Memoire sur la fermentation alcoolique
(Memoir on Alcoholic Fermentation)
by Louis Pasteur, published 1860, here Google-translated to English. The original text is available at https://archive.org/details/s696id13664740/page/323/mode/1up.
Contents:
PART 1: WHAT HAPPENS TO SUGAR IN ALCOHOLIC FERMENTATION.
§ 1. — Historical Overview of the Current State of Science Regarding the Products of Alcoholic Fermentation.
[The rest of the sections of Part 1 are not included in this translation.]
PART TWO: WHAT HAPPENS TO BREWER'S YEAST IN ALCOHOLIC FERMENTATION.
§ 1. — History of the current state of science on brewer's yeast and its modifications during alcoholic fermentation.
§ II. — The nitrogen in yeast is never converted into ammonia during alcoholic fermentation. Far from ammonia being formed, any ammonia that is added may even disappear.
§ III. — Production of yeast in a medium made up of sugar, an ammonia salt and phosphates.
§ IV. — Study of the relationship between yeast and sugar. What happens to the nitrogen in yeast during alcoholic fermentation.
§ V. — In alcoholic fermentation, part of the sugar is fixed on the yeast in the form of cellulose.
§ VI. — In all alcoholic fermentation, a portion of the sugar is fixed in yeast in the form of fats.
§ VII. — Permanent vitality of yeast globules.
§ VIII. — Applications of some of the results of this Memoir to the composition of fermented liquids. Specific studies on wine.
Memoir on Alcoholic Fermentation
Introduction
I call alcoholic fermentation the fermentation that sugar undergoes under the influence of the ferment known as brewer's yeast. It is this fermentation that produces wine and all alcoholic beverages. It is also the model for a multitude of similar phenomena that are designated, according to general usage, by the generic term "fermentation" followed by the name of one of the essential products of the particular phenomenon under consideration.
From this convention regarding the adopted nomenclature, it follows that the expression "alcoholic fermentation" cannot designate every fermentation phenomenon in which alcohol is produced; for there can be various kinds sharing this common characteristic.
If we could not agree beforehand on which of these very distinct phenomena should bear the name alcoholic fermentation, to the exclusion of the others, we would inevitably give rise to a confusion of language that would quickly shift from words to ideas, and would sow discord in studies already obscure enough that we must scrupulously avoid artificial complications.
Moreover, any hesitation regarding the words alcoholic fermentation and their true meaning seemed impossible to me, since they were applied by Lavoisier, Gay-Lussac, and Thénard to the fermentation of sugar by brewer's yeast. It would be pointless to fail to follow the example of these illustrious masters who laid the foundations of our initial knowledge on this subject.
In the first part of my work, I study what happens to sugar through alcoholic fermentation, and in the second, I deal more specifically with the ferment, its nature, and the transformations it undergoes. In order to better highlight the progress resulting from my research, I have prefaced each part with a historical summary of the state of science at the time I began to study alcoholic fermentation.
PART ONE.
WHAT HAPPENS TO SUGAR IN ALCOHOLIC FERMENTATION.
§ 1. — Historical Overview of the Current State of Science Regarding the Products of Alcoholic Fermentation.
Lavoisier was the first to present the most judicious views on the products of alcoholic fermentation. The Memoir included in his Elements of Chemistry on this subject is singularly curious. Excessively flawed in its numerical determinations, it is admirable if considered from the point of view of general and philosophical ideas. It is there that one finds these beautiful words: "Nothing is created either in the operations of art or in those of nature, and one can posit this principle, that in every operation there is an equal quantity of matter before and after the operation, that the quality and quantity of the principles are the same, and that there are only changes, transformations. It is on this principle that the entire art of conducting experiments in chemistry is founded."
In any case, the operations reported by Lavoisier only confirm his preconceived views as a result of compensating for considerable errors. Without going into detail, I will simply say that Lavoisier starts from the following composition of sugar (1):
< Footnote: (1) Lavoisier, Eléments de Chimie, Volume I, page 139, 2nd edition. End footnote >
Hydrogen: 8
Oxygen: 64
Carbon: 28
[Total: 100]
where the carbon is off by 14 percent. A few pages later, Lavoisier provides a complete table of his numerical results, in which, as he points out, the sum of the weights of alcohol and carbonic acid produced during fermentation is almost equal to the weight of the sugar that fermented, and the equation holds true for each of the separate elements. The under-error for carbonic acid is very large and is compensated by an equivalent over-error for alcohol.
