UMC - Grandes Marques et Maisons de Champagne

Champagne Guest Book

The Making of Sparkling Wine

Perseverance and the spirit of invention finally perfected the techniques involved in the production of champagne and it was only when this had been achieved that a huge growth in its sales became possible. As late as 1866, in the fifth edition of Jullien’sTopographie de tous les vignobles connus (Topography of all Known Vineyards), he declares that the phenomena that create or destroy the quality of sparkling wines are so astonishing that they cannot be explained. As William Younger very justly said, up until the middle of the century champagne has been a difficult wine, young and unstable, sometimes sparkling in the natural sense of the term, sometimes simply "cremant". Scientific research and the improvement of techniques has enabled it to become a consistent and trustworthy wine, that is of better quality and less expensive [1].

This progress was due to not only French but also German and Swiss scientists, who followed in the footsteps of Lavoisier, and amongst whom numbered Chaptal, Cadet de Vaux, Thénard, Nicolas de Saussure, Mitscherlich, Gay Lussac, Leuwenhoek, Appert, Cagniard de La Tour, Pasteur, Berthelot, Claude Bernard, Liebig, Buchner. Credit is also due to the researchers in Champagne who, while less famous, played a key role in the practical domain; amongst these were, to name just the central figures, Herpin, François, Maumené, Robinet, Cordier, Salleron,, and Manceau. Vine growers and merchants also played a part in the development of techniques, offering both ideas and financial support, and began to systematically reap the benefits of the possibilities that became available towards the end of the century. Couanon and Convert noted in 1900 that the practice of carrying out quick chemical tests of musts and wines has become widespread amongst vineyard owners and merchants [2].


The wine that undergoes primary fermentation is made with particular care in Champagne when it is destined to become sparkling; this is made clear by André Jullien who wrote, Champagne is undisputedly the region where industry has made the most progress in the manipulation of wines. They are clarified with extreme care and bottled only when they are perfectly clear. The vinification equipment was maintained in excellent condition, this included the innumerable barrels (each marked with letters and numbers that identified where the wine came from, the particular marc, grapes and pressing), the drawing off tanks, and the large funnels which were still made of copper in order to avoid any risk of ferric casse.

There were no notable differences in terms of wine-making technique between Champagne and the other wine regions of France.

The advice of Chaptal was universally followed. He wrote at the beginning of the century on the subject of "sugaring" the must, that the addition of sugar has the double advantage of considerably increasing the ’spirituosité’ (alcohol content) of the wine53 , and of preventing the acidic degeneration to which weaker wines are subject. The usual objective was to reach an alcohol content of between 11.5° and 12° at the end of fermentation. Some wine makers in order to obtain the same result would, add a few litres of well-flavoured spirit of cognac, after the wine had fermented. Parmentier, for his part, advised grape syrup, or concentrated grape juice, which he said sweetened the wines of Champagne more advantageously than sugar candy.

Before the barrels were filled with the sweetened must, they would be disinfected by burning a wick or sheet of sulphur suspended inside them from an iron hook. Maumené states that at the time (1871) it was customary to mix aromatic powders with the sulphur such as clove, cinnamon, flowers of lavender, thyme, marjoram, violet, and Florence iris. He adds that the smallest quantity of these substances can be beneficial to the wine’s flavour.

Below is a rough outline of the process of transformation from must to wine as it was conducted at the end of the nineteenth century:

During the 15 or 20 days of the primary fermentation known as "turbulence" [3] it is necessary to carefully top up the barrel three or four times with good wine. The bung hole is covered with a vine leaf weighed down with sand, which enables the carbon dioxide generated by the fermentation to escape. After about twenty days, when one is sure that this stage has finished, the barrel is filled up. The wine stays in these barrels until Christmas and continues to throw off a deposit; fermentation eventually stops due to the cold and the wine clears. It is then drawn off towards the end of December, during a spell of clear dry weather, into barrels of the same size where it finishes clarifying.

The wine is thus good to be drawn off when, due to the cold, most of the salts and organic matter has precipitated. For a long time wine makers were at the mercy of the fluctuations in winter temperatures, and it was not until about 1890 that the first refrigeration systems began to be installed in cellars.

Before drawing off the wine was fined with tannin, tartaric acid and isinglass. Maumené states that in the 1870s egg white was also used for fining, as was casein and dried blood powder. Drawing off was, from the start of the century, carried out with a pump, of which the use was instigated in Champagne and then successively in other wine producing areas. Drawing off was an opportunity to adjust the colour if necessary using animal charcoal. In the second part of the nineteenth century wines began to be filtered, but fining was still preferred.


The stage that follows drawing off was that of the blending of the wines, an operation that consisted of mixing together the various types of wine that made up the ’cuvées’ (blends); these were made from wines of the same year that were made from various "crus" (different types of grapes from various locations), that were then skilfully blended, following the rules set down in the eighteenth century by Dom Pérignon for vin gris and by the first producers of sparkling wine. Cavoleau described the blending process in 1822 as the making of assortimens (arrangements). In the Notice Historique sur le Vin de Champagne, the official document of 1889, it consisted of recoupages ou assemblages.

The word assemblage (literally "assembly") was used comparatively late in the context of blending, it also referred to the arrangement in the cellar of the barrels in parallel lines, each line containing a certain number of barrels of each ’cru’, reflecting the proportion that was used when the various blends were made up.

At the end of the nineteenth century cuvée was defined as the combination of wines of which the master of the house has personally, after tasting, decided the composition and quantities, with a view to creating a homogeneous and harmonious blend, in which the bouquets are combined, improved and complementary to each other. In fact blending was a necessity for the wine producer in Champagne, as it had been for Dom Pérignon, and for similar reasons. As Robinet, an inhabitant of Epernay, wrote, in a wine producing region like Champagne, where property is divided ad infinitum, each vineyard owner only harvests a small quantity of wine; merchants are therefore... obliged to blend wines coming from a large number of properties. Moreover, for producers this practice represented the best use of the available wines. André Jullien gives a good example, When it has been hot the best wines of the 5th class are used for the preparation of the sparkling wines of the 3rd class, adding to them about 1/10th of the wines of the second or third pressing from Ay or Mareuil which, by bringing extra body and alcohol, render them more likely to keep. It has indeed been proven, and not only in Champagne, that blending enhances overall quality. It is practised in the Bordeaux region and numerous other vineyard areas. Lenoir affirmed in 1828, writing about wine in general, that blends, when they are well matched and made in the right proportions, always produce better wines than any of the individual constituents. And Horace had already written, I have neither the vines of Falernum, nor the slopes of Formia, to correct with a happy blend the wine from my own grapes.

Champagne was, and still is, essentially a brand name wine and blending also gave producers the opportunity to remain faithful to a style adopted by their house and known to its customers. By making use of the various wines stored in its cellars the same style could be recreated every year, or even, if it was judged appropriate, changed over a few years in a series of imperceptible transitions. Champagne was sometimes specially blended for foreign customers, whose tastes were often different to those of the French market. As Cavoleau wrote in 1827, to send a wine to Frankfurt that had pleased Paris would be to risk seeing it remain on the sender’s account. The idea was developed by J. de Saint-André in 1892, Customers in one country will like their champagne young, elegant and sweet, while those of another want champagne that is old, full-bodied and dry, it is obvious that the elements which make up these wines will not be the same.

Any improvement in the quality of the wine was obviously commercially desirable, but on a purely technical level producers used blending to finalize the adjustment of the colour, and above all to improve the chance of obtaining a genuinely sparkling wine. If we are to believe Maizière on this point the best wines were not always the ones that succeeded, for he wrote in 1846, It is still only by chance that one can have a great sparkling wine made from the best quality grapes. As in the eighteenth century the accepted blending practice was to mix wines made from white grapes with those made from black grapes in order to maximize the possible sparkling effect. The rule was to blend, before bottling, wines of various crus with complementary characteristics, and on the basis of what was available, but always in such a way as to obtain the best wine possible, by uniting the various qualities attributed to each vineyard. But once assembled, or blended, the crus would completely lose their individuality; and even if one was to predominate the customer would not know which one of them it was. And there were some who regretted this; for example in a Bibliophiles Note inserted in Louis-Perrier’s Mémoire sur le Vin de Champagne we find, Is this not to remove from our best wines their true character? By this general fusion of the grapes from a dozen parishes in the same vat we lose the wine of Ay’s taste of peaches, the wine of Avenay’s taste of strawberries, Hautvillers’ taste of hazelnuts, and Pierry’s taste of gunflint.

Blending was not limited to the various crus of the year in question, and could include wines from previous years dating back to the middle of the century. Dr Guyot called this operation recoulage ("repouring") and wrote on this subject, When the drawn wine is the product of an unremarkable or mediocre year, some wine from a particularly good year is added, that has been kept aside specifically for this purpose. For example the unexceptional wine of 1847 was blended at 10, 15, and 20% with the great wine of 1846. He notes that the consistent quality of the great houses depends on this practice and that a house without a stock of old wines of the best quality would, in the event of a series of mediocre years, be totally lost, with no hope of maintaining its reputation. These stocks made up what the Syndicat du Commerce des Vins de Champagne (which became the Syndicat de Grandes Marques, and then in 1994, the Union des Maisons de Champagne) called in 1889 the vins de réserve and which have been referred to as such ever since.

If only the wines of a single year were blended then one had what is today called a millésime (vintage champagne). This word did not exist in the nineteenth century55 and the concept was far from well-defined. From the early 1830s the date began to be mentioned on some labels, but rarely, and only to indicate a truly exceptional year, particularly for the British and American markets. However at around the same time wine merchants and connoisseurs, especially in England, would often refer to the year of production even when it was not indicated. It was only in 1865 that the practice began of regularly putting on the market, every two three years in France and more frequently in England, champagnes that had in principal been made without the addition of vins de réserve; and yet the vintage was still only occasionally shown on the label, until at least the 1870s. The only guarantee that they were sold with was simple good faith in the seller, writes André Simon. He states that every year from 1881 to 1891 was sold as a vintage in England while on the continent non-vintage was the rule. And as there was no legislation that obliged the year displayed to correspond to the year of production of the wine in the bottle it was without shame that a champagne "of 1882" was sold that was in fact a blend of two or three different years. Feuerheerd even went as far as to write in 1899 that he doubted that a wine of Champagne consisting purely of a single year had ever been put in bottles. Good houses nevertheless made it a point of honour to offer a champagne that essentially consisted of the wine of the year displayed, while others added only a token quantity or none at all.

From the 1870s the best brands made champagnes for which the wine of the year displayed really was the principal ingredient, and the custom of putting the year on the cork began to establish itself. Some producers did not go along with this until fairly late, such as Charles Heidsieck (1889), and Pommery (1892).

Vizetelly considered that great vintages never occurred more than twice in ten years, and that during the same period there would generally be one or two other fairly good vintages (651). Here, according to his research and observations, and on the basis of other subsequent accounts, are the very good years of the nineteenth century: 1802, 1806, 1811, 1815, 1818, 1822, 1825, 1834, 1840, 1842, 1846, 1848, 1857, 1865, 1868, 1874, 1880, 1884, 1889, 1892, 1893, 1898, and 1899. It should be added that 1875, the year that gave the biggest harvest of the century, was also praised for its quality, and that 1900 offered a very promising start to the twentieth century.

One year, 1811, or the Year of the Comet, deserves special mention. At the start of the century nature had not always been generous, and there had been much complaining in 1805, when the detestable wine, baptised "the conscript", spoilt the barrels, and also in 1808 and 1809. But then came 1811 which produced a wine that was incomparable according to all reports, both in quantity and quality. Now, it so happened that a particularly impressive comet had appeared that year. By a natural association of these two marvels the wine was baptised the Wine of the Comet. 1846 was also a particularly impressive year, as was 1874, which left such an impression that twenty years later R.J. Lloyd Price’s farewell Ode to Pommery 1874 appeared in the Vanity Fair of 27 December 1894, ending with the tragic lines, Good-bye then, Pommery Seventy-Four! With respectful sips we leave you, saddened to think that never again shall such wine pass our lips.

The various operations involved in the preparation of the blend were carried out during January and February (234). By the end of the century these had evolved into an unchanging rite that it is even today still essentially the same .

