Evidence for a possible early Frontignane bottle / Eléments en faveur d’une frontignane primitive

Papier invité de Tim Eldridge / invited paper by Tim Eldridge

-> Résumé en français en fin d’article!

Evidence for a possible early Frontignane bottle


The mystery of the origin of the Bordeaux wine bottle and why it is known colloquially as the Frontignane is a subject of considerable interest, and not only in France. The excellent twin papers by Dominique Dabas & André Orsini, De la bouteille bordelaise[1], have helped to popularise the mystery and provide clues to the identity of the early Frontignane. It is from these papers that I drew inspiration to conduct my own search to uncover the origins of the wine bottle that was destined to conquer the world.

Yet the identity of the Frontignane is just one clue to an even older mystery; the identity of the traditional Frontignan wine bottle that preceded the Frontignane. As the name suggests, it appears that the Frontignane began as a bottle designed for storing and transporting Frontignan wine, which was spoiled by prolonged storage in oak barrels. According to one legend, the traditional Frontignan bottle provided the template for the Frontignane. I may have the answer to the mystery of the Frontignan, but the history of the two bottles is so intertwined that an explanation of one mystery could not be complete without an explanation of the other. First it is necessary to address the issue of the Frontignane.

Initially I set out to find early French ‘black glass’ bottles that could provide clues to the identity of the earliest Frontignanes. In the process I stumbled across what may actually be one of them (see picture).

The evidence for the illustrated bottle being an early Frontignane comes from the first of the Dabas & Orsini papers, as well as archaeological investigations conducted in Quebec. Because most of the early 18th century trade with New France originated from Bordeaux and nearby ports, including the wine trade[2], the archaeology of Quebec provides a unique window into the evolution of wine bottles manufactured in the Bordeaux region, including the Frontignane.

To be convincing, any explanation of the origin of the Frontignane must be consistent with the historical accounts, including those in the Dabas & Orsini paper. Accordingly, where relevant, each part of my argument is prefaced with an excerpt from the paper.




« … nous avons découvert, dans les registres comptables de la verrerie de Bourg (1734 à 1739) la mention suivante : 14 novembre 1735 : Reçu du sieur Letellier, négociant de Bordeaux, la somme de 3900 Livres Tournoi pour payement de vingt mille huit cents bouteilles frontignanes d’un demy pot…»


 “… we discovered in the accounts of the Bourg glassworks (1734 to 1739) the following mention: “November 14, 1735: Received from Mr Letellier, trader of Bordeaux, the sum of 3900 Pounds as payment for twenty thousand eight hundred frontignane bottles of a half pot… ”

This is one of the earliest known accounts of a Frontignane. If the Frontignane originated in the vicinity of Bordeaux as a distinct type of black glass bottle, then the earliest Frontignane must date from between 14 November 1735 and 1723 when the first black glass bottle factory commenced production in Bordeaux. The Bourg-sur-Gironde glassworks commenced production in 1725.

I estimate the age of my bottle at c1730 plus or minus 5-10 years, but that is not the only evidence I rely upon. It is also very similar to two bottles (#467 and #517) “à col court, épaule tombante” (short neck, sloped shoulder) recovered from an archaeological site in Quebec City and independently estimated to date from c1720[3]. Bottle #467 is pictured below and Table 1 compares the dimensions of the three bottles.



Table 1: Comparison of my bottle with bottles #467 and #517

My bottle #467 #517
Dimensions (mm)
Total height 205 222 214
External diameter of the mouth 25 24 27
Height of the neck 90 95 85
Diameter at the shoulder 119 117 123
Height of the body 65-70 75 75
Diameter of the base 112 112 118
Diameter of the pontil mark 24 24 26
Dimension (ml)
Capacity 796 ~900 ?


The only significant difference between my bottle and #467 is in the height of the neck and body, which results in the different capacity.