But, in any case, and within the limits of accuracy of the analyses of the time on these substances, all of Lavoisier's contemporaries must have thought that sugar, under the influence of brewer's yeast, split into alcohol and carbonic acid. Lavoisier summarized the conclusions of his research as follows: "The effects of wine fermentation thus amount to separating sugar, which is an oxide, into two parts; oxygenating one at the expense of the other to form carbonic acid; and deoxygenating the other in favor of the first to form a combustible substance, which is alcohol; so that if it were possible to recombine these two substances, alcohol and carbonic acid, sugar would be reformed."
Twenty years later, the analyses of Gay-Lussac and Thénard, and those of de Saussure, definitively established the composition of sugar and alcohol. If the conclusions of Lavoisier's work had not been consistent with these new analyses, there is no doubt that they would have been revised by experiment, and that the great inaccuracy of Lavoisier's measurements would have been even more apparent; but on the contrary, it then became easy to demonstrate theoretically that by adding alcohol and carbonic acid, the composition of sugar could be reproduced. This is what Gay-Lussac soon pointed out. In a letter to M. Clément, published in the Annales de Chimie for 1815, Gay-Lussac, after discussing recent analyses of [ethylene], alcohol, ether, and sugar, arrives at this theoretical conclusion:
“If we now assume that the products provided by the ferment can be neglected relative to alcohol and carbonic acid, which are the only perceptible results of fermentation, we will find that, given 100 parts of sugar, 51.34 are converted into alcohol and 48.66 into carbonic acid during fermentation” (1).
< Footnote: (1) Gay-Lussac, Annales de Chimie, vol. XCV, p. 318. End footnote >
These few lines by Gay-Lussac established the opinion of chemists. However, they contained an error which was very judiciously brought to light by Messrs. Dumas and Boullay in 1828. These skilled chemists showed that Gay-Lussac's numbers are only true for sugars with the formula C12H12O12 whereas Gay-Lussac applied them to cane sugar. From this came the consequence, very well indicated by Messrs. Dumas and Boullay, that cane sugar can ferment without assimilating the elements of a water molecule (2).
< Footnote: (2) Dumas and Boulay, Annales de Chimie et de Physique, volume XXXVII, page 45; 1828.
“The theory of fermentation established by Gay-Lussac thus leaves something to be desired; but this is no longer the case as soon as ether is substituted for alcohol in the theoretical composition of sugar. The most perfect agreement is then restored between theory and experience...”
“Grape and starch sugars appear to differ mainly from cane sugar in that they are composed in such a way that they can actually be represented by carbonic acid and alcohol...” End footnote >
Two years later, Mr. Dubrunfaut went further and experimentally recognized that cane sugar, before fermenting, was transformed into uncrystallizable sugar. In any case, it is easy to recognize from reading contemporary works that, from the time of Gay-Lussac's letter onward, there were no longer any doubts; and, when the use of chemical formulas had definitively become established in science, everyone expressed the alcoholic fermentation of sugars under the influence of brewer's yeast by the equation: C12H12O12 = 2 (C4H6O2) + 4 CO2.
But let it be clearly noted that Gay-Lussac did not conduct any experiments; and in all of this, what I see as most real is the illusion produced by the possible equation between sugar on the one hand, and alcohol and carbonic acid on the other. I wanted to show through these historical details that the study of alcoholic fermentation, despite the importance of the subject, had not yet been established on a sound scientific basis and that the generally accepted equation was merely the expression of a theory unsupported by any precise measurements.
Lavoisier's work contains a valuable result concerning the formation of a small quantity of an organic acid during alcoholic fermentation, a fact confirmed by Mr. Thénard and by all the observers who have studied this fermentation. The nature of this acid is poorly understood. Lavoisier states that it is acetic acid, while modern authors assert that it is lactic acid. No sustained work on this point has yet been found in published literature. The assertion concerning lactic acid probably entered the scientific community when it was learned that Mr. Dubrunfaut had obtained a large quantity of lactic acid in certain alcoholic fermentations.