The first stage was to carry out an analysis of the wines, to examine their qualities and faults, their colour, acidity, and alcohol content, and to compare them with the characteristics of previous wines as noted in the livre des cuvées (book of blends). This would enable one to form an idea of the possible and desirable blends. Next was the dégustation (tasting) which consisted of a number of stages carried out by several practitioners, under the guidance of the cellar master, but the overall responsibility fell to the head of the house who would personally participate in the final stage and sometimes in all of the stages, either alone or in the company of his or her assistants. For the heads of houses this was one of their most important prerogatives, and the final blend was regarded as being of their own composition. Here is how this operation was described by Salleron in 1886, In the morning the tasters would assemble in the inner sanctum of the tasting office, a veritable oenological temple, where, in silence, they would eat a pomme d’api (a type of small apple). The wines would be tasted and classified: some would be put down for the first class blend, others for the second or third, and some would be saved. The tastes of the various countries where the wines were to be sold was a primary consideration; blends of previous years served as rallying points; everything would be examined, weighed, calculated and the final figures decided on: the blend would be made up of so many barrels of Ay, Bouzy etc.

The next operation was the blending of the selected wines. This was done by arranging the barrels and pouring their contents in the proportions indicated into large ’foudres’, immense recipients that could hold 200 and sometimes even 250 barrels, in order to obtain a consistent wine, with exactly the same strength, quality and character. A mixing mechanism with paddles gently stirs the huge volume of liquid to ensure perfect homogeneity. When the blending is completed it is transferred into new barrels, fined with extra care 58 and then left for a while to rest. Finally they are drawn off and taken down to the cellar until mid April, which is the usual time for bottling


We now know why, before drawing off, it is necessary to add some sugar to the blend and the yeasts which will covert it into alcohol and carbon dioxide. It was only in the 1830s that people started to understand the reasons behind this technique. Previously they had operated on an empirical basis because the role of these elements in alcoholic fermentation was still unknown, despite their presence being essential in order for it to take place. The result was a great deal of irregularity in the sparkling nature of wines, left as they were to become sparkling naturally in what were hoped to be suitable conditions.

Here is what Cavoleau wrote in 1827: Wines bottled to be sparkling do not all form bubbles in the same way. There are some in which the bubbles can be seen almost immediately after two weeks in the bottle; others take several months; some require a change of temperature and these are brought back up from the cellar; some wait for the increase in the sap in August; others start just when everyone has given up waiting for them; and finally there are those which must be poured back into barrels the following year and mixed with some wine from the new harvest, which has the characteristic of being particularly sparkling, such as that made from the white grapes of the Côte d’Avize. The sparkling nature of the wine of Champagne, its initiation and its various other aspects, might be compared to... an elusive deity, pursued by the most experienced merchants and vineyard owners. As late as 1846 Maizière was asking when the art of obtaining, not by chance and without loss, a great sparkling wine from a pure blend of the best grapes, is going to be understood.

Cavoleau might also have mentioned the casse (breakages), which was as rife in his time as it had been in the eighteenth century. Usually one would expect to lose 10 to 20% to breakages, but it could happen that an entire blend would be destroyed. Jullien warned that it is not wise to cross a cellar without the protection of a wire mesh mask.

Attempts to protect against this curse sometimes involved some strange inventions. Maizière put into service, but with little success, the para-casse, a huge airtight cylinder which was supposed to enable the sparkling process to take place in 15,000 bottles at a time; and in the Report to the Academy of Rheims by the Commission charged with the examination of various processes relating to wine making, Sutaine describes unsuccessful attempts at acupuncture by passing a hollow needle through the cork in order to let some carbon dioxide gas escape and thereby prevent breakages, the trials were carried out on a small part of 2,000 bottles, using Doctor Rousseau of Epernay’s trocar. In the same report we learn of the best means of preventing the disaster of excessive breakages. It is suggested that the bottles are uncorked and emptied of about one fifth of their contents, and then refilled, this operation having the inconvenience of taking a fairly long time and of incurring considerable costs or yet another suggestion is to imitate those who pour their wine back into barrels and then rebottle it immediately.

But as he finishes Sutaine mentions the only possible hope of solving the problem by recommending a small work, which is already well-known, to all the wine merchants dealing in the wine of Champagne, by Mr. François, a former pharmacist in Châlons-sur-Marne, and which is entitled: Treatise on the Preparation of Sparkling White Wines. Indeed it was this thin brochure that appeared in 1837, the work of a modest researcher, which was the initial and vital link in the series of trials and studies that would lead to a proper understanding of the phenomenon of how wine becomes sparkling and ultimately, as a result, the practical applications of this understanding, how to control the sparkling process and virtually eliminate the risk of breakages. At the start of the century everyone was unaware of the role of sugar in secondary fermentation [4]. It was only observed that sparkling wines gain their sparkling property from being enclosed in glass before they have completed their fermentation. Sugar was sometimes added, but usually with the idea that it would help the wine to keep. Cadet de Vaux wrote in 1803 that wines never turn to vinegar so long as there still exists a portion of sugar and that this observation justifies the practice, above all for the sparkling wine of Champagne, of adding a little sugar at the bottling stage to ensure that it keeps without changing.

It would seem to have been a little before 1820 that some wine makers began to add sugar with the hope that it would improve the sparkling process, as Roques remarked in 1821, To render champagne more sparkling some sugar candy dissolved in white wine may be added (550). If this was done earlier then it was not mentioned: Chaptal, Jullien, Macculoch, Cavoleau, and Lenoir are all silent on the matter. Towards 1830 the addition of sugar to the blend was acknowledged as favouring the sparkling process but there was still no technique for establishing the optimum amount, which would have to take into account the quantity of residual sugar from the initial fermentation, a quantity that no one knew how to measure. It was from this deadlock that François struggled to find a way out. Having observed that wines lack sparkling properties when there is not a sufficient quantity of sugar, and that there are enormous numbers of breakages when there is too much of this substance he set himself the task of establishing the rational use of sugar.

After numerous experiments he developed a technique which involved first the evaporation of the alcoholic content of a given volume of wine to be bottled, and then the evaluation, using Cadet de Vaux’s gleuco-oenometer and a table of correspondence, the weight of sugar found naturally in a litre of the wine being analysed. This became known as the François reduction, which Monceau said was practical if not very precise. The result obtained was however only part of the story, because while one now knew how to measure the residual sugar, one still did not know with any precision the total quantity necessary for a successful sparkling process and thus the correct amount to add still remained unknown.

Due to François major errors could now be avoided, but it was still very much a question of broad assessments that were backed up by experience and confirmed by tastings. This did however enable the proportion of breakages to gradually fall, albeit with some spectacular exceptions, to somewhere between 3 and 8%. François moreover considered such a percentage to be normal because, he wrote, if one goes down to 1 or 2% of breakages then one runs the risk of fermentation stopping, which would then result in a sparkling effect that lacked the appropriate intensity.

The pharmacist of Châlons must be given great credit for this observation because it draws attention to the relationship that exists between the weight of sugar contained in the wine and the production of carbon dioxide. On the practical side he invented a very useful dosing process that enabled a significant reduction in breakages from the 1840s onwards. The 1899 Bulletin of the house of Moët & Chandon, states that the François reduction still serves as the cellar master’s guide in the bottling operation, and remains the same today as when it was created by its author. In the Notice Historique sur le Vin de Champagne, which it prepared for the 1899 Exhibition, the Syndicat du Commerce des Vins de Champagne (which became the Syndicat de Grandes Marques, and then in 1994, Union des Maisons de Champagne) notes that since Mr François’ important discovery, the trade in sparkling wines has considerably increased.

François died prematurely in 1838 at the age of 4660, but studies of the sparkling process continued. In 1874 Maumené, a former professor of chemistry and holder of the municipal chair of Rheims, focused his research on the ability of wine to absorb carbon dioxide, a process that he had first noticed in 1858. He drew up a table that showed the pressure that would result as a function of the amount of sugar that was added. Unfortunately his method for establishing the absorption capacity was too imprecise to be applied with a great deal of success. In 1877 Robinet came up with a reduction method that was faster than that of François, but less precise, and in 1886, Salleron invented an absorptiomètre which involved a manometer and a pump for calculating the absorption capacity of a wine, and advised calculating the residual sugar by chemical analysis based on the reduction effect of sugar on salts of copper. Research and inventions continued but it was not until the last years of the nineteenth century that, having established that about four grams of sugar were necessary to create a pressure of one atmosphere, it finally became possible to calculate, given the quantity of residual sugar (calculated by various means), the absorption capacity of the wine, its alcohol content and the temperature, the exact quantity of sugar to add in order to obtain the six atmospheres of an intensely sparkling wine.

Parallèlement aux recherches sur l’emploi du sucre, on s’occupe des ferments [5], dont Chaptal a reconnu l’importance au début du siècle :

Le sucre existe dans le raisin et c’est surtout à lui qu’est dû l’alcool qui résulte de sa décomposition ; mais ce sucre est constamment mêlé avec un corps doux plus ou moins abondant, et qui sert de ferment ; nous appellerons ce principe doux levain, levure [6], ce que Lenoir confirme en 1828 en écrivant qu’aucune fermentation alcoolique ne peut avoir lieu sans la présence d’un ferment [7] qui en est l’agent en transformant le sucre en alcool et en gaz carbonique.
Il faut revenir au pharmacien François, qui procède à une étude approfondie des ferments et indique les moyens de les éliminer lorsqu’ils sont en excès, en les précipitant par l’acide sulfureux, l’alcool (ou l’eau-de-vie) et le tanin. Il note également que la quantité de ferment est en raison directe de la présence de l’acide tartrique, qui est son dissolvant naturel [8]. Mais il n’établit pas de relation directe entre les ferments et la prise de mousse. On y songe cependant. Sutaine écrit dans le rapport précité de 1843 : Nous connaissons la manière de peser le vin et le degré de sucre qu’il doit contenir : que la science nous apprenne maintenant à opérer de même sur le ferment.

La connaissance de la nature des ferments et de leur rôle dans la fermentation alcoolique fait de grands progrès dans la deuxième moitié du XIXe siècle.
Cagniard de La Tour, reprenant les travaux de Leuwenhoek démontre que ce sont des végétaux. Pasteur établit d’une manière indiscutable la relation systématique qui existe entre la présence de levures et la fermentation du milieu. Buchner découvre que les ferments agissent par l’intermédiaire de substances qu’ils sécrètent, les enzymes. Hansen met au point un procédé permettant d’obtenir des cultures pures à partir d’une seule cellule de levure. Maumené découvre que les aléas du tirage sont la conséquence de l’existence et des fonctions de la levure et, le premier, envisage d’utiliser les ferments pour la prise de mousse. Il note que leur action est différente selon les crus dont ils proviennent et qu’il n’est pas nécessaire d’en avoir une grande quantité, la qualité primant la quantité. Salleron les étudie dans leurs rapports avec le sucre, l’alcool et la température, il en ajoute expérimentalement dans les vins de tirage, avec d’excellents résultats et il préconise de ne pas trop pousser le soutirage après assemblage, afin d’en conserver en quantité suffisante dans le vin [9] . À partir des années 1880, ayant acquis une meilleure connaissance du rôle des levures, les producteurs les utilisent petit à petit pour la seconde fermentation, avant de le faire systématiquement au début du XXe siècle. On sélectionne les ferments par culture dans les laboratoires régionaux car on s’est aperçu que les levures cultivées de crus de Champagne accomplissent le travail de la prise de mousse beaucoup plus facilement, plus régulièrement et mieux que n’importe quel autre ferment et ont la faculté de faciliter la formation du dépôt dans les bouteilles et de simplifier beaucoup le travail du vin. Avec elles, on a l’immense avantage d’être absolument certain de réussir la prise de mousse [10]

Mais hélas ! on n’en est pas encore là au milieu du siècle, et même dans les années 1880 malgré les progrès que l’on vient de voir. Il n’y a pas encore de résultat définitif dans la recherche du contrôle de la prise de mousse. La casse continue, souvent d’ailleurs par suite d’une mauvaise application des travaux des chercheurs, et la production du vinaigre encore largement alimentée par le vin de casse.

