It has long been known that wines have a natural tendency to be very slightly sparkling; this is due to the carbon dioxide that can be present in varying degrees. Amongst the factors involved are the relative quantities of sugar, the types of yeast that are present and the temperature. Their combined effects can in certain cases result in a sparkling wine. Sugar is supplied by the grapes, as are the yeasts, which are the agents of fermentation. So long as fermentation remains unfinished a sugary residue will remain in the wine with the potential to produce carbon dioxide and alcohol. If fermentation begins in a barrel and ends in a sealed bottle then the carbon dioxide can no longer escape. This results in the spontaneous "fixing" of the gas; the amount of gas, and thus the extent to which a wine is sparkling, will depend on the factors cited above.

It so happened that from about the last third of the seventeenth century onwards the wines of Champagne had all the characteristics of wines that were likely to be sparkling. We know that the climate was harsher than it is now; in any case it was colder in Champagne than in the vineyards further to the south. This meant that the grapes were usually picked before reaching full maturity; the wines were therefore acidic and low in alcohol, which is ideal for the fixing of carbon dioxide. At the start of the nineteenth century the chemist Cadet de Vaux wrote that the sparkling and tart white wine of Champagne is made from grapes that are not totally ripe [1]. The white wine that came from black grapes, known then as vin gris, could be kept more easily than other white wines with sugar remaining at the end of fermentation. We have already seen elsewhere that the maceration of grapes in the vat is part of the process of making vin gris, and now that, the fermentation of the must obtained by the pressing of cool grapes is generally less complete than that of a must which ferments in contact with the various parts of the bunch. As cellars were not heated, the cold slowed the action of the yeast. Fermentation that began as soon as the harvest was finished tended to drag on, to become languissante (languid) [2] and then to stop only to start again after the winter in the warmth of the spring. This was known as the montée de la sève (rising of the sap), or the pousse du vin ("new growth" of the wine), which was to be studied by Pasteur.

Once solid bottles and air-tight corks became available in Champagne it would have been quite normal for vins gris that were drawn off and refilled in the spring or summer so that they kept better, to be sparkling. The bubbles came from the carbon dioxide resulting from the conversion of residual sugar by yeasts revitalized by the rise in the ambient temperature, and this effect was possibly boosted by the carbon dioxide produced by the malolactic fermentation that was also stimulated by the warmer temperature. It thus seems likely that sparkling wine did not have an inventor as such, and in fact Edward Hyams declared simply that champagne invented itself. The phenomenon was the logical consequence of new wine-producing techniques for white wine, and the bottling practices of a northern region, the empirical materialisation of an idea that was in the air and which followed from practical observations. It was no doubt widely noticed, and while it was rejected by some, it was warmly welcomed by others, who attempted to make wines that were intentionally sparkling. People were beginning to question the exact causes of the bubbles, and it is to Canon Godinot that we owe the first technical study of the particularities of the sparkling wine of Champagne, published in 1718. Views are strongly divided, he wrote, on the principles of this type of Wine; some believe it is the strength of the ’drogues’ that are added, which make the wine froth so strongly, others attribute the froth to the greenness of the Wines, because most of those which froth are extremely green; finally others attribute it to the effect of the Moon, believing it depends on the time when the Wine is put into flagons [3].

But the sparkling effect was capricious; sometimes it would be absent, or it would be so violent as to break the bottles, or it would be insufficient producing an effect known as sabler (literally "to sand"). In a letter of the 20 December 1736 to Philippe-Valentin Bertin du Rocheret, Abbot Bignon writes of two wines, both of them ’sablant’ perfectly, but neither could be called sparkling (B 21). The verb sabler is thus used for wines which, rather than sparkling, present a few bubbles of carbon dioxide at their surface, as is today the case with some Swiss red wines, which have to faire l’étoile ("make a star") in order to be appreciated by connoisseurs. Great efforts were thus made to obtain a reliable sparkling effect by empirical methods and some Wine Merchants often put in alum, spirits of Wine, pigeon droppings, and many other drogues, that made it froth most extraordinarily. Everyone was in complete agreement on the capital importance of the timing of bottling, which had to take place when the warmth of spring and summer restarted the fermentation, as we have just seen. There is no artifice necessary, wrote Canon Godinot, one is always sure of having a perfectly sparkling wine if it is put in flagons between the 10 and the 14 of the Moon of March. Nicolas Bidet confirms that one usually chose the full moon of March to draw off the wine in bottles [4] and Abbot Rozier specifies at the end of the eighteenth century, Wine bottled in March is more sparkling than that which is bottled in August, and if one waits for the month of the following October or December, then it will not be sparkling [5].