The following propositions therefore summarize current knowledge on the products of sugar transformation in alcoholic fermentation:
1. Cane sugar, C12H11O11, after being transformed into the sugar of acidic fruits, with the composition C12H12O12, ferments and separates into alcohol and carbonic acid. The sum of the weights of the alcohol and carbonic acid represents almost exactly the weight of the sugar.
2. A very small quantity of sugar is transformed isomerically into lactic acid of the same composition as sugar.
I will show shortly that the first proposition is never exact, that it is only a rather crude approximation of the truth, and as for the second, it is entirely erroneous with regard to the nature of the acid in fermentation, which is in no case lactic acid, unless the fermentation is fortuitously complicated by a completely different fermentation, namely lactic acid fermentation.
I will now present successively, with all the appropriate details, the new results of my research. I present them with confidence to chemists, because I have devoted meticulous care to their study and, above all, because, having repeated my experiments many times under varied conditions, I believe I have succeeded in distinguishing the general laws of the phenomena by disentangling them from the accidental complications that have cast much obscurity over the history of alcoholic fermentation.
PART TWO.
WHAT HAPPENS TO BREWER'S YEAST IN ALCOHOLIC FERMENTATION.
§ 1. — History of the current state of science on brewer's yeast and its modifications during alcoholic fermentation.
In 1680, Lewenhoek studied brewer's yeast under a microscope and found it to be composed of very small spherical or ovoid globules. However, the chemical nature of this substance was unknown. In a memoir on fermentation, awarded a prize in 1787 by the Academy of Florence and read to the Philomathic Society of Paris in 1799, Fabroni, an Italian scholar, amidst many erroneous views and facts, compared and identified yeast with gluten. This was progress. It gave yeast a place among organic products. It was to assimilate it to the substances then called animal, that is to say, those that provide ammonia upon distillation (1).
< Footnote: (1) “Fermentation is simply the decomposition of one substance by another, such as that of a carbonate by an acid or of sugar by nitric acid. The substance that decomposes sugar is the animal-gelatinous substance; it resides in specific utricles in grapes as well as in wheat. When pressing grapes, this glutenous substance is mixed with sugar, as if an acid and a carbonate were being poured into a vessel; as soon as the two substances come into contact, effervescence or fermentation begins, as occurs in any other chemical operation.” – Fabroni.
See Fourcroy's critical summary of Fabroni's work, Annales de Chimie, Volume XXXI; 1799. End footnote >
This assertion by Fabroni brought the question of the nature of ferment to the forefront. It was believed that Lavoisier had solved the difficulties of fermentation with regard to fermentable matter, but no one had any idea about the nature of the substance that caused the sugar to split.
Thus, in Year VIII [referring to year 8 in the french revolutionary calendar, corresponding to the period 23. Sep. 1799 to 22. Sep. 1800 in the Gregorian calendar, see https://frcal.qt.ax/9/1 ], a year after the publication in France of Fabroni's work, the Physical and Mathematical Sciences class of the Institute proposed the following question as a prize topic: "What are the characteristics that distinguish, in vegetable and animal matter, those materials used as fermenting agents from those to which they subject fermentation?"
Encouraged by this proposal from the Institute, Mr. Thénard attempted to solve the problem, and in Year XI [23. Sep. 1802 to 23. Sep. 1803], he published a remarkable paper in which he dealt primarily with the nature of ferment, its origin, and its alteration during the fermentation process. Here is a summary:
All natural sugary juices, when left to ferment spontaneously, produce a deposit resembling brewer's yeast and, like it, possessing the ability to ferment pure sugar water. This yeast is of an animal nature, meaning it is nitrogenous and produces a significant amount of ammonia during distillation. During fermentation, the yeast gradually loses its nitrogen and some of it is destroyed, transforming into soluble products (1).
< Footnote: (1) Thenard, Annales de Chimie, vol. XLVI, p. 294; 1803. End footnote >
By reducing the quantities in the single experiment reported by Mr. Thénard to 100 parts, 20 parts of fresh brewer's yeast and 100 parts of sugar, after complete fermentation, left 13.7 parts of an insoluble residue, still active, which, exhausted by contact with a new quantity of sugar, was reduced to 10 parts.
“This last residue was white, exhibited all the properties of wood, and exerted no action on a new quantity of sugar water.”