Il n’est que de lire Vizetelly qui, en 1879, raconte comme suit sa visite dans une cave champenoise : Les bruits causés ici et là par les bouteilles qui explosent assaillent l’oreille, et tandis que l’écho va en mourant, il s’y mêle le son argentin du vin qui s’échappe, coulant en cascade jusqu’au bas des tas et trouvant son chemin sur les côtés inclinés du sol jusqu’à l’étroite gouttière centrale [11]. En 1894 encore, Thudichum écrit que lorsque la casse atteint plus de 8%, le vin doit être débouché ou transporté dans un endroit plus frais, comme on l’a vu faire au début du siècle, et que pendant ce travail le personnel porte des masques et gants [12].
Une casse de 8% est plus supportable que les 20% des années 1840 mais elle est encore considérable. D’où différentes pratiques, déjà connues, pour essayer de sauver la cuvée, surtout par le refroidissement. Maumené le conseille sous forme d’aération, de jets d’eau froide sur les bouteilles, de blocs de glace entreposés dans les caveaux, en précisant qu’il faut nettoyer rapidement les caves souillées par les vins qui se sont répandus, sans quoi ils s’oxyderaient et feraient monter la température de la cave à 18 ou 20° . On opère aussi des changements d’emplacement en descendant les bouteilles dans une cave plus froide, lorsqu’une pression de 8 atmosphères est indiquée par l’aphromètre que Maumené a inventé pour pouvoir suivre la progression de la fermentation.

Parmi les barrières que l’on cherche à opposer au fléau destructeur de la casse, on doit noter, comme au XVIIIe siècle, la recherche de meilleures bouteilles. Au milieu du siècle la champenoise, ainsi nommée depuis 1800 [13], est encore assez pansue, puis elle s’allonge, ce qui lui donne une meilleure résistance. Pendant longtemps elle est faite très irrégulièrement. On lit dans l’édition de 1866 de la Topographie de tous les vignobles connus de Jullien que la qualité des matières servant à la fabrication des bouteilles, et peut-être aussi le degré de feu qu’elles ont subi, contribuent à diminuer ou à conserver la mousse des vins de Champagne. On ne doit donc pas être étonné de trouver, dans le même panier, des bouteilles dont le vin mousse plus ou moins fortement, et d’autres où il ne mousse pas du tout.
Pour ne pas s’exposer à une casse qui résulterait de l’emploi de bouteilles déjà fatiguées par la pression d’une première prise de mousse, on n’utilise généralement que des bouteilles neuves [14]. Et comme les bouteilles sont d’une solidité inégale, on cherche un moyen d’en contrôler la résistance avant emploi.

Sutaine écrit en 1843 dans son rapport précité qu’il deviendra indispensable que toutes les bouteilles destinées aux vins mousseux soient éprouvées dans les verreries, et que celles-là seulement qui auraient résisté à la pression d’un certain nombre d’atmosphères soient livrées à la consommation. En fait, c’est au niveau du producteur que se fait l’essai préalable des bouteilles. Il dispose pour cela depuis les années 1830 de la machine à pression Colardeau, à laquelle le docteur Rousseau, médecin sparnacien et très actif chercheur, déjà rencontré à propos de l’acupuncture, apporte quelques perfectionnements avec un appareil de son invention, le brise-bouteilles. Ces procédés sont fort imprécis car, comme le fera remarquer Salleron, qui avait pour sa part inventé un élasticimètre, des bouteilles résistant à 30 atmosphères pendant deux ou trois minutes peuvent ne pas supporter des pressions continues de 8 atmosphères. Maizière, autre chercheur infatigable, préconise de vérifier la solidité de la bouteille en en mesurant l’épaisseur à l’épaule, la partie la plus mince, avec un compas de son invention. Mais les résultats sont décevants.

À la fin du XIXe siècle on n’a donc encore trouvé aucun remède totalement efficace contre la casse, qui ne s’abaissera dans des proportions admissibles qu’au début du XXe, lorsque l’on aura acquis définitivement la maîtrise de la deuxième fermentation.

On a bien compris combien est irrégulière et déconcertante, à l’époque, la production du gaz carbonique pendant la prise de mousse. La pression monte plus ou moins rapidement, avec des pointes de 8 à 12 atmosphères pouvant atteindre exceptionnellement 20 à 30 atmosphères. Elle s’établit en fin d’opération à 2 atmosphères environ en 1840, d’après le rapport de Sutaine, et à 4, 5 ou 6 en 1870 d’après Maumené , ce qui correspond déjà à la pression du champagne d’aujourd’hui. Deux atmosphères, par contre, c’est peu. Cela suffit pour faire sauter le bouchon, mais la mousse ne persiste pas longtemps dans le verre. On doit à Maizière une jolie évocation du comportement du champagne sortant de sa bouteille où le lyrisme de l’homme de goût rejoint la précision du technicien; elle se termine par la description de l’ascension accélérée d’une colonne centrale de perles brillantes, qui naissent au fond du verre, et se succèdent rapidement pendant une minute (379), ce qui n’est guère !
Il est aisé, cependant, de prendre conscience de l’augmentation de pression constatée au cours du XIXe siècle en se reportant à la littérature et aux dessins de l’époque. Pour faire mieux mousser, comme le faisaient leurs prédécesseurs du XVIIIe siècle, les buveurs que décrivent les hommes de lettres de l’Empire et de la Restauration, que dessinent Gavarni, Deveria, Grandville, versent de très haut le champagne dans les verres, ce que facilite l’usage de le servir très froid mais ce qui serait impossible avec une forte pression. À la fin du siècle, cette habitude s’est perdue car elle est devenue impraticable en raison de l’abondance habituelle de la mousse.

Il peut arriver cependant, et c’est assez fréquent, que les vins pèchent par une mousse trop faible, la pression restant très au-dessous de ce qu’elle devrait être normalement, atteignant seulement une atmosphère au lieu de deux, ou deux au lieu de quatre à cinq. Ces vins de petite mousse, résultat d’un échec [15], ces vins appelés au XVIIIe siècle les demi-mousseux, on en fait depuis les années 1820 type particulier de champagne, le crémant [16].

Voici ce qu’en dit en 1822 André Jullien : Chassant le bouchon avec moins de force, pétillant moins dans le verre, leur mousse forme une nappe d’écume qui couvre la liqueur et se dissipe au bout de quelques instants; ils ont, sur les vins grands mousseux, l’avantage de conserver plus de qualités vineuses et d’être moins piquants; leur prix est ordinairement plus élevé parce qu’ils sont fort recherchés par un certain nombre d’amateurs, et que, ne devant leur qualité qu’à l’un de ces phénomènes bizarres qui se manifestent dans les vins de Champagne, on n’est pas à portée de s’en procurer en aussi grande quantité qu’on eut le désirer [17].

En 1862, Moët & Chandon adresse à ses clients un prix courant qui précise que le vin Crémant d’Ay ne peut être obtenu que dans les meilleurs crûs et dans les années tout-à fait remarquables; il est offert en blanc et en rosé et vaut 25 % de plus que les autres qualités.

Mais le grand mousseux est celui qui constitue l’essentiel des expéditions, celui qui est consommé universellement, celui qui s’élève en gerbes jusqu’au plafond [18], pour la plus grande joie des caricaturistes à qui il offre matière toutes sortes de situations cocasses. Maumené indique en 1874 que sa mousse est marchande, forte ou très forte, selon que la pression atteint, en atmosphères, respectivement à 4½, 4½ à 5½, 5½ à 6 , la mousse forte étant celle que cherche à obtenir systématiquement Salleron en 1886. Le degré d’alcool du grand-mousseux terminé est de l’ordre de 11,5° à 12,5°.

La tisane de Champagne existe toujours. Par sa pression, elle s’apparente au crémant, mais elle a moins de finesse. Le Dictionnaire de la langue française de Littré la définit comme un vin de Champagne plus doux, plus sucré, moins spiritueux et André Jullien précise que les vins de tisane sont principalement fournis par les crus de la Côte d’Avize [19].


Sachant ainsi ce que l’on veut ou espère obtenir comme mousse, on se prépare au tirage des vins qui ont été préalablement assemblés. L’époque choisie est, comme au XVIIIe siècle et pour les mêmes raisons, celle du retour des beaux jours, en principe mai et juin. On est toujours lié par le concept de la reprise de la fermentation interrompue, si bien qu’il serait plus exact de dire qu’à l’époque le champagne résulte d’une fermentation unique, mais discontinue, bien plutôt que d’une seconde fermentation. Comme le faisaient les auteurs du siècle précédent, John Macculoch l’exprime en 1821 en écrivant que le retour des températures élevées explique le renouveau de la fermentation qui a lieu au printemps, après qu’elle a été partiellement ou totalement suspendue par le froid de l’hiver [20] . Nous voici au printemps, écrit Salleron, la nature entière revient à la vie, tous les êtres subissent l’action mystérieuse qui doit assurer la perpétuité des espèces par la génération... Les globules de ferment sortent de leur torpeur et attendent le sucre nécessaire à leur nouvelle évolution. C’est le moment opportun pour procéder au tirage. On s’affranchit peu à peu des anciens impératifs du cycle lunaire, mais on reste très attentif à la température, et à bon droit car les ferments ne reprendront vie que si elle leur est propice. C’est ce que François exprime en écrivant : La mise en bouteille ne doit avoir lieu qu’autant qu’une température de douze degrés ait régné depuis quelque temps, aussi on ne doit pas s’attacher précisément à tel mois ou à tel lune, si cette condition n’existe pas.
Pour le docteur Guyot, le sucre restant naturellement dans le vin est le combustible de base de la prise de mousse, et on s’efforce de faire en sorte qu’il se trouve exactement limité à la proportion nécessaire pour qu’on obtienne la bonne mousse . C’est la doctrine des œnologues de l’époque mais elle n’est pas admise par tous puisque l’on peut lire dans le Vigneron champenois d’avril 1883 que dans la vinification la première chose à obtenir est la fermentation entière du sucre, théorie en avance sur son temps. En réalité, on cherche généralement à limiter la fermentation. Dans le Nouveau dictionnaire des dictionnaires, en 1898, Mgr Guérin, reprenant un texte du docteur Guyot, écrit que l’on met le vin à l’abri d’une première fermentation trop complète, afin qu’il garde une notable quantité du sucre du moût. Vizetelly explique très précisément que si c’est nécessaire on ralentit la fermentation en transférant les tonneaux dans un cellier plus froid, car il est essentiel que le vin retienne une large proportion de son sucre naturel pour assurer sa future effervescence.
On prétend, sans preuves, dans les notices des maisons que la faculté qu’ont les vins blancs de garder leur sucre naturel est loin d’être aussi prononcée dans tous les vins que dans les vins de Champagne [21], que différant en cela de la plupart des vins ordinaires, le vin de Champagne ne perd pas tout son sucre dans la première fermentation [22]. En réalité, c’est un problème de conduite de la première fermentation, et du fait que ces documents de la fin du siècle ont pour objet la promotion des marques, que l’on n’y trouve pas mention d’addition de sucre, on peut seulement conclure que l’on s’efforce à l’époque de présenter le champagne comme un produit absolument naturel.
Dans la pratique, on prépare une liqueur de titrage [23] en faisant dissoudre dans du vin de cuvée la quantité de sucre nécessaire, sucre candi de canne, provenant des colonies, en général raffiné à Nantes, et on y ajoute les ferments lorsqu’ils sont en usage. Afin d’éviter la maladie de la graisse, on y joint du tanin à la dose de 15 à 20 gr/hecto . En outre, si l’acidité est insuffisante, on la complète par une addition proportionnée d’acide tartrique ou mieux d’acide citrique, et si le degré alcoolique est trop bas, on le remonte par une addition de fine champagne.
C’est à ce stade qu’est préparé à l’époque le champagne rosé, le vin paillé ou œil-de-perdrix du XVIIIe siècle, que l’on fait mousser de la même façon que le blanc, et cela depuis le début du XIXe puisqu’il figure en tant que vin rosé mousseux sur un bulletin d’expédition de 1775 de la maison Vve Clicquot et sur la carte des Frères Provençaux de 1814 [24] , œil-de-perdrix mousseux sur le prix courant de 1819 de Wichelhausen à Zurich, champagne pink sur celui de 1823 de The London Wine Company.

Pour obtenir la teinte voulue, on ajoute dans la liqueur de titrage une quantité de teinte, dite de Fismes; c’est le seul procédé cité par Robinet, qui ne l’indique d’ailleurs qu’en passant. Mais on peut aussi, avant tirage, ajouter du vin rouge à la cuvée, méthode que préfère Vizetelly qui estime que l’utilisation de la teinte de Fismes est le fait de producteurs de réputation douteuse.