Pascal Birolleau, dans sa thèse sur les verreries de Guyenne (1) précise, qu’en 1785, l’ouvrage le plus réclamé est « la grande frontignane du poids de 20 à 22 onces, dont la jauge est de 3 roquilles, les huit faisant le pot, mesure de Bourg égale au pot de Bordeaux soit 0,8515 litre… »


Pascal Birolleau, in his thesis on the Guyenne glassworks(1) specifies that, in 1785, the most sought after product was “the large frontignane weighing 20 to 22 ounces, with a gauge of three roquilles, eight of which make a Bourg pot, equal to the Bordeaux pot, or 0.8515 litres …”

It is possible that the capacity of the Frontignane changed in the 50 years between 1735 and 1785, but so far I have found no evidence for this so I am assuming it remained the same. If Pascal Birolleau’s account is accepted then the large Frontignane had a very small capacity (3 / 8 x 0.8515 litres = 319.3ml).

According to one English historical source, “At Bordeaux, 181.18 pots = 100 English gallons, and each = 127.5 cubic inches”[4]. Therefore, one Bordeaux pot equals 0.552 English gallons (100 / 181.18). In 1819 the English wine gallon equalled 231 cubic inches or 3.785 litres, so one Bordeaux pot is equal to 2.089 litres (0.552 x 3.785).

If one Bordeaux pot consists of eight roquilles then one roquille equals 261.1ml (2.089 / 8) and the capacity of the large Frontignane, equivalent to three roquilles, equals 783.4ml (3 x 261.1).

My bottle holds 796g of water when filled to the brim, or approximately 796ml, assuming one millilitre weighs one gram. If an allowance is made for a cork, the capacity is almost exactly the calculated capacity of the large Frontignane.

One roquille in this case corresponds to what was more commonly known in the 18th century as a demiard or demi-septier (one roquille was normally half of one demiard or demi-septier). This apparent confusion of terms was not unusual in France. For example, the same English source states that “The pot is divided into 2 chopines or 4 demi-septiers”, indicating that 1 pot = 2 chopines = 4 demiards = 8 roquilles, when more commonly the units of volume were expressed as 1 pot = 2 pintes = 4 chopines = 8 demiards = 16 roquilles.

It is possible that some of this confusion may be due to the French government’s failed attempt in 1735 to enforce the Paris pint (pinte de Paris), equal to 0.93 litres, as the standard wine measure throughout France. Incidentally, this might also explain why some early Bordeaux wine bottles seem to have a capacity closer to 900ml than half of one Bordeaux pot (1.045 litres).



« … lors des contrôles réalisés par la jurade de Bourg dans la verrerie de la ville, le 22 février 1741 : «Je me suis certorié que toutes les bouteilles qu’on y fait soient au moins du poids porté par l’arrêt du conseil du 20 déc. dernier, à l’exception seulement de celles que les verriers de cet endroit font pour le vin de Frontignan, d’une jauge extraordinaire et qui ne font qu’environ 20 à 21 onces ; on ne vends point ces dernières au public et on ne les débite qu’en faveur des marchands qui commercent en vin de Frontignan et qui les demandent dans la forme et du poids qu’elles font »

Translation :

… during the controls carried out by the jurade [the royal governing institution] of Bourg in the glassworks of the city, on February 22, 1741: “I certify that all the bottles that in fact are there are at least of the weight carried by the order of the council of 20 Dec. last, except only for those which the glassmakers of this place do make for the wine of Frontignan, of an extraordinary gauge, and which are made to weigh only approximately 20 to 21 ounces; one does not sell these last ones to the public and distributes them only in favour of the merchants who trade in the wine of Frontignan and require them in the form and of the weight that they make” (13).

The unit of measurement in 1741 was the Paris pound (livre de Paris), consisting of 16 ounces (onces). One ounce equals 30.59g, so 21 ounces equals 642.39g (21 x 30.59). The weight of my bottle is 653g (21.35 ounces), consistent with the weight of the bottles reportedly made for the merchants who traded in Frontignan wine.



I acquired my bottle from a place inland from the port of La Rochelle. The bottle was dirty and had not been in a collection so it is more likely to have originated from that region.