This being established it was now a matter of finding the grapes and the wines most suited to being sparkling. Because it had been noticed that grapes that were not fully mature made better sparkling wines, no others were wanted. In general, wrote Maupin, it is better to use grapes that are a little green rather than ripe. Sometimes this was taken too far, and the wine was acidic and unpleasant; here is what PhilippeValentin Bertin du Rocheret had to say about one such wine, it is as green and hard as an oak, and as dry as the devil (B 21). But the wines destined to be sparkling were chosen from the best vintages, and for the best results the preferred course of action was to blend either the grapes or the wines, as we saw was done from the beginnings of vin gris. Abbot Pluche explains, I have vines of varying quality; if I want to unite the good qualities in one wine, and improve one with the other, I can do it, either at the vine by mixing grapes of different cantons, or in the cellar by mixing the different wines that I will have pressed. It was the grapes that Dom Pérignon mixed in the press, and still in 1780, the treatise already cited of Dudoyer de Vauventrier recalls that the mixing of grapes of different vines in the press in Champagne is a point of fact of public notoriety, adding that at Pierry in the cellars of the bourgeois owners there was no wine that could properly be said to come from Pierry, the wines being made up from grapes from Pierry but also from Moussy, Vinay, Ablois, Cuy, Cramant, Épernay, Ay, Dizy, Hautvillersand others that had been mixed and of which the juices had been extracted together and mixed in the press. Nevertheless it did also happen that it was not until the cellar stage that the musts of wines from different vines were mixed with a view to obtaining and maintaining quality.

Initially the only good quality vins gris available were those of the Marne Valley, it was not until the very early years of the nineteenth that the choicest grapes of the Montagne de Rheims were used for sparkling wines. An important event at the start of the century, forty years ago, wrote Maizière in 1846, was drawing of the fine wines of the Montagne de Rheims for the creation of sparkling wines. But soon the vineyards of the current Côte des Blancs would also be enlisted, as it was noticed that their white grapes produced more "energetic" sparkling wines than those made with the vin gris from black grapes. Bidet specifies the white grapes of Cramant, Avise, le Mesnil, Auger and others as being easier to use for the production of sparkling wines. From this came an idea that gradually spread, which was the blending of vins gris with wines made from white grapes, thus benefiting from the acclaimed qualities of the former and the reliable sparkling properties of the latter. Wines made from the white grapes of the Côte d’Avize, mixed with those of Ay and other first class vineyards, make sparkling wines which combine all their qualities, wrote Jullien. And Chappaz stated in 1951 that effervescence is now more frequent in musts made with white grapes.

However as Armand Maizière said, nature herself had placed the ultimate solution to the issue of making wines sparkle elsewhere than in a simple blend; but it was only in the nineteenth century that technical progress enabled reliably sparkling wines to be made with a degree of control. The question is whether before this sugar was used to boost the levels of carbon dioxide. It was used in England, but there is nothing to indicate that it was used in the eighteenth century in Champagne, or at least not systematically. It is not mentioned in oenology manuals, and there is no sugar listed in the inventories that were drawn up in the event of a sparkling wine producer’s death. The word liqueur is often used in wine-related texts of the period. Some have deduced from this that sugar was added to musts to aid fermentation, but this is to misunderstand the meaning of liqueur. It refers primarily to wine, with which it is synonymous, especially in literature. In Furetière’s Dictionnaire Universel, it is said par excellence of wines, and particularly of those that are the most agreeable. One also describes, on tasting wines, the presence of natural sugar, being a positive feature for some and a flaw for others. Abbot Bignon, in a letter to Philippe-Valentin Bertin du Rocheret, dated the 20 January 1734, tells him that the wines of 1733 resemble those of 1715 on account of a liqueur that we anticipate will last some time, but such a charming flaw comes only from the exceptional maturity of an excellent grape. No doubt it will be considered a flaw in Paris, but it is a quality much sought after in foreign countries (B 21). Whether a wine had too much or not enough liqueur was thus a question of taste, and whenever necessary attempts were made, with varying degrees of success, to remove an excess; with this objective Canon Godinot recommends adding a pint of milk, letting it rest, and then drawing off. It is true that sugar or honey was sometimes added to wine, but it was always in small quantities and only as part of a remedial process to correct a fault in the wine, or in a domestic recipe, such as that which instructs us to infuse a bottle of white wine with half a pound of crystallized sugar and two ’gros’ of cinnamon, together with a full spoonful of elder flowers. Nicolas Bidet advises with simple good sense the addition of some honey eau-de-vie in order to soften a harsh and green wine. But it was only in the nineteenth century that sugar was systematically used in the preparation of sparkling wines, and even then it was for a long time purely guesswork. At this stage it was still a case of trusting to nature, as is suggested by Malavois de la Ganne in 1735 when he wrote, This year has been good for sparkling wines, all those that were drawn had this quality.