Il est curieux de constater que le champagne rosé est toujours marginal. On en parle depuis le début du siècle, mais beaucoup de traités œnologiques, tel celui de Maumené, n’y font même pas allusion. Quant aux amateurs, ils s’en détournent parfois, comme Alexander Henderson, qui écrit en 1824 que le champagne rosé est moins demandé et qu’il n’a en réalité rien qui justifie qu’on lui donne la préférence [25], ou Cyrus Redding qui affirme en 1833 qu’aucun connaisseur ne voudrait en boire s’il pouvait obtenir les autres catégories [26]. Le champagne rosé est en réalité un vin au succès intermittent, réapparaissant de temps à autre avec une popularité passagère due à son aspect attractif, mais qui ne séduit jamais les véritables amateurs de champagne.
À côté du champagne rosé, il existe le champagne normal légèrement rose, produit dans les années où la maturité du raisin est excessive. Selon Vizetelly [27], ce champagne slightly pink est devenu à la mode en Angleterre depuis le vin de 1874, mais le fait n’est pas nouveau. Sur un avertissement joint vers 1840 aux expéditions de la maison Moët, on lit que les meilleurs vins blancs de Champagne se faisant avec le raisin noir, une plus grande maturité, dans les années chaudes, leur donne une légère nuance de rose, qui, bien loin de leur nuire, est une preuve d’excellente qualité.

Quant au champagne rouge, il s’en est fait un peu (on peut citer notamment un champagne rouge Giesler 1887) mais on n’en a guère parlé, son succès, si succès il y a eu, n’ayant pu être que très limité.

La liqueur de titrage et le vin de cuvée sont déversés ensemble dans de grands foudres, d’une capacité qui atteint couramment 600 à 750 hl à la fin du siècle. Des palettes mécaniques y brassent le mélange afin d’en assurer l’homogénéité et d’aérer le vin qui doit contenir l’oxygène nécessaire à la vie des ferments. Selon les conseils de François, on remet souvent en tonneaux pour attendre la reprise de la fermentation et procéder alors seulement au tirage. Cette méthode est encore préconisée par Salleron en 1886 lorsqu’il écrit : L’habitude est de ne mettre en bouteilles que lorsque le vin a subi un commencement de fermentation, et on s’en trouve bien.

Les bouteilles utilisées à la fin du XIXe siècle sont indiquées ci-après avec leur contenance: le double magnum, 2,90 l à 3 l, le magnum, 1,45 l à 1,50 l, la champenoise, 0,78 l à 0,84 l, la pinte impériale [28], 0,56 l à 0,60 l, la ½ champenoise 68 , 0,39 l à 0,42 l, la ¼ champenoise, 0,18 l à 0,20 l. La champenoise a un poids d’environ 1 kg. Elle est habituellement remplie à 0,80 l.

Après avoir été triées, et éventuellement testées, les bouteilles sont soigneusement nettoyées.
On les décape avec de la grenaille de plomb, qui a le défaut de rester souvent dans les bouteilles, et des chaînes de fer portant à chaque extrémité une sorte de molette d’éperon, ou avec des perles de verre, ou encore avec des brosses et du sable (au début du XXe siècle, on utilisera de la limaille de fer). On les lave à l’aide de canules ou de machines à laver. Ce sont d’ordinaire les femmes qui font ce travail, les laveuses, et qui le contrôlent, les mireuses.

Les bouchons sont en liège, de Catalogne et d’Andalousie pour les meilleurs. Ils sont d’un seul bloc, ou, à partir de 1858, faits de deux parties, collées dans la longueur ou la largeur par de la gutta-percha; on en trouve même, vers 1870, qui sont faits de six fragments. Pour les fabriquer, de nombreux Catalans viennent s’installer en Champagne au cours du XIXe siècle.
Certains sont employés dans les maisons de champagne au service des bouchons, d’autres travaillent dans les bouchonneries installées sur place par des Espagnols.
Les bouchons sont examinés un par un avant usage, parfois à l’aide d’une machine à essayer les bouchons, inventée par Salleron, mais le plus souvent à la main, afin de déceler les imperfections qui pourraient occasionner le goût de bouchon, qui tient de la moisissure et d’une amertume particulière. Un bouchon défectueux peut être responsable de la mauvaise obturation d’une bouteille qui deviendra une couleuse, une recouleuse [29], noms que l’on donne à une bouteille mal bouchée ou munie d’un bouchon de mauvaise qualité qui laisse échapper du vin et du gaz [30]
Dans la salle des tirages travaillent jusqu’à 200 personnes, hommes essentiellement, occupées au remplissage des bouteilles, dont sont chargés les emplisseurs, au bouchage et au ficelage. Des enfants font passer les bouteilles des uns aux autres. À partir de 1825, on utilise dans les grandes maisons des tireuses à six, huit ou dix becs, d’origine anglaise.

Pour le boucheur, faire pénétrer un bouchon de liège dans un goulot de bouteille d’un diamètre deux fois plus petit n’est pas une opération facile.

Sutaine, dans le rapport précité, écrit en 1843 : Le temps n’est pas encore bien loin de nous où les tonneliers ne connaissaient, pour déprimer le bouchon et le forcer d’entrer dans la bouteille, d’autre moyen que de le serrer entre leurs dents, moyen nuisible à la santé de quelques-uns, et, dans tous les cas, fort peu énergique et peu convenable.

L’ouvrier avait alors comme seul outil d’enfoncement un maillet appelé batte, dont il donnait trois coups sur chaque bouchon, à la cadence de 1700 bouteilles par jour [31]. En 1827, apparaît en Champagne une machine à boucher fabriquée en Bourgogne restée imparfaite malgré des améliorations successives et remplacées par les machines à maillet Leroy (vers 1840) et Maurice (1848) qui donnent enfin satisfaction. Dans les années 1860, M. Charbonnier, d’Épernay, fabrique pour le bouchage une machine à mouton glissant entre deux rainures verticales, dont l’idée lui aurait été donnée, si on en croit le Vigneron champenois du 29 mars 1899, par un de ses apprentis qui avait vu tomber le couperet de la guillotine, alors que celle-ci était dressée le 23 juin 1862 pour une exécution capitale sur la place Louis-Philippe, aujourd’hui place de la République, à Épernay. Il est de fait que cette machine à boucher est connue sous le nom de guillotine. Elle est améliorée par Delorme qui combine le maillet et le mouton.
Parallèlement aux recherches en matière de machines à boucher, dont le rendement s’élève ainsi à 5000 bouteilles par jour, on met au point des machines à ramollir les bouchons à sec, puis on adopte la méthode du trempage des bouchons dans l’eau chaude. Si bien qu’à la fin du siècle, l’opération du bouchage s’effectue d’une manière satisfaisante et que, par voie de conséquence, la proportion des recouleuses diminue.

Le ficelage suit le bouchage. Afin de prévenir toute sortie intempestive du bouchon, on le fixe à la bague de la bouteille par deux ficelles huilées posées en croix par le ficeleur qui doit dérouler la pelote de ficelle, faire un noeud tirer sur les deux bouts de la ficelle et couper, tout cela deux fois pour chaque bouteille puisqu’il y a deux ficelles, avec une cadence de 1 000 à 1 200 bouteilles par jour ! [32] Le travail se fait à la main, avec l’aide à partir de 1844 du ficeleur sparnacien, dispositif inventé par Maurice, puis, à partir des années 1870, à la machine avec le ficeloir mécanique de Nicaise. Au ficeleur succède souvent le metteur de fil qui pose un fil de fer également en croix. Dans certaines maisons, on procède au calottage du bouchon avant la pose de la 2e ficelle en le coiffant par une calotte en fer-blanc portant deux rainures en croix pour recevoir le passage de la 2e ficelle et du fil de fer. Pour faciliter ces opérations, le bouchon n’a pas été tout à fait enfoncé dans la bouteille et se trouve en saillie. Un renflement se forme sous la pression de ses liens et on dit qu’il fait champignon, en prenant l’aspect caractéristique du bouchon de champagne.

Vers 1850, on commence à utiliser une sorte de muselière en fil de fer appelée agrafe de tirage, posée par une machine à agrafer. C’est l’origine du muselet utilisé aujourd’hui pour le bouchon d’expédition. Un peu plus tard on se sert aussi d’une bride, prenant parfois appui sur la cannelure d’une plaque de fer-blanc appliquée sur la tête du bouchon et se fixant sous la bague de la bouteille. À la fin du siècle, la bride supplante l’agrafe, dont elle prend le nom. Elle remplace petit à petit ficelles et fil de fer.

After they had been corked and tied the bottles were placed vertically in willow baskets with several compartments, usually six, which were known as six-cases, but sometimes with substantially more. These would then be placed in the cave (cellar), rather than the cellier (storeroom), because the former, being underground, was cooler and fermentation would thus be slower and more regular, which would reduce the risk of explosions. Some champagne makers, however, preferred to start them off in the cellier, until the secondary fermentation was well under away. The underground cellars had several levels, generally three, and barrels and baskets were transferred between levels via vertical shafts of between one and two square metres in area, known as essors, which also served to provide ventilation.

The cellars were lit by candles, in cast iron lampes de cave, which consisted of a tripod with a candlestick holder that could be adjusted to an appropriate height70. In the second half of the century lighting was achieved with large reflectors placed at the bottom of 15 to 20 metre deep shafts (284). When electricity became available cellars were equipped with with two parallel wires on which a wooden crosspiece would sit which had a strip of copper at each end enabling it to receive the current. In the middle of the piece of wood there was a flexible wire with a lamp on the end, thus creating a mobile lighting system.
Bottles were carried down by hand or most often attached to loops of chain, sometimes known as descendeurs. In cellars that could be entered at ground level baskets of 36 or 48 bottles were transported on little carts.

The next stage was known as entreillage and consisted of stacking the bottles on boards. This involved creating thick walls of bottles, made up of twelve to fifteen double rows, separated by wooden boards, head to tail, all the bottoms of the bottles being on the outside of a row and all the necks on the inside. In this way, if a bottle exploded it would be pushed out and all the pieces of glass and wine would be expelled to the outside of the stack. During this stacking procedure each bottle was marked with a white line which would serve as a guide in future handling operations.

The process of the carbon dioxide becoming fixed in the wine began three or four days after bottling. Guyot notes, rather curiously, in 1860, that the signal that the process of the wine becoming sparkling has begun is the explosion and breaking of several of the bottles in the stack, and even in the twentieth century Moreau-Bérillon was writing that the first bottle to explode was a joy for everyone because it meant that the bottling had been successful and that the process of the wine becoming sparkling was taking place as it should. But the increase in pressure could be followed on a more scientific basis with the use of the aphromètre.

The moment when the champagne was brought out of the cellar was closely related to that chosen for when it would become available for sale, it being understood that champagne needed to spend some time on the boards before reaching its full potential. This period, which varied for different wines, was generally two or three years, there being some slight differences in opinions on the matter, and this is confirmed by the fact that the champagne that one found on the market at that time was usually three years old. Vizetelly wrote that the wine of 1874 was put on the market in 1877, i.e. after three years, as is the case with most champagnes. When it was judged that the time had come to place all or part of a cuvée on the market then the next stage was the désentreillage, which involved unstacking all the bottles and preparing to remove the deposit.


After exploding bottles, and from the beginning of the production of sparkling champagne, the merchants main worry was the removal of the deposit resulting from the yeasts that die during fermentation and form lees made up of globules of the fermenting agent bound together by organic vegetable matter.

In the very early days of the nineteenth century, as in the last century, and with the problems that have been noted, wine was decanted from one bottle into another, leaving the deposit behind73. In 1821 Macculoch recorded that this method was still used for champagne of the best quality. This was an expensive operation. André Jullien stipulated that: Thoroughly clear wine cannot be obtained unless a certain quantity is left with the lees. This residue, known as the "bas-vin" (low wine), is not entirely lost, but its value is halved; to the extent that it is often only used in the wine that is given to the workers.

