, la vinification et les vins dans le Médoc, Bordeaux, 1855, pp. 492 to 493 for Saint-Pierre, and pp. 504 and 518 for Cantenac and Angludet.
• Charles Cocks and Édouard Féret, Bordeaux et ses vins classés par ordre de mérite, 2nd edition, Bordeaux, 1868, the reproduction of the official table.
• Charles Cocks and Édouard Féret, Bordeaux et ses vins classés par ordre de mérite, 3rd edition, Bordeaux, 1874, the official table and the Saint-Pierre and Angludet footnotes.
• Charles Cocks and Édouard Féret, Bordeaux et ses vins classés par ordre de mérite, 5th edition, Bordeaux, 1886, Cantenac.
• Charles Cocks and Édouard Féret, Bordeaux et ses vins classés par ordre de mérite, 6th edition, Bordeaux, 1893, Cantenac, the Angludet entry.
• Charles Cocks and Édouard Féret, Bordeaux et ses vins classés par ordre de mérite, 7th edition, Bordeaux, 1898, Cantenac and Saint-Julien.
• Charles Cocks and Édouard Féret, Bordeaux et ses vins classés par ordre de mérite, 9th edition, Bordeaux, 1922, Cantenac and Saint-Julien.
• Charles Cocks and Édouard Féret, Bordeaux and its Wines Classified in Order of Merit, 2nd English edition, Bordeaux, 1883, Cantenac and Saint-Julien, for Angludet, Montbrun, and Saint-Pierre.

Modern scholarship.

• Dewey Markham Jr., 1855: A History of the Bordeaux Classification, John Wiley and Sons, New York, 1997, for the composition of the brokers' document and its 2 amendments.

The estate's own account.

• Domaines Henri Martin, the Château Gloria page, for Martin's birth at Gruaud-Larose, the first hectare, the registration of the name in 1942 and its origin, the purchase from Duhart-Milon, the toast, the American article of 1976, and the acquisition of Saint-Pierre. domaines-henri-martin.com

How this was checked

Every period quotation was read in the original text, in a private library of 244,000+ wine documents, the oldest from 1373. Where an English edition is quoted its wording is kept, spelling and all; the other translations from the French are mine, with the French beneath the lines that matter most so that you can check one against the other. If a page cited here does not say what I say it says, write to me at antonio@bordeauxreviews.com and I will correct it.

To cite this article: Antonio Assumpção, "Does the Classification rank the soil?", Bordeaux Reviews, Château Case Files 08/70, 2026, https://bordeauxreviews.com/case-files/does-the-classification-rank-the-soil/.

Antonio "Guto" Assumpção stands at the intersection of law and wine. Admitted to practice law in Florida, New York, and Rio de Janeiro, he is also a French Wine Scholar and a Master of Bordeaux by the Wine Scholar Guild and holds a WSET Level 3 certification. Antonio has been studying, collecting, and tasting Bordeaux wines for the past 20 years, and his articles blend a detailed analysis with a connoisseur's insight into the world of wine. Follow him on Instagram at @BordeauxReviews.