Perhaps because New France was first settled from south-west France, most trade with the French colonies was based out of Bordeaux and La Rochelle[5], and most of the wine sold in the colonies came from either south-west France, Spain or Portugal[6]. The bottle factories around Bordeaux also supplied export markets, so it is likely that French black glass bottles found in Quebec originated from the Bordeaux region. Furthermore, any Bordeaux wine bottles excavated in Quebec can be reliably attributed to the period 1723-1760, because the trade in French goods abruptly ceased after the British conquest of New France in 1760.

At one time I thought my bottle might be an early Champagne bottle because it looks similar to two early Champagnes pictured in the book by Willly Van Den Bossche [plates 144(1) and 259(1)]. However, these bottles have quite different weights and capacities so the resemblance is likely to be superficial[7].


Other considerations

Figure 11 of Camille Lapointe’s book illustrates dark green bottles in the classic French ‘flowerpot’ style (pot-de-fleurs) in three different sizes: 500ml, 825ml and 1 litre. The 825ml size is similar to the calculated capacity of the large Frontignane (783.4ml), my bottle (796ml) and a c1820-30 sealed Bordeaux wine bottle in my collection (820ml). The seal on this last bottle reads “Grave” (a prestigious Bordeaux wine region at that time) confirming that it is a true Frontignane.

It is tempting to speculate that Frontignane originally meant a triple demiard or three-quarters of a pint, which at some point became three quarters of a litre (750ml) as it is today. Interestingly, the average reported capacity of the four full-size sealed Bordeaux wine bottles pictured in Willy Van Den Bossche’ book is 752.5ml [see plates 149(4), 266(2), 266(3) and 267(3)]. These bottles all date from the period 1858-1874, well after metric became the norm.

If this theory is correct then the style of the earliest Frontignanes could have varied significantly, provided that the capacity remained the same. However, late 18th century Frontignanes are always an elongated flowerpot shape [for example, see Willy Van Den Bossche plate 149(1)], which almost certainly evolved from a squat flowerpot form. The Quebec archaeological record confirms that a squat flowerpot shape already existed in 1745[8], but the shape of the Frontignane before then is uncertain.

It has been speculated that another c1730 bottle similar to my bottle is “probably an example of the early flower-pot form” [9]. Although my bottle and bottles #467 and #517 are all slightly larger at the shoulder than the base, further research appears necessary to test this theory. If true, it is possible that the Frontignane has always been a flowerpot shape, as well as a triple demiard capacity bottle.


One last problem

The only historical account of the Frontignane I cannot yet fully explain is the 1735 reference to “frontignanes d’un demy pot” (Frontignanes of half a pot). As discussed, half a Bordeaux pot is a pint or a chopine, depending on the historical source, and equivalent to 1.045 litres – not 3 roquilles as was the reported capacity of the large Frontignane in 1785.

For a while I thought that my theory about the Frontignane being a triple demiard bottle must be flawed until something caught my attention that made me wonder whether there is more to this historical account than first meets the eye. The book by Camille Lapointe includes an analysis of deceased estates (les inventaires après décès), mostly from the first half of the 18th century. At least one inventory refers to a black glass bottle with a short neck, a capacity of one pint comprised of three demiards, and which does not take a pint (“qui ne tient pas pinte”)[10].

This contemporary reference to a bottle that apparently looks like a normal pint bottle but does not hold a pint sounds remarkably like my bottle, which looks like a pint bottle (i.e. #467) except that it has a capacity of three demiards.

This interpretation is supported by the pioneering paper by Bernard Ducasse examining the origins of the Bordeaux wine bottle[11]. According to the author:

« Lorsque la Jurade de Sainte-Foy, en 1735 et 1748, interdit les « Bouteilles fortes d’Angleterre ou de Rouen », elle précise que ces bouteilles défendues ne font que les trois quarts de la pinte …


When the Jurade of Sainte-Foy, in 1735 and 1748, banned « Heavy bottles of England or Rouen », it was specified that the authorised bottles are to be made three quarters of a pint [i.e. three demiards] …

Sainte-Foy is an appellation in the eastern part of the Bordeaux wine region known principally for its sweet white wines.



I cannot yet say for certain that my bottle is a Frontignane but so far it appears to satisfy all of the requirements to be one. Only time and further research will tell whether it is.