There was thus much research into ways of improving the decanting process and the idea developed of ejecting the deposit directly from the bottle, having let it settle by gravity on the underside of the cork. This required that the bottles were kept upside down, or neck downwards, until the operation had been carried out. In order to achieve this they were placed in holes on racks that were fixed to the walls, with the result that the stage following désentreillage was called the mise sur pointe ("putting on the ends"). This caused the deposit to slide down towards the cork, and was followed by dégorgement which consisted of expelling the deposit74. André Jullien described these manoeuvres as follows in 1813:

The bottle is held by the neck, initially in a horizontal position; it is rotated on itself until the deposit is loosened and collects in a single mass, in the middle of the lower cavity; the same movement is continued as the bottle is inclined in order to move the deposit towards the neck; the bottle is then placed in one of the holes made in a board, in such a way that the inclination given to it while rotating it is maintained. A second bottle is then dealt with in the same way; each worker usually continues this operation until he has completed one hundred bottles. When the board is covered the first bottle is again rotated and inclined slightly more, and then returned to the rack maintaining the new inclination. The operation is repeated on the other bottles and then recommenced for a third time; but in this last manipulation the deposit descends into the neck of the bottle, which is placed completely upside down in the rack. The wine is left to settle for a few days, until the deposit has become dense and attached to the cork; the bottle is then uncorked without returning it to an upright position and the wine that escapes pushes out the deposit. In this way as little wine as possible is lost; the bottle is returned to an upright position and recorked.

The deposit generally remained on the side of the bottle in a fairly thick, sticky film,. During the fixing of the carbon dioxide in the wine, care was taken to move the stacks of bottles and to vigorously shake each bottle in the process in order to detach the deposit; this was what Robinet called retenir les vins and which was more commonly referred to as a changement de tas (stack change). To do this a bottle was taken by the neck and shaken in a pendulum like movement. The aim of this operation, known as the coup de poignet (shake of the wrist), was to ripen the deposit and to render it more mobile , putting the wine molecules in contact with all the gas molecules. The changing of stacks also provided an opportunity to clean up leaked wine and broken glass.

For stubborn deposits, Jullien invented mechanical brushes to detach the deposit which attaches itself to bottles of sparkling champagne, in order to be able to subsequently disgorge them. But before the latter took place the wine was fined in the bottle, with the help of a n° 3 powder and an aerating pump for fining sparkling wines in bottles. It was often the case that the deposit became firmly attached to the side of the bottle, forming what was referred to as a masque if it was in a single layer, or a griffe (claw) if it displayed web-like folds spreading from a point in the neck. When stacks were moved, in order to avoid the formation of "masques", the deposit should always resettle in the same place on the glass ; bottles were thus returned to the same position in the new stack, workers would use the white marks that had been made during the entreillage. In order to démasquer (remove the mask), the bottle was given a few blows with a hammer, which was known as électrisage, or there was also tapotage, which consisted of vigorously knocking the bottle on a wooden bar, the edge of a rack or table. These operations were also carried out by machines à démasquer, and machines à électriser for which a patent was filed in 1867.

The technique whereby the bottle was removed from the rack several times and the deposit was gently encouraged to move down to the cork, by twisting the bottle and putting it back neck down, was too slow. It did not meet the merchant’s requirements; from the commercial point of view they needed a faster procedure, and this was finally found in 1818. That year, wrote Vizetelly, a man named Antoine de Müller, who worked for Madame Clicquot, thought that there should be a way that bottles could stay on the racks during the deposit moving process, and also that the holes should be made angled so that bottles could be then be inclined at various angles. Tradition has it that Madame Clicquot herself studied the problem, going as far as having one of her tables cut up for the purpose of carrying out experiments; according to Victor Fiévet she even went down to the cellars at night to examine the bottles in order to verify the results of the new procedure.

A hole cut at an angle, if it is wide enough, does indeed offer more possibilities than the simple cylindrical hole in the boards which Jullien describes; but it would seem that the practice of placing the bottles in such boards at a fixed angle of 25° to 30° was continued by some for a considerable period of time77. We know from Doctor Guyot however, that in 1860 the holes were made in such a fashion that the bottles could be placed be placed at a great angle, then more vertically, and finally completely vertical. This is the practice today, and is the best way of obtaining a gradual and total collection of the deposit without worrying about it dispersing each time the bottle is turned.

As late as 1860 and, again according to Dr. Guyot, the planks were in oak, arranged as shelves along the walls of the cellars and supported by four feet. Cavoleau had already described in 1827 "a device for holding bottles upside down" in the form of wide, strong planks drilled with three rows of holes, in which could be placed 3 or 4,000 bottles at one time, adding that workers sat on stools which they moved up and down the rows. To loosen the deposit and tip the bottle closer to the vertical (remuage), the worker would rest his elbow on his knee. However, when there were several levels of shelves on a wall he would only be able to reach the upper shelves by standing.

In the 1830s the bottles were moved fairly infrequently, every two or three days, the bottle remaining up-ended for ten to fifteen days. But by 1889, the Syndicat du Commerce, in its official brochure, was saying that the bottles are given a sharp, light twist and states that this continues every day for between six weeks and two months, the gradual up-ending of the bottle described above by Jullien being carried out at the same time. This is the technique used today.

At the end of the century the person who performed this operation was known as a remueur, and could handle from 20 to 25,000 bottles a day. It is true that in some cellars the operation was made considerably easier from the beginning of the 1840s by the invention of the pupitre (a kind of free standing rack) which, for the remuage of the bottles, gradually replaced the horizontal planks which did, however, remain in use for quite some time. In the house of Vve Clicquot-Ponsardin, wrote Bertall in 1878, the bottles are not put in ’pupitres’, like in most of the establishments in Rheims, but on horizontal tables to avoid the situation of the bottles on the lower levels being neglected (48). Indeed, Monsieur Werlé feared that the remuage of the bottles in pupitres would not be as regular as that carried out on tables or shelves but, in fact, with experienced and conscientious workers, the results are better. A pupitre consisted of two very thick, wide wooden planks, joined by hinges and forming an acute angle when the opposite ends were placed on the ground. Each plank had ten rows, each one six holes wide. A pupitre thus held 120 bottles and, while being large, actually took up less space than the horizontal planks. Furthermore, it could be folded when it was not in use. The remueur worked standing up and thus had access to all the rows.

Once all of the deposit had collected on the cork, the bottles were taken out of the pupitre. They were then mises en masse79, which meant that they were maintained in their upside down vertical position in stacks of several layers, generally three to five, the neck of one bottle resting in the punt (concave underside) of the one beneath it. They would stay there until the elimination of the deposit stage, which would take place after a few weeks or months and, in exceptional cases, after a year or more, while the wine slowly completed its chemical maturation.

We have already heard, from Jullien’s explanations, how the dégorgement process used to be carried out in its early days. To facilitate this operation a hook was used with which the string and wire could easily be removed; this hook was also used to remove the cork, which was never very firmly fixed in the bottle.

The person who carried out this operation (the dégorgeur) would sometimes use a pair of pliers to get the cork out, or at least to control it as it left the bottle. Despite the increase in internal pressure extraction could be difficult and, in such cases a machine à retirer les bouchons (machine to remove corks) was used, which was a sort of powerful cork screw with a crank, stll known as a machine à déboucher (uncorking machine).
Here is how Maumené described the disgorgement process in 1874: the worker takes the bottle, keeping it upside down, and laying it on his fore-arm, removes the wire and strings using a hook; the cork begins to slide out, he holds it in with the index finger of his left hand, and controls its exit with ’disgorging pliers’ or a tool known as a ’lobster claw’, which he holds in his right hand. He pulls the cork out with a quick jerking movement towards the small tilted barrel into which spurts the four or five centilitres of foaming wine that contain the deposit. The worker passes his finger through the foaming wine to expel any of the impurities that might remain, at the same time creating movement in the wine with a few light taps of the hook and by continually turning the bottle in his hands. Maumené warns that all too often the bottles explode in the hands of the dégorgeur, causing him serious injury (394). To protect against this danger he was given leather sleeves to wear and sometimes a mask. Finally, it should be added that once the deposit had been expelled with the cork, the worker would quickly return the bottle to an upright position in order not to lose too much of the precious liquid, the ejected wine, known as the vin de dégorgement, entered the category of the bas-vins (low wines).

Here is how Maumené described the disgorgement process ™™The degorgement was sometimes called the degorgeage, for example by Cavoleau and by Bertall. Redding points out that the most suitable word would have been degagement ]] in 1874: the worker takes the bottle, keeping it upside down, and laying it on his fore-arm, removes the wire and strings using a hook; the cork begins to slide out, he holds it in with the index finger of his left hand, and controls its exit with ’disgorging pliers’ or a tool known as a ’lobster claw’, which he holds in his right hand. He pulls the cork out with a quick jerking movement towards the small tilted barrel into which spurts the four or five centilitres of foaming wine that contain the deposit. The worker passes his finger through the foaming wine to expel any of the impurities that might remain, at the same time creating movement in the wine with a few light taps of the hook and by continually turning the bottle in his hands. Maumené warns that all too often the bottles explode in the hands of the dégorgeur, causing him serious injury (394). To protect against this danger he was given leather sleeves to wear and sometimes a mask. Finally, it should be added that once the deposit had been expelled with the cork, the worker would quickly return the bottle to an upright position in order not to lose too much of the precious liquid (62), the ejected wine, known as the vin de dégorgement, entered the category of the bas-vins (low wines).

However skilful the dégorgeur the level of champagne in the bottle would inevitably have gone down, and this was topped up with champagne of the same cuvée (batch) from another bottle. This operation was called remplissage (refilling). The problem was that it resulted in a considerable waste when the wine was highly sparkling because the foam that forms as soon as a bottle is opened fills the empty space, and pushes out most of the wine that is poured in. It was at this stage that bottles that had leaked (les recouleuses) were topped up, and they of course required a much greater quantity of new wine to be added.

Efforts were made to make the operation more efficient. Jullien invented a double aerated funnel, for filling the highly foaming wines of Champagne when the deposits in the bottles are removed in the operation known as ’dégorgement’, in order that filling bottle and the bottle to be filled can exchange their contents without contact with external air. Maumené developed a garde-mousse ("foam guard") or aphrotèque to empty a bottle under pressure after dégorgement, to refill it with pure carbon dioxide and draw again on the wine previously introduced into the bottle. But none of these processes or any of the les machines à remplir (refilling machines) used from about the 1840s provided a proper solution to this problem, which was not to be resolved satisfactorily until the twentieth century.

In 1884 there was, however, a significant improvement with a process that was patented by Armand Walfard for dégorgement à la glace (ice disgorging). Nevertheless dégorgement without ice remained the rule up until comparatively recently for businesses who could not afford the high costs of the equipment82; in order to distinguish between the two techniques disgorgement without ice was called dégorgement à la volée (flying disgorgement).

The dégorgeur would take advantage of the bottle being open to check the quality of the wine by smelling it. He would also wipe the inside of the neck of the bottle with his finger to remove any residual deposit, and ensure that there was no projection of the cork which, resulting in a fold, could lead to an escape of gas and produce a ’recouleuse’ (dripper) (59). Any suspect bottles would be eliminated and workers received a bonus for any that they found which, in 1913, stood at 50 centimes.

These techniques still left room for error, and fermentation sometimes recommenced after decanting and disgorging, especially at the beginning of the century when it was fairly frequent, and almost considered normal. Once it was in the customer’s cellar, however clear and bright the wine might seem, wrote André Jullien in 1813, it still contains imperceptible particles which have a tendency to precipitate. And he advises leaving the bottles alone once they are in the cellar because any that are moved, even if it is just to look at them, are likely to form, some time later, a more substantial deposit than those bottles that are left undisturbed, especially during three periods, namely when the vines are growing, flowering, and when the grapes are mature. Producers thus made every effort that champagne was perfectly clear, at least when it was delivered.

If the dispatch of the champagne was delayed, wrote Cavoleau in 1827, it would not leave the merchant’s cellar without being disgorged and recorked a second, and sometimes even a third time, if it had been disturbed during the two preceding weeks. A small deposit would always form. It is hardly necessary to mention the costliness of these additional manipulations. And if we are to believe Kirwan, wines were filtered into new bottles, even as late as 1864. He wrote in his Host and Guest concerning champagne that throws a second deposit: It is the custom, in the great cellars of Champagne, to filter the wine into new bottles.