These results, to which I will return shortly, have been cited in all the works and have served as the basis for all discussions on the theory of fermentation. I must add immediately that there is, in Mr. Thénard's Memoir and even in the latest edition of his Treatise on Chemistry, a curious passage on the nitrogen in the ferment.
“New research,” he said, “worthy of the full attention of chemists, must be carried out on the decomposition that the ferment undergoes. It will be necessary to see what becomes of the nitrogen in the decomposed ferment. It is not found mixed with carbon dioxide; it does not enter into the composition of the insoluble white matter; it is not part of a very small quantity of highly soluble matter found in the liquid with the alcohol. Alcohol does not contain any; so the question of what becomes of the nitrogen in the ferment is still to be resolved (2).”
< Footnote: (2) Thenard, Treatise on Chemistry, 6th edition, volume V, page 65. End footnote >
Although this passage, as I shall show, contains several errors, it testifies to Mr. Thénard's preoccupation with the nitrogen in ferment, and moreover, to the fact that this eminent chemist, who had made a particular study of the subject, did not accept the opinions of authors on the transformation of nitrogen in yeast into ammonia. It was, it seems, Döbereiner who first announced that the nitrogen in yeast was present in the liquor in the form of ammonium salt, an assertion that was accepted by all chemists. It gradually made its way into elementary works, with the exception of Thénard's.
A few years after the publication of Thénard's Memoir, Gay-Lussac made public a very extraordinary result. Examining Mr. Appert's methods for preserving plant and animal substances, he noticed that grape must that had been preserved without alteration for a whole year began to ferment a few days after being transferred. This fact, attributed to Appert, led Gay-Lussac to the experiments that everyone knows, and from which it follows “that oxygen is necessary to begin fermentation, but not to continue it” (1).
< Footnote: (1) Gay-Lussac, Annales de Chimie, vol. LXXVI, p. 247, 1810. End footnote >
To find a new advance worthy of mention in the history of fermentation, after the preceding work and that of Mr. Colin, which I have already mentioned, one must go back to 1835 and 1837, to Mr. Cagniard de Latour (2). Taking up the old, very incomplete microscopic observations of Lewenhoek, which he was, moreover, unaware of, Mr. Cagniard de Latour introduced a new idea into the studies we are discussing. Before him, yeast had been considered an immediate substance of plants, which had the property of precipitating in the presence of fermentable sugars. Mr. Cagniard de Latour recognized “that yeast was a mass of globules capable of reproducing by budding, and not simply organic or chemical matter, as was supposed.”
< Footnote: (2) It would be unfair not to mention, in a history of fermentation and even in connection with Mr. Cagniard de Latour’s observations, the earlier microscopic research conducted by Mr. Desmazières and published in 1825, ten years before Mr. Cagniard de Latour’s work, in the Annales des sciences naturelles. It is true that Mr. Desmazières did not study yeast itself, but he deserves credit for having studied under the microscope and clearly described the composition of the film that forms on the surface of beer and which Persoon, in 1822, had called Mycoderma cervisiae.
Mr. Desmazières' work must have paved the way for the microscopic study of brewer's yeast and helped to better see and understand its structure. At that time, microscopic research was much more difficult than it is today, and Mr. Cagniard de Latour himself reports that in 1820, having used a very imperfect microscope, he had believed that yeast was like very fine sand composed of crystalloid grains (Memoir cited, page 208).
Mr. Desmazières recognized that the film in question was formed of a multitude of hyaline, ovoid capsules, which, according to him, can fuse end to end to form more or less branched tubes, etc. He further recognized that these globules possess particular movements; he is convinced of their animal life and classifies them among the infusoria.
It is evident that Mr. Desmazières confused Brownian motion with actual vital movement. R. Brown's research was not published until 1828. (Annals of Natural Sciences, 1st series, volume XIV, page 341, and Additional Remarks, volume XIX, page 104.) It is in his Additional Remarks that R. Brown corrects his initial impressions and settles on the following opinion:
“The extremely delicate particles of solid matter, whether obtained from organic or inorganic substances, when suspended in water or some other aqueous fluid, exhibit movements which, according to their irregularity and apparent independence, resemble to a remarkable degree the slowest movements of some of the simplest infusorial animalcules.”