Merchants could always sell wines that refused to sparkle as still wines, but it was a last resort, and they would often resign themselves to the considerable expense of attempting to correct the apathy of wines which did not sparkle [6]; this involved bringing the bottles back up from the cellar, pouring the contents into barrels and trying to improve the blend. However in doing this there was a risk of jumping out of the frying pan and into the fire: the newly blended wine could be excessively sparkling, which would then result in bottles exploding. This phenomenon, known as la casse ("breakage"), was the nightmare of every producer of sparkling wines who would obviously have to take it into account when establishing his prices, which were already necessarily higher than all the still wines of Champagne; thus in his journal on 17 October 1747 Philippe-Valentin Bertin du Rocheret noted that concerning M. Motheux: the risk of breakage is his (B 35). Indeed very frequently bottles were unable to withstand the increase in pressure of carbon dioxide during fermentation. Huge numbers exploded, sometimes 50% or more of a batch. Examples abound. Here is one from the records of one of the biggest merchants: In 1746, I drew 6000 bottles of a wine extremely high in liqueur: only 120 bottles remain. In 1747, there was less liqueur: I lost one third in breakages. In 1748, it was higher in alcohol and with less liqueur: I only lost a sixth. Another producer records that on 6 April 1760, 1,100 bottles out of 2,000 had already broken.

In order to try and reduce this danger Canon Godinot sensibly advised leaving a space in the bottles when they were being filled, without this the wine, when it starts working during the various seasons of the year, will break a great quantity of flagons, even so it will break many, despite all the precautions that we take (254). More effective is to find good cellars, as prescribed by Nicolas Bidet: in order to protect against breakage as much as possible, [cellars] should be neither too high nor too deep, neither too warm in winter, nor too cold in the summer. He adds that when one takes the bottles down to the cellar after drawing they should be "entreiller" which means they should be laid down on a trellis of wooden slats. And he advises the construction of basins of cement with vessels to receive the wine if it should spill (51). One can still find cellars today in Champagne with channels for the recuperation of vin de casse. Sometimes it was collected and put in barrels or bottles, but as Cavoleau says, unless it is one of these extraordinary instances when the explosions are furious and the wine flows in streams, it is rare that one can put to good use wines from breakages; the least spoilt is no more than a base wine, at the most likely to be suitable for the production of vinegar.

Lacking real remedies the only recourse was to better bottles, of which Nicolas Bidet recounts the history in 1759:

It used to be that bottles were used indiscriminately, no account being taken of their shape, quality or size; some used flat bottles covered with willow, the glass of which was as thin as a drinking glass, and thus very fragile, and of an indeterminate capacity. Others used round bottles, of which the base was very wide, the body short and the neck much longer than the body, the base very thick, and the body very thin; the least effort from the wine would separate the base from the body of the bottle. Finally bottles were made in the shape of apples, of which the neck sunk into the highest part of the body; this not only resulted in a shape that was unattractive but also disadvantageous, as much for placing them on trellises in the cellars, as for putting them in baskets for deliveries, and for putting in corks. The disadvantages of this bottle becoming apparent, the people of Champagne decided to give to their bottles the shape of a pear.

Producers looked for supplies of bottles in Champagne, where glass factories multiplied, and in Lorraine: The use of round flagons is very common in Champagne, wrote Godinot. As there is a lot of wood in the Province many glass factories have been established, which mostly make nothing but flagons. Glass-workers competed to supply the most robust bottles, but still there were complaints about their lack of solidity. In a report that is kept in the Bibliothèque Nationale (Joly de Fleury Collection, 264), Master Legras, a notable of Rheims, complains of the irreparable damage that the poorly constructed bottles from the Sainte-Menehould glass factory was causing to all of Champagne. There were eleven glass-working ovens in Argonne, and also, significantly, some English glass-workers.

The champagne bottle made its official appearance on 8 March 1735, accompanied by a royal decree, which stipulated that it should in future contain a pint, of Paris measure, and must not weigh less than twenty five ounces. The same decree also makes provision for half and quarter bottles in the same proportion and for double sized bottles and beyond. At around the same time bottles began to be stamped either with the producers’ arms, as at the Abbey of Saint-Basle in Verzy, or with those of the customer. Godinot states that there are Lords who have their own bottles made . This custom had existed in England since the middle of the seventeenth century, in order to help purchasers using their own bottles to identify them more easily when they were filled from the barrel at the wine merchant. As we have already seen the word flacon (flagon) rather than bottle was used in Champagne, and appears in the Journal des Sçavans of the 7 June 1706. Elsewhere at that time in France a flacon was used to mean a large bottle with a screw top (Dictionnaire de l’Académie, 1694 edition). The word thus had a particular meaning in Champagne, those which you call ’flacons’ in your Champagne, wrote Abbot Bignon on 2 March 1741 to Philippe-Valentin Bertin du Rocheret (B21). The reason was probably that through the system of tying up the corks in use in Champagne they were probably as secure as if they had been screwed in. In his manuscript Malavois de la Ganne uses flacon up to 1735 after which he uses bouteille and flacon interchangeably.