After the dégorgement (the first if there were several), the next stage was the opération du vin83, more commonly known as dosage, which involved softening the wine with a dose of sugar, with a view to adjusting its characteristics in terms of taste, which depended on the type of wine that was being made, and who it was being made for. The necessity for dosage arises because champagne loses all of its sugar during the secondary fermentation. If this has been completely successful, then any sweetness will have disappeared as a result of the total transformation of the sugar; depending on the taste of the clientele for whom the champagne was intended, it was necessary to restore this sweetness, to a greater or lesser degree. It was probably in the 1830s that dosage became a standard procedure84. The demand for substantially sweetened champagne grew quickly, especially as the practice of drinking it very cold (frappé) at the end of the meal became more established as the century progressed. Maizière wrote in 1846 that it was a matter of making champagne more and more worthy of taking precedence, at dessert, over all the other wines of the world, by making it as perfect as possible. As we shall see, this was not universally accepted, but the official doctrine, as the dawn of the twentieth century approached, was that it was necessary to dose heavily: an information brochure prepared by the Syndicat du Commerce des Vins de Champagne (which became the Syndicat de Grandes Marques, and then in 1994, the Union des Maisons de Champagne) declared that following fermentation in the bottle, the wine has almost completely lost its sugar and becomes virtually undrinkable.

It would seem that in France it was necessary to wait until 1890 for a market in lightly dosed champagne to emerge. However, in the Anglo-Saxon countries there was a demand for dry champagne from about the 1850s onwards. Several merchants in Rheims, Ay and Epernay sold English Cuvées that had much less added sugar (589), and in 1855 Moët & Chandon printed labels for sec and dry champagne for the American market, and for the English market in 1861. Dry wines were popular in North America, but in fact it was in Great Britain that there was a real taste for dry champagne, because there, more than in France or the Unites States, champagne was drunk throughout the meal.

Here is Thomas Walker’s opinion on the matter in 1835: Champagne should be served at the very start of the meal, it should be placed on the table so that everyone can serve themselves as they please. Charles Monselet, describing in La Cuisinière Poétique meals aboard a transatlantic liner, wrote: Champagne is generally drunk, many English considering this a table wine, and in 1855 one finds in the Exposition de Troyes Illustrée cited by Plonquet, that a large number of English drink nothing else but champagne from the start to the finish of their meal. William Younger confirms this, writing in reference to the same year of 1855, that there was thus established a habit that would prevail until the end of the Victorian era.

However, as early as 1848 a London wine merchant by the name of Mr. Burne had asked Perrier-Jouët to send him his 1846 cuvée without any dosage, so that he could sell it more easily as a fine table wine than the champagne that was usually sold in the dessert wine category, in which it came up against competition from port and madeira. This ended in failure, due partly to the total absence of the addition of sugar; the strategy may have been successful if the champagne had been very lightly dosed. And even after 1855, and all through the nineteenth century, many British champagne drinkers, like today’s sherry drinkers, liked bottles which said "dry" on the label but which in fact were slightly sweet.

Whatever the case may be, dry champagne became popular in Great Britain, especially in London, and even became associated with a certain snobishness [33]. An imaginary conversation during a champagne dinner appeared in Punch on 12 April 1862, written by John Leech. An uncle, having given a glass of champagne to his young nephew, says to him, "George, my boy, here is a glass of champagne for you", and his nephew replies, "H’m ! Terribly sweet. Good for the ladies. I have arrived at a stage in life when I must confess that I like my wine dry". Many people, however, continued to prefer their champagne very sweet and might say of somebody whose tastes and manners they did not much care for, "No doubt they pretend to like dry champagne!" Even it was relatively slow [34], the transition was irreversible and most of the big houses were sending dry champagne to Great Britain, and to a lesser extent to the United States, from around 1860. This was especially true of Ayala, whose 1865 vintage was said to, during the 1870s, have converted the Prince of Wales, the future Edward VII to dry champagne (16), of Bollinger, also with the 1865 vintage, of Pommery, who had a great success with their 1874 vintage87 , that went on the market in 1879, and of Vve Clicquot, who sold a dry 1857 and shortly after 1870 introduced a perfectly dry wine.

The quantity of sugar added to champagnes destined for England progressively decreased, at least it did for some of them, and from 1865, in addition to sec and dry, champagnes were sent that were described as extra-sec, very-dry and extra-dry. In the list of the main brands compiled by Vizetelly in Facts about Champagne, one finds, in 1879, nineteen houses offering one or several of these degrees of dryness, and for eleven others it was possible, without it being specifically mentioned, that they too had them available. Even though the British market had developed considerably as a result of the introduction of dry champagnes, some producers remained very reticent concerning lightly dosed wines. Louis Roederer junior is reputed to have said that he would not, as long he was alive, bow down before false gods.

In the 1870s the cult of sweet champagne slowly but surely approached its end (589) on the British market, where the escalation of dry wines continued with a new type, called brut, labels for which began to appear shortly after 1876, in particular for the wines of 1874. In the terminology of champagne technique the word brut then referred to wine that had just been disgorged, and which was still in its natural state. It could only be drunk like this in years of outstanding quality and when the wine was high in alcohol, or when the wine was very old (351) but this was the exception. Similarly in 1929, the fifth edition of Weinmann’s Manuel du Travail des Vins Mousseux informs us that: After disgorging, sparkling wine is brut, it is not generally agreeable to drink in this condition. A little liqueur is added even to so-called brut wines.
To satisfy the British market some merchants supplied a wine that was labelled brut but which was actually dosed with between 10 and 30 grams of sugar per bottle (650). In 1879, in the aforementioned list in Facts about Champagne, there are four brut champagnes, Moët & Chandon, Pfungst, Pommery et Greno and De Saint-Marceaux, which, three years later, were joined by G.H. Mumm and Roper Frères in a similar list published, again by Vizetelly, in A History of Champagne.

In France, however, it was not until around 1910 that heavily dosed champagne began to lose ground to dry champagne. This can be established from the merchants’ archives. But Cyril Ray provides an interesting confirmation by comparing two editions of the Gourmet’s Guide to Europe. In that of 1903 the author recounts that, having asked a merchant in Champagne for some vin brut, in a hunting party, he received as a reply, "Ah! You drink that poison". In the 1911 edition the anecdote has disappeared. It is thus hardly an exaggeration to say, as does Patrick Forbes, that dry champagne was a British invention.

At the end of the nineteenth century there were thus two categories of champagne: dry champagnes, which were labelled with various terms depending on their degree of dryness; and heavily sweetened champagnes, which were not classified in any particular way, except in Anglo-Saxon countries where they were sold as medium-dry, and the sweetest as full or rich. We already know that the British liked their wines very dry. The Americans liked them markedly less dry. In Germany, Belgium, Austria, and France, sweet champagne is preferred. This remained true for a long time, even for drinking during meals. In 1885, when Guy de Maupassant’s Bel-Ami dines at the Café Riche with Madame de Marelle, the latter asks for, "Some iced champagne, the best, sweet champagne for example!" Vizetelly, however, specifies that the French drank moderately sweet wines, while for the Germans it was necessary to dose the wine four times as much as for the English.

As for the Russians, they liked their wine as sweet as a young woman’s hot toddy and at one point it was even the fashion in France, although probably not very good form, to make a Romanoff by pouring champagne into glasses that contained a substantial quantity of a very sweet, coloured liqueur.

The amount of sugar added to the bottle thus varied considerably depending on its destination. This was expressed in a percentage which corresponded to the number of centilitres of sugar solution added to the bottle. Thus for a dose of 10%, 10 centilitres was added. We will see later how this solution was prepared, for there were several types, with varying proportions of sugar. But for the time being we may suppose, with Maumené, that it was made up of 150 kilos of sugar and ten litres of brandy for 125 litres of wine, which corresponds to very close to 11 grams of sugar per centilitre of what was referred to as liqueur. In 1882, and therefore in Maumené’s time, the dosages were, according to Vizetelly, as follows: 2 to 6% for the British or 22 to 66 grams of sugar per bottle, known as the goût anglais; 10 to 15% for the Americans, or 110 to 165 grams, known as the goût américain; 15 to 18% for the Germans and the French, or 165 to 200 grams, known as the goût français; and 25 to 30% for the Russians, or 275 to 330 grams, which was virtually a syrup, one third of the bottle’s volume being sugar solution!

It is hard to explain today how the Russians could possibly have liked this drink that could barely be described as champagne, perhaps a clue lies in the advice of the house of Pommery whose publicity material in 1891 specified that it had to be drunk outrageously cold. Even in other countries, including Great Britain, a large proportion of the champagne drunk [35] was, by today’s standards, terribly sweet.

While dosage was essentially dependent on the tastes of the customers, it also depended, but to a lesser extent, on the age and character of the wine to be dosed. It should also be noted that the same percentage might not, for two different reasons, have the same value, because each producer had his or her own ideas about the relative concentration of the sugar solution. Due to the general tendency towards less sweet champagne, the average dose in 1913 was no more than 0.5% for England, 8% for France, 10% for Germany and 20% for Russia, the dosage for America also decreased, possibly to an even greater degree, for Pommery’s Notice sur le Vin de Champagne remarked that in the United States, from around 1891, there has been a demand for wines almost as dry as those in England. In terms of total production, about 60% of champagne was dosed at 5% or above, as against 80% in 1882.

Dosing was carried out by adding to the bottle what was called the liqueur d’expédition ("dispatching liqueur"). This was the name given to the solution used for dosage from about the middle of the century, which contained varying proportions of sugar depending on producers.

Here is what Maumené had to say on the matter in his Traité Théorique et Pratique du Travail des Vins (Practical and Theoretical Treatise on Making of wines): The liqueur is theoretically a solution made by dissolving pure sugar in wine; but in practice it is a much more complex liquid, for which everyone has their own recipe, depending on the public’s tastes and on the wine maker’s ability to accommodate these tastes. He then lists the various ingredients that were usually employed in a liqueur being prepared for the English market: sugar, water, champagne of the same vintage, port, spirit of cognac, ordinary eau-de-vie of cognac, dark eau-de-vie of cognac, Fismes colouring agent, kirsch, raspberry flavoured alcohol, saturated alum solution, tartaric acid and tannin.

The basic liqueur, made up of sugar, water, and some wine of Champagne, was boiled until it had reduced and then the other ingredients were added. This rather culinary nature of this operation did not please everyone. Dr.Guyot wrote: Bad wines gain nothing and good wines suffer, losing, at the very least, their original character and flavour (283). And the Vigneron Champenois, in the 16 September 1885 issue, thought it necessary to protest against the calumny according to which the liqueur d’expédition contains kirsch, port, etc., attesting that it was an exceptional and fraudulent practice, although we have just seen that Dr. Maumené considered it normal, and even in 1894, Monsieur d’Avenel, in the October issue of Revue des Deux-Mondes, reported the use of the same ingredients in the liqueur, to which he adds vanilla colouring.

The liqueur was also used to darken the colour of the champagne, if this was deemed necessary. Robinet wrote in the fifth edition of the Manuel Général des Vins that the customer likes or is used to having a wine of a certain colour; but this is different every year, if the customer is used to a lightly coloured wine, then the producer... gives his liqueur an appropriate colour that is then passed on to the wine and gives satisfaction. He added that to colour the liqueur, the colorant known as ’teinte de Fismes’ is used in Champagne.

The spirit used in the liqueur d’expédition was generally 82° spirit of cognac, according to
Robinet, who also wrote that if the wine is to travel at sea, the proportion of alcohol in the dosage is increased, because this increases the dissolving power of the wine, the gas is less likely to expand and the risk of explosion is reduced (542). Great importance was attached to the choice of sugar. Maumené, in the work already cited, warns of the special and disagreeable smell of beetroot molasses, which used to come from imperfect processing, and recommends sugar candy from cane as the only type which may be mixed with wine in considerable proportions without causing it to lose its scent or flavour. For the wine it was generally the best quality that was used. As Robinet noted, the aim of "operating" on a bottle of wine is to improve it, the wine that is added must therefore be chosen with care (542). And the Vigneron Champenois of September 1885 stipulates the use of old wines that no longer have any tendency to ferment.

The cellar master decreed the quantity of liqueur to be added. The necessary space in the bottle had in principal been allowed when the remplissage had been carried out. If, however, if there was not enough space then the excess would be removed from the bottle by the doseur, but this was a slow process, during which the wine would be exposed to air and lose some of its carbon dioxide, resulting in reduced sparkling qualities and detracting from the finesse of the bouquet. Next, using a small measure with a spout, the doser would gently pour in the liqueur, continuously rotating the bottle, so that the liqueur slowly descended into the wine without agitating it; if the wine was overly disturbed then this could result in gas being generated and the wine foaming out of the bottle. This was a delicate operation, as was remplissage (partial topping up with champagne of the same batch after disgorgement) and for the same reasons. The consistency of the dosage depended on the person who carried it out, and there were inevitably slight variations between bottles. This is why the dosing machines, known as doseuses, of which early versions appeared in 1844, represented great progress; the procedure could be executed without exposing the wine to air or dust and the dose to be added could not only be varied but was also guaranteed to be the same for each bottle. Small producers did nevertheless continue to use manual dosing into the twentieth century.