Such is the movement exhibited by yeast globules, or the smaller ones that sometimes accompany them. It was these movements that had misled Mr. Desmazières and several naturalists before him. Mr. R. Brown himself revisited these observations twice to fully understand them. His initial impressions had been completely wrong. End footnote >
From his observations, Cagniard de Latour concluded “that it is very probably through some effect of their growth that the yeast globules release carbonic acid from the sugar liquid and convert it into a liquor (1).”
< Footnote: (1) Cagniard de Latour, Annales de Chimie et de Physique, 2nd series; vol. LXVIII. End footnote >
This opinion immediately found a powerful opponent in Mr. Liebig. In his view, ferment is an extremely perishable substance, which decomposes and stimulates fermentation as a result of the alteration it undergoes, by disrupting and disassembling the molecular group of the fermentable matter. After recalling the properties of yeast, he states:
“The facts we have just presented demonstrate the existence of a new cause that engenders decompositions and combinations. This cause is nothing other than the movement that a decomposing body communicates to other substances in which the elements are held together with a very weak affinity. Brewer's yeast, and in general all putrefying animal and vegetable matter, transfers to other bodies the state of decomposition in which they themselves find themselves; the movement which, by disturbing the equilibrium, is imparted to their own elements is also communicated to the elements of the bodies which are in contact with them” (1).
< Footnote: (1) Liebig, Annales de Chimie et de Physique, 2nd series, volume LXXI, page 178.
Liebig, Lettres sur la Chimie, French translation, letters 16 and 28.
Liebig, Traité de Chimie organique. Introduction, page 29.
Note on the Memoir of Mr. Cagniard de Latour and Gay-Lussac's observation on the influence of oxygen in fermentation. When Mr. Cagniard de Latour made his initial observations known, some people, as happens all too often, tried to diminish their merit and spread the rumor that these results had already been published in Germany by Dr. Schwann. It is indeed true that Messrs. Cagniard de Latour and Schwann, each working independently, arrived at almost the same conclusions regarding the nature of yeast around the same time. But it is no less certain that the priority of publication belongs to Mr. Cagniard de Latour. Here is a passage from Mr. Schwann's Memoir, published in Poggendorff's Annals, Volume XLI, 1837:
“This dissertation is a verbatim reproduction of the one read in the first days of February of this year (1837) on my behalf by Professor Muller at the meeting of the Society of Friends of Natural Sciences in Jena. Soon after, I received the journal of the Institute dated November 23, 1836, in which I saw that Mr. Cagniard de Latour had made similar observations on the fermentation of beer, observations which had remained unknown to me until then.”
But Schwann's work contains valuable observations that shed much light on the origin of spontaneous fermentations and that allow for a different interpretation of Appert's experiments than Gay-Lussac did. End footnote >
Recent historical studies published by Mr. Chevreul in the Journal des Savants, February 1856 issue, reveal that Stahl had already expressed ideas similar to those of Mr. Liebig on the causes of fermentation.
Mr. Liebig developed his opinions in most of his works with a persistence and conviction that gradually led to their acceptance. Today, they are generally accepted in Germany and France. Messrs. Fremy and Boutron applied them to lactic fermentation with a modification that has been generally adopted. The dominant idea of their work is this: In substances capable of acting as ferments, the fermentative character varies with the degree of decomposition of the substance. At various stages of its decomposition, it is an alcoholic ferment, a lactic ferment, etc., depending on the more or less advanced stage of its decomposition.
Mr. Liebig's ideas were also developed and supported in the excellent work that Mr. Gerhardt left upon his death (1).
< Footnote: (1) Gerhardt, Volume IV, page 537. End footnote >
In my opinion, this is the main reason for the gradual success that Mr. Liebig's ideas have acquired among chemists. For twenty years, a great number of phenomena have been discovered that are classified alongside alcoholic fermentation proper, and in which it has seemed impossible to recognize the existence of particular cryptogamic plants, but in all of them there was matter, once alive, in the process of decomposition.
And, for example, if one places dissolved sugar mixed with chalk with any nitrogenous animal matter—caseum, fibrin, gelatin, rennet, an animal membrane, etc.—one sees the sugar gradually become lactic acid. Now, these animal substances are of very diverse structure, nature, and form, and the final effect on sugar is the same. Only one thing appears to be similar in these nitrogenous substances: their gradual decomposition. The correlation is thus evident between the transformation of sugar into lactic acid and a process of decomposition of a nitrogenous substance.