In this paper I have tried to avoid discussing the fascinating history of Frontignan wine so as not to unduly confuse matters. In my next paper I will discuss the evidence for a possible early Frontignan bottle from approximately the same period.

Tim Eldridge

Fern Tree, Tasmania


14 October 2014


Résumé en français / French summary

L’article de Tim Eldridge est consacré à l’origine et au nom d’une bouteille française typique, connue sous le nom de « frontignane ».

Tim propose une analyse serrée d’une bouteille de sa collection (illustration), provenant vraisemblablement de la région de La Rochelle et qui pourrait être considérée comme une frontignane primitive. La démonstration s’appuie sur une comparaison soigneuse des mesures (taille, poids, contenance) de la bouteille présentée et de ce que l’on sait, par les sources historiques, des mesures de la frontignane de la 1ère moitié du 18ème siècle. Les dimensions de la bouteille de Tim sont notamment très proches de celles d’une bouteille retrouvée à Québec ; seule la contenance est un peu inférieure. La question des contenances est donc étudiée de manière très documentée dans l’article.

[Commentaire webmaster : Personnellement, je ne pense pas que la question des contenances soit décisive, dans la mesure où la normalisation de celles-ci ne semble pas avoir été de mise, ni généralement respectée, au 18ème siècle.]

L’âge de la bouteille étudiée (vers 1730) est confirmé par les exemples publiés et par le matériel archéologique provenant de Québec.

Si elle est bien originaire des environs de Bordeaux, la frontignane a pu être soufflée dès 1723 par les verreries locales (1725 pour Bourg-sur-Gironde) et exportée vers le Nouveau Monde, et en particulier le Canada, jusqu’à la conquête de ce territoire par l’Angleterre, en 1760. Au passage, les bouteilles retrouvées à Québec sont donc d’une aide précieuse pour la datation de leurs équivalents français, dans l’intervalle 1723-1760.

La frontignane tire probablement son nom de la bouteille destinée initialement à contenir du vin de Frontignan. À la fin du 18ème siècle, elle est devenue la bouteille bordelaise par excellence. Tim Eldridge explore ici l’origine de la frontignane bordelaise (vers 1730) et annonce une prochaine publication consacrée à la bouteille destinée au vin de Frontignan.


[1] De la bouteille bordelaise: 1e Partie & 2e Partie, Dominique Dabas and Andre Orsini, 2005

[2] De la qualite des vins en Nouvelle-France, Marc Lafrance, 1992

[3] Le verre et les terres cuites communes de la maison Perthuis a Place-Royale, Camille Lapointe, 1997

[4] see definition of “Pot” in The Cyclopædia; Or, Universal Dictionary of Arts, Sciences, and Literature, Volume 28, Abraham Rees, 1819

[5] The Pattern of French Colonial Shipping to Canada before 1760, James S Pritchard, 1976

[6] De la qualite des vins en Nouvelle-France, Marc Lafrance, 1992

[7] Champagne bottles must be strong to tolerate a high internal pressure and so are always heavy bottles.

[8] See figure 4, Glass Bottle Push-ups and Pontil Marks, Olive Jones, Historical Archaeology 1971

[9] See figure 6, Bottles in the Place Royale Collection by Paul Hanrahan

[10] See Tableau 3, p.41 of Le verre et les terres cuites communes de la maison Perthuis a Place-Royale, Camille Lapointe, 1997

[11] Anciennes Bouteilles Soufflees Et Marquees by Bernard Ducasse, published in Vignobles Et Vins D’Aquitaine, Federation Historique Du Sud-Ouest, 1970

One comment on “Evidence for a possible early Frontignane bottle / Eléments en faveur d’une frontignane primitive

Votre commentaire

Entrez vos coordonnées ci-dessous ou cliquez sur une icône pour vous connecter:

Logo WordPress.com

Vous commentez à l’aide de votre compte WordPress.com. Déconnexion /  Changer )

Image Twitter

Vous commentez à l’aide de votre compte Twitter. Déconnexion /  Changer )

Photo Facebook

Vous commentez à l’aide de votre compte Facebook. Déconnexion /  Changer )

Connexion à %s