After the addition of the liqueur, the next stage was égalisage which was the final topping up of the bottle by hand by the égaliseur ("equalizer"), or by a machine, the latter, as time progressed, becoming increasingly advanced and eventually performing all three of these stages (remplissage, dosage and égalisage).

The bottles now had to be corked and retied with string just as they had been at the first bottling. This time, however, the cork was not pushed in quite as far, thereby creating more latitude for it to "mushroom". Once again the destination of the bottles was a factor, the corks in principal being pushed in further for those intended for export. These final corks or bouchons d’expédition were chosen with particular care and branded with the name of the house. String and wire were still used to hold them in, and for a little while longer tests for instantaneous uncorking were carried out, such as the needle system. But by the end of the 1870s the metal muzzle or cage, known as an agrafe, that had previously been used in the initial bottling, began to be used to crown and control the cork after the final corking, becoming known as a muselet, and being applied mechanically. This was called muselage (muzzling) and the first machine to carry out this operation was developed around 1880 by Lemaire, in Epernay. However, some customers remained attached to the traditional technique, and up until the First World War ficelage (attachment with strings) and muselage coexisted. All of these processes, from dégorgement to muselage, were carried out in an area called the chantier de dégorgement or the chantier d’opération. Bottles were moved along on revolving stands called rondoirs.


The bottles, duly corked and tied were then sent to the cellier d’expédition (dispatch cellar), where the first procedure was to make sure that the liqueur had properly combined with the wine by giving each bottle a quick shake with a wrist action similar to that used when the stacks were being moved around. To perfect the wine’s condition it was left undisturbed for two months, Maumené explained that the wine and the liqueur have a mild effect on each other, through which the process is finally completed. This interlude was used to test the quality of the new corks. At the time of dispatch each cork is carefully examined [36]. If it is considered in any way defective then the bottle is withdrawn and not dispatched. This results in a loss of five to ten percent and sometimes more. Bottles that are considered satisfactory are washed and dried; the corks are covered with wax or white or gold tin foil. This was called surbouchage ("overcorking"), which dates from 1856. There was also an operation which Legrand called capsulage (351), and which Guyot mentions in 1860, specifying that the capsule consists of a general envelopment of the cork and the ridge of glass at the top of the bottle’s neck ; it was thus a sort of cap. The capsule was put on mechanically from 1865 by a machine called a capsuleuse. The original purpose of surbouchage and capsulage was to stop rodents gnawing through the strings that held in the corks.

It was around 1820 that printed labels replaced handwritten petits billets de vin (little wine tickets) on bottles of champagne. As soon as these appeared they stood out on account of their elegant design, that was often embellished with gold and silver. They were stuck on by hand. Some houses did not use them, also dispensing with surbouchage, at least for bottles that were to be exported. An 1862 Moët & Chandon price list for Belgium declares that the house’s practice is to dispatch wines with tarred corks and in bottles without labels; if tin sheets and labels are required then this must be mentioned when placing orders.

Finally the bottles were packed in willow baskets for orders on the Continent or, from about 1830, in wooden crates, usually made of poplar, for more distant orders, with the destination branded onto the lid. There were baskets of 10, 12, 15, 18, 20 and 25 bottles and more (up to 130), which were closed with ten brass hooks attached by the fermeurs ("closers"). There are crates of 12, 25, 50 and even 120 bottles for sending to Hungary and Russia.

The packing was carried out by hand, as was the labelling, until at least 1900. A great deal of care was taken. Here is an account from André Jullien in 1845: In Champagne, the insides of the baskets and crates are covered with sheets of cardboard and each bottle is wrapped in a sheet of paper, over which is placed a twist of straw. For America and the other southern countries, the bottles are protected against the excessive heat by packing them with salt... the bottles are placed between two layers of salt and each row is separated by a bed of straw (317). Straw cases were also used, which were produced mechanically from 1866. Once closed, the baskets and crates were encased in strips of wood or metal and were ready to depart.

All of the tasks in the preparation of champagne, from the decanting of the must to the packing, represented a vast amount of labour. Some houses employed several hundred workers, known as tonneliers (coopers), even though a comparatively small number of them were actually involved in making barrels; they were supervised by a maître-tonnelier, an important figure who was responsible to the merchant for the quality of his brand of champagne. Each bottle would pass through the hands of forty five workers (351). In the small houses there were, however, sometimes only twenty, ten, or even five workers, and each of them would carry out several tasks. The champagne industry employed large numbers of women in the nineteenth century and took on children to help with the bottling. All the staff were efficient and aware that they were producing a product in which quality was paramount.

Cleanliness was considered important. Tonneliers wore blue shirts and white aprons. The maître-tonnelier hired them, fixed their wages, and maintained accounts; he distributed tasks, supervised the preparation of the wines, the handling of the bottles, he tasted the batches of wine, gave his opinion on its various qualities, prepared orders for dispatch, received supplies, and generally occupied himself with the staff, equipment, materials, and all the orders that left the champagne house.

The lot of a tonnelier was better than that of workers in most other industries, but was still somewhat precarious compared to today’s workers. As was the norm in those days, the working day was very long, even if it was less in 1861, with sixty hours a week, than in 1806 when the usual week was eighty-four hours, of which sixty-six were actual work, days starting at between five and six-thirty in the morning, depending on the time of year, and always finishing at eight in the evening, with three quarters of an hour for breakfast and afternoon tea, and an hour and a half for lunch at midday.

The work became simpler and faster as mechanical progress was made, which was often linked with progress in chemistry. Machines, noted the Vigneron Champenois in 1874, supply with great speed a result identical to that obtained by manual labour, and avoids all the faults inevitable in this kind of work. Their profitability was guaranteed not only because of the economies that could be made in terms of labour and time but also because of the reductions in the loss of wine at various stages of production.

As we have already seen concerning the problem of losses due to exploding bottles, there was considerable research carried out into improving production methods. Often benefiting from financial support from the merchants, research was carried out in every aspect imaginable, as can be seen from the list of patents filed from 1843 onwards. Amongst these there are even projects that were never followed up, such as machines à dégorger, in 1852, and machines for shaking the bottles in 1875 and 1876, and again in 1889.

To draw up a list of all the devices that were invented during the nineteenth century, most of which appeared before 1880, would require an entire volume. It will suffice here to recall, amongst the main ones that were in use at the end of the century, the wine pump, the bottle and cork checker, the bottle washer, the bottler, the corker and string tier, the capsuleuse, the muzzler, the decanter, the tapper, the filler, the doser, and the topper-upper, not forgetting the aphromètre and numerous laboratory instruments.

The wine machine industry was thereafter closely linked to the champagne industry, as, moreover, were many professions, listed as follows by Roux-Ferrand in Mœurs Champenoises (Customs of Champagne)(1861): makers of barrels and bottles, cork makers and retailleurs ("recutters")who would restore old corks and sell them to be used for bottling sparkling mineral water and beer, makers of muzzles, crates, baskets, lithographs, merchants of all sorts of raw materials: sugar, alcohol, wire, string, packing paper, cardboard, tar, tin sheets, capsules, tannin, fish glue, tartaric acid, etc.


Was champagne good in the nineteenth century? There is no cut and dried answer. Tastes have changed and, due to its heavy dosage, it would be unlikely to please today’s drinkers of champagne, with the exception of some of the cuvées blended for the English market. From an objective point of view, it has to conceded that, apart from the generally excellent champagnes produced by the biggest houses, quality was not consistent, due to techniques that had yet to be perfected and a total absence of any official control. The art historian Spire Blondel wrote in 1894 on the subject of champagne: When travelling one should be very cautious if one does not know the cellar of one’s hotel.

Be that as it may, champagne was a popular drink of the time. It is true that for many people that it sparkled and foamed was all that really mattered, its quality as a wine being made even more secondary by the fact it was served extremely cold, and it was thus difficult to appreciate any subtleties in its flavour. As for the informed amateurs, they would seek out the very good champagnes, which were in the minority because they had to fulfil three criteria that were at that time rarely fulfilled, namely: they had to be well made, have been well aged and kept for a sufficient amount of time in the producer’s cellars, and, of course, be from an excellent vintage. At the start of the century champagne was frequently defective. Cyrus Redding wrote that good champagne is one of the most wholesome wines but that bad champagne is more commonly pernicious and he describes wines which, in 1818, were of very poor quality, dressed with sugar and alcohol. In 1821 Jullien recommended, in the Notice sur les Vins that he wrote for Archambault’s Le Cuisinier Econome, on the subject of sparkling white wines, that they should not be kept too long because they will eventually become oily and unpleasant to drink. We also know that corks were not as reliable in those days and could bring disappointment.

As technical advances were made, accidents became less frequent, especially those that stemmed from a defective corking procedure, and those that resulted in the formation of a second deposit in the customer’s cellar.

The problem of graisse, a bacterial disease that gives wine an oily appearance, was finally resolved. Herpin wrote in 1819 that, in the wines of Champagne white or yellowish deposits are the most dangerous: with the slightest movement they spread through the wine, either as flecks or in thick globules, or most often in fatty threads which extend in all directions. In Champagne the entire bottled harvest is often afflicted with devastating transformation. This nightmarish description is matched by the bizarre list of experimental and unsatisfactory concoctions that were added to the wine when it was still in the barrel such as fresh meat, horseradish, radishes and turnips! Herpin recommends the use of cream of tartar, but the true remedy was found by François in 1835 and, as a result of his advice, the disease known as "graisse" was eliminated by the use of tannin, added on the basis of 15 to 20 grams/hectolitre, in an alcoholic solution three to four weeks before bottling.

Good champagnes, as we have noted, were usually sold after being aged for three years, and sometimes more. Unfortunately, too much champagne was put on the market just a few months after bottling, sometimes as a result of the bottling itself not having been successful; the champagne of 1818, which was mentioned by Cyrus Redding, was sold after just two months in the bottle! These overly green champagnes were always in the heavily dosed category, this type of champagne, explained a notice in 1891, having the immense advantage of being able to be delivered and drunk very young.

The choice of years and areas where the grapes were grown that made the wines that went into the blends was of prime importance. Producers were still at the mercy of nature, we have already seen the losing battle that was fought against natural disasters in the nineteenth century; in the twenty years from 1871 to 1890, there were fifteen that were either mediocre or bad. It was only in the periods of plentiful harvests that the vins de réserve could properly play their role. A happy combination of improved technique and a series of generous years enabled André Simon to write in 1908: Never have such good champagnes been made than in the last twenty years (589). But, in the meantime a lot of mediocre wines were produced, and attempts were made to salvage them with the addition of sugar and alcohol.

For the crus, no such manoeuvres were possible; the houses that bought their grapes from the best crus produced the best blends. Grapes from the more modest crus were used by the same houses to make their second choice blends, this principle sometimes continuing as far as a fifth choice blend, and by those producers, both small and large, who had deliberately chosen to sell at the lower end of the market. But in order to satisfy the growing demand for champagne, everyone began to extend their grape and wine purchases to areas beyond the traditional crus. At the beginning of the nineteenth century, the vineyards of the department of the Marne were the only ones to supply this famous wine (318). And in fact it would have been more accurate to say that it was only those located in the Marne valley and the cliffs of the Ile-de-France.

Initially grapes were sought from crus on the edge of the department, in the areas of Vitry-le-François and Sézanne (48). This then extended to the rest of Champagne, i.e. the Haute-Marne, the Aisne and above all the Aube. Champagne’s capacity to supply good quality wines for champagne production had not been exhausted when producers, in the second half of the century, began to turn to other wine producing regions likely to be able to supply them at better prices with wines from climates more suited to growing vines, and of which the transport had become much cheaper since the arrival of the railroad.

By 1873 the situation was such that Paul Urbain and Léon Jouron wrote: The secondary wines of Champagne, which contain only two thirds and sometimes less of local wines...are made with grapes harvested in vineyards that are entirely secondary in Champagne, and mainly with white wines from Burgundy, the Midi and Saumur, which are cheaper than any from our own vineyards. This phenomenon was not only true of champagne. The Wine Trade Review of 15 January 1874 declared that the high prices of the true wines of Champagne and Bordeaux locally have resulted in several speculative houses substituting imitations.