The word carafon was also used throughout the kingdom, but comparatively rarely as for example in this text of 1724 that mentions the high consumption in recent years of bottles made of strong glass, commonly known as carafons. Here again there was an evolution in glass-making terminology, for we read in the Dictionnaire de l’Académie, in the 1690 edition: Caraffon - Large bottle made of thick glass with a long neck, used to chill drinks in an ice-bucket. It was probably due to the similarity in shape that the bottles in Champagne, which also had long necks, were sometimes referred to as caraffons or carafons.

Good quality bottles needed good quality stoppers. Only cork could withstand the high pressure inside the bottles. Spain, who was the main producer, had a monopoly on the manufacture of the corks used for bottling sparkling wine in Champagne until the cork makers established themselves in the region in the 1740s. One cannot take too much care in choosing them, wrote Canon Godinot, wine can spoil in certain bottles because of defective corks.

According to Nicolas Bidet they should be an inch and a half in length and, to ensure air-tightness, the cooper drawing the wine off into bottles attaches the cork with string knotted in a cross, after which, turning the bottle, he dips the end into melted wax mixture up to the ring, so that the string is tarred, but others are content to soak their ball of string in linseed or nut oil, it then becomes as hard as catgut and lasts several years in the cellar without rotting.

The aforementioned royal decree of 1735 stipulates that the corks must be attached with a string with three threads, tightly twisted and knotted in a cross on the cork. From 1760 onwards wire made of iron or brass gradually overtook string. Note that the use of string was not reserved exclusively for sparkling wines and was also used for some still wines. Abbot Pluche states that one can seal corks if one wants, to prevent "misunderstandings and infidelities".

In addition to breakages there was another disagreeable phenomenon to upset the producer of sparkling wine, which was the presence of a deposit, an unwelcome guest that could cloud the wine. It was yet known that it was formed by the multiplication of yeast cells during fermentation and there was no way known of getting rid of it before the wine was sold without losing the precious carbon dioxide. However the deposit was less then than it is now, because before the wine was bottled, having been drawn off, there was no sugar or yeast added. Sometimes a decanting process was undertaken which consisted of changing the bottle and hopefully leaving behind the deposit, but since this was suspended in the wine the operation must have been far from satisfactory. Pressure was lost and refilling was difficult. If the deposit was very light then it would be carried into the new bottle with the wine; if the pressure was too high when the bottles were opened then the gas would expand with such force that the deposit would be disturbed and dispersed throughout the wine (316). At the end of the eighteenth century planks with holes in them began to be used into which bottles would be placed neck downwards; some of the deposit would then, by gravity, settle on the cork but the rest would remain stuck to the sides of the bottle. René Gandilhon states that the only inventory following death in which he found planks with holes for standing up wine is only dated 1784 (242). There is no documentation to show that a satisfactory solution was found to the problem of deposits before the nineteenth century.

The unpredictability of the behaviour of sparkling wines led to champagnes with varying degrees of pressure inside the bottle being offered on the market, and these were given different appellations There was thus mousseux (sparkling), which had existed since the beginning, this was the most common and was also known as pétillant. and then from 1729 there was grand mousseux, and from 1736 demi-mousseux. The grand mousseux, sometimes called sautebouchon ("cork-popper"), or even sauteur ("jumper"), contained more pressure than the mousseux; it was estimated to be three atmospheres. This is very low compared to the five or six atmospheres of the twentieth century; there is no doubt that initially champagne did not sparkle as much as it does now. For proof of this one need only examine Lancret’s picture, Déjeuner de jambon, in which champagne is poured from a great height into small flutes without it foaming over the sides of the glass, a feat that would today be inconceivable. When the cork comes flying out up to the ceiling, and the wine spurts out like a jet of water, half-emptying the flagon, this must be considered the exception which proves the rule. The demi-mousseux would present a light foam that would delicately whiten the glass and then quickly disappear. Such wines were said to crème ("to cream"). There was also the ptysanne or tisane de Champagne, of modest quality, which could be lightly foaming or with no effervescence at all; according to Chaptal, the vin de tisane was made with the second "cut" of the grapes.