Legally there was nothing to stop this practice which, however, harmed the reputation of Champagne’s wines and constituted very reprehensible trickery over the nature, origin and quality of the merchandise delivered. It was frequently denounced in the press and specialist literature as a known and regrettable activity. Bertall wrote: The sparkling wines of Champagne are known to be made with wines from several provinces.

Champagne producers, attempting to avoid such accusations, would keep quiet about their erring ways and base calculations on exaggerated areas and yields, thereby avoiding having to include in their records grapes bought elsewhere. However, given the impossibility of completely denying that such practices took place, they endeavoured to cut their losses and would declare that it was a case, as Guyot wrote, of a few false brothers and that these falsifications are still rarer in Champagne than in most other areas, where great reputations are always exploited by audacious speculation. One would like to believe that this fraud was not widespread91, and that it hardly affected the big houses, but there was no doubt that it did exist and the consequences were damaging for all the merchants, and therefore for the vine growers, the public being aware that, with the exception of the best cuvées, the quantities of champagne that were produced would require something akin to the miracle of Cana.

Every effort was made, with the methods then available, to make these foreign wines as similar as possible to those coming from the slopes of Champagne. Thus Weinmann wrote in 1899 that for the exotic wines, the must is fermented with yeasts cultivated in Champagne. But the result could never be the same, and it was still fraud. So much so that, in order not to be cheated, it is necessary to buy only the best crus from the oldest and most reputable houses in Champagne, houses which, having laboriously and honestly acquired a fine reputation, could not allow themselves to descend to such methods of production.

It was thus well established in the nineteenth century that there were two markedly different categories of champagne. There was that of the true champagne lovers which, as it was more expensive, was also that of wealthy people who liked to think of themselves as connoisseurs. These great wines kept better than the others and Vizetelly wrote that they could be drunk after ten or fifteen years of age, having then acquired a supreme excellence. Then there was the champagne of the more common sort, which was less refined and cheaper, but of which the image remained sufficiently attractive to ensure its success throughout the century. Both categories existed in several types which we have already come across in the course of this chapter but of which it will be useful here to present a summary.

At the end of the century the following sparkling wines were sold: grand mousseux ("highly sparkling"), mousseux, crémant or demi-mousseux (there was also sometimes crémant demi-mousseux, known on the English market as "creaming"), and tisane de Champagne; there was millésimé and non millésimé (non -vintage); rosé (known on the English market as "brown sparkling champagne", to distinguish it from "pale sparkling champagne"), œil-de-perdrix ("partridge eye"); brut, sec (and very dry, extra dry, extra sec, très sec, dry) and dosed champagne, which was known as medium-dry or full or rich in Anglo-Saxon countries, but to which no particular term was applied in France. A champagne could of course have several of the above characteristics, being, for example, crémant, rosé, sec and millésimé all at the same time.

Until the Second Empire the word champagne did not generally figure on the label. The name of a prestigious cru was used such as Ay or Aï, (or even Aÿ), Sillery, Verzenay, Bouzy, followed by mousseux, or grand-mousseux, as in, for example, Sillery mousseux. There was even a Aï Royal Fleur de Sillery mousseux !

This was a strange custom, given that the contents of the bottle usually had nothing to do with the locality mentioned and might even have been, as we have just seen, made with wine from Lorraine or Anjou. In 1842 the indignation of the Comité Vinicole du Département de la Marne was roused in response to a questionnaire from the Comité Vinicole Central: It is derisory to give to our sparkling wines, on shining gold and solver labels, the name of Sillery mousseux. That Sillery delivers its harvest from 20 to 25 hectares, that Ay puts its mark on 1,000 hectares, Mareuil on 800, and so on; but that the trading houses, who did not spend a penny purchasing any of the harvests from these vineyards can be said to have vats full of Sillery, Ay and Mareuil.

Despite protestations this practice continued up until the First World War; on the subject of the blending of crus sold under the name of one of them, Roche write in 1908: This procedure is now widespread. It was true that the custom had developed in high society and in literature to frequently replace the word champagne with the name of a prestigious cru. In Les Natchez, Chateaubriand wrote that the Duc d’Aumale tasted the wine of Aï before the milk of his wet nurse.

At the same time, from the 1860s onwards, the word champagne appeared on some labels [37] but was for a long time accompanied by mousseux. This remained the case until the beginning of the twentieth century.

The trading name of the producer was usually on the label, but this was not always the case, especially at the beginning of the century. It could be totally absent, or replaced, or accompanied, by a phrase, not necessarily registered as a brand, that referred to a well known event or to a particular clientele. This will be dealt with later in the study of the commerce of champagne. Finally, the label would carry an indication of one of the numerous grades on which the producer based his or her production and prices. In 1890, for example, the house of Dufaut proposed seven grades, namely, from the best to the most modest, Grands Vins de Réserve, Carte Blanche, Bouzy, Fleur de Sillery, Grand Crémant, Royal Sillery, Ay mousseux. Many houses used the word carte, in association with various colours, which were generally, in decreasing order of excellence, gold, white, green or blue, and black. From the 1870s onwards some houses kept carte blanche for their dosed champagne. Mumm, instead of carte, used cordon: Cordon Rouge, for America only.

To establish the order of merit in little known brands one would have to look at the prices, the terms used to qualify the various qualities often being somewhat fanciful. Royal, Supérieur, Spécial, Extra, Extra Supérieur, Grand, Crème (which sometimes applied to crémant, but not always), Fleur was often supplied to modest champagnes. Réserve and Cuvée Réservée was usually applied to the best, as was Cabinet which, in French as in English, was used in reference to the fashion for private cabinets (rooms), sometimes in the absence of any other term identifying it as champagne, such as the Grand Vin de Cabinet de Koch Fils of 1884. The big houses increasingly developed the habit of keeping the names of crus or unusual titles for their inferior quality wines, their great champagnes carrying nothing more than the name of the house.

The label would not describe the physical properties or characteristics of the wine, which producers endeavoured to make comply as far as possible with the tastes of the clientele.

Thus, Robinet wrote, when one is making sparkling wines destined for England or Germany it is obvious that the elements are not the same. The first country requires full-flavoured wines that are high in alcohol; the other requires finer, lighter, more pleasant wines. Vizetelly gives other examples, such as China, India and the other hot countries, which require light, dry wines, and Australia and the Cape, to which we send very strong wines.

Only the very good houses could create champagnes every year that were consistent and varied, by combining the possibilities that were offered to them by the excellent grapes that came from various vines, their vins de réserve, various degrees of dosage and various ageing periods.

There were thus a lot of different champagnes in the nineteenth century. Choice was a matter of personal taste. Louis de Chevigné resolves the dilemma with an eclectic approach:

... But joyful champagne
Kept cool in my cellar does await me.
Which do you want ? From the valley or the mountain ?
Aï, so proud, claims first place;
But I prefer a Sillery crémant.
Let us unite these Kings of Champagne,
Says the chaplain, and drink them both gaily.



[1YOUNGER (William). Gods, men and wine. London, 1966.

[2COUANON and F. CONVERT. Board of admission’s report and installation of classes 36 and 60: Viticulture. Wines and wine spirits. International Universal Exhibition of 1900. Saint-Cloud, 1900.

[3The appearance of turbulence was the first phase of fermentation. Also called bouillage (frothing) this was the period of intense activity that at the end of the nineteenth century was still left to continue, fairly fast, until it was more or less finished, depending on the composition of the must and the temperature of the cellar.

[4It would seem to be in error that Manceau, in his Theory of Sparkling Wines, suggests that during the eighteenth century the necessity of adding a little sugar to the wine to make it sparkling was acknowledged.

[5On emploie généralement au XIXe siècle le mot ferment et au XXe celui de levure. Salleron écrit en 1886: le ferment alcoolique, la levure comme on l’appelle communément.

[6CHAPTAL (Jean-Antoine, comte). L’Art de faire le vin par M. le Comte Chaptal. Paris, 1819. CHASTELAIN (Dam Pierre). Voir JADART.

[7LENOIR (A.). Traité de la culture de la vigne et de la vinification. Paris, 1828.

[8FRANÇOIS. Traité sur le travail des vins blancs mousseux. Châlons, 1837.

[9SALLERON (J.). Etudes sur le vin mousseux. Paris, 1886.

[10WEINMANN (J.). Manuel du travail des vins mousseux. Épernay, 1929.

[11VOGUÉ (Comte Bertrand de). Madame Veuve Clicquot à la conquête pacifique de la Russie. Reims,

[12TOEPFFER (Rodolphe). Histoire de M. Cryptogamme. Genève, 1846.

[13BARRELET (James). La Verrerie en France de l’époque gallo-romaine à nos jours. Paris, 1953.

[14GUYOT (Dr jules). Culture de la vigne et vinification. Paris, 1860.

[15En 1892, J. de Saint-André écrivait: Si la quantité de sucre contenue dans le vin est trop faible, on manque la mousse, c’est-à-dire que celle-ci n’est pas suffisante, on dit alors que le vin est crémant . En 1918 encore, Pacottet et Guittonneau disaient que le vin crémant résulte souvent d’une cuvée manquée.

[16Le crémant était au XIX e siècle appelé creaming en Grande-Bretagne. On le trouvait sur les prix-courants de The London Wine Company, en 1823, et de Charles Harris, en 1829.

[17JULLIEN (André). Topographie de tous les vignobles connus. Paris, 1816, 1822, 1832, 1866 (Se édition, revue et corrigée et augmenlée par C.E. 7ullien).

[18RAISSON (Horace). Gode Gourmand. Paris, 1829.

[19JULLIEN (André). Topographie de tous les vignobles connus. Paris, 1816, 1822, 1832, 1866 (Se édition, revue et corrigée et augmenlée par C.E. 7ullien).

[20LURINE et BOUVIER. Physiologie du vin de Champagne par Deux buveurs d’eau. Paris, 1841.

[21Notice sur le vin de Champagne (Maison Veuve Pommery, Fils et Cie. Reims, 1891.

[22BONNEDAME (Raphaël). Notice sur la Maison Moët & Chandon. Épernay, 1894.

[23Guyot en 1860 employait l’expression liqueur à vin, et Robinet en 1877, et encore en 1900, celle de liqueur de titrage. Aujourd’hui on dit liqueur de tirage.

[24ARON (Jean-Paul). Le Mangeur du XIX e siècle. Paris, 1973.

[25HENDERSON (Alexander). History of Ancient and Modern Wines. Londres, 1824.

[26REDDING (Cyrus). À History and description of Modern Wines. Londres, 1833.

[27VIZETELLY (Henry). A History of champagne with notes on the other sparkling wines of France. Londres, 1882.

[28C’est pour le marché britannique qu’a été conçue la pinte, contenant un huitième de gallon, citée en 1879 par Vizetelly. Selon André Simon la pinte impériale était à l’ère victorienne la taille idéale pour un homme tempéré qui pouvait considérer la bouteille comme un peu plus que ce qu’il souhaitait pour son repas, mais qui ne se serait pas satisfait d’une demi-bouteille.

[29Pour Littré, le mot recouleuse est un adjectif: Se dit, en Champagne, des bouteilles dont le vin s’échappe à travers le bouchon.

[30ROBINET (E.). Manuel général des vins. Fabrication des vins mousseaux. Paris, 1877

[31CAVOLEAU. Œnologie française ou Statistique de tous les vignobles et de toutes les boissons vineuses et spiritueuses de France, suivie de considérations générales sur la culture de la vigne. Paris, 1827.

[32MAUMENÉ (E. J.). Catalogue des instruments et appareils pour le travail des vins mousseux inventés par MM. E. Maumené et L. Jaunay. Corbeil, 1873.

[33According to Cyril Ray, the British middle classes had an ostentatious taste for dry champagne.

[34Professor Saintsbury observed that the transition from sweet to dry had not completely finished in Great Britain when he began to keep his cellar book in 1884.

[35There were not only rich and medium-dry champagnes sold in Great Britain for sweet champagne drinkers, but also champagnes à la Russe.(Russian style) for those who liked their champagne especially sweet.

[36The testing of the corks was entrusted to the releveur (translator’s note: nowadays this word is used to describe someone who reads meters).

[37In the Anglo-Saxon markets, the word champagne, also written champaign, had been used earlier than in France. For example, Verzenay champaign 1846.