Research by Mr. Colin on alcoholic fermentation, dating from 1825, had already established similar facts. This chemist had recognized that a wide variety of animal substances could cause sugar to split into alcohol and carbonic acid.
However, a remarkable circumstance should have aroused attention and prompted caution, at least concerning alcoholic fermentation. Indeed, after the publication of Mr. Cagniard de Latour's observations, Mr. Turpin, who had been tasked with reporting them to the Academy, studied, at Mr. Thenard's request, the deposit that forms during the alcoholic fermentation of sugar by decoction of egg white, and found that it consisted solely of brewer's yeast globules. If one of the substances used by Mr. Colin, albumin [egg white], only caused alcoholic fermentation by producing yeast, it was presumed that all other nitrogenous substances behaved similarly, and therefore their diversity proved nothing regarding Mr. Liebig's theory.
But I hasten to add that nothing of the sort existed, or so it was believed, in the very diverse and numerous cases of lactic fermentation. All observers agree that there is only a chemical decomposition of the animal matter. The facts relating to this fermentation and several other phenomena of the same kind therefore had a decisive influence on the theory.
Thus, the idea of Mr. Cagniard de Latour, which had initially enjoyed some credibility, was gradually abandoned. Many people, at least, did not dispute that brewer's yeast was organized, but they believed that it was partially decomposed by fermentation, as Mr. Thénard had stated, and that, resembling all other nitrogenous substances acting as fermenting agents, it was to this property that it owed its action on sugar. This was Mr. Liebig's view.
Berzelius did not share Mr. Liebig's ideas, while also rejecting those of Messrs. Cagniard de Latour and Schwann. For him, fermentation was a contact action. He did not even believe in the existence of a living organism in yeast. “This was merely a chemical product that precipitated during beer fermentation and took the ordinary form of non-crystalline, even inorganic, precipitates, of small balls that group together one after another in the form of a chain of beads (1).”
Elsewhere he expresses himself thus: “It is clear that when products of organized bodies decompose in water and dissolved matter precipitates, the latter must take on a form, and that, since many do not take on regular geometric forms, other forms dependent on the nature of these bodies must result, which also influence the determination of these forms in living nature, hence it is quite natural that they imitate the simplest forms of the productions of plant life. However, form alone does not yet constitute life.”
< Footnote: (1) Berzelius, Comptes rendus des travaux de Chimie, page 277; 1843.
Berzelius, Comptes rendus des travaux de Chimie; 1845.
In his important writings on fermentation, Mr. Mitscherlich did not hesitate to admit that yeast is indeed organized. Nevertheless, he shared Berzelius’s views on the way yeast interacts with sugar. (Annales de Chimie et de Physique, 3rd series, volume VII, page 31.) Mr. Mitscherlich states:
“The ferment globules therefore behave with respect to sugar, or with respect to sugar and water, which contain the elements of carbonic acid and alcohol, exactly like platinum sponge with respect to hydrogen peroxide.”
This opinion was more recently expressed by Mr. Berthelot, who, moreover, it seems to me, somewhat associated it with the ideas of Mr. Liebig. I show in the second part of my work that the facts on which Mr. Mitscherlich relies are entirely corroborated. End footnote >
Let us now say a few words about the chemical composition of yeast. In the excellent collection of works he published on plant development, Mr. Payen gave the following immediate composition of yeast (1):
Nitrogenous matter: 62.73
Cellulose envelopes: 29.37
Fatty substances: 2.10
Mineral matter: 5.80
[Total] 100.00
< Footnote: (1) Payen, 1839. See Mémoires des Savants étrangers; volume IX, page 32, 1846. End footnote >
Elemental analysis yielded, as one might expect, results that varied depending on the observer, the washing and purification methods used, and the origin of the yeast submitted for analysis.
Dumas (Traité de Chimie):
Carbon: 50.6
Hydrogen: 7.3
Nitrogen: 15.0
Oxygen + Sulfur + Phosphorus: 27.1
[Total] 100.00
Mitscherlich (Eléments de Chimie):
Carbon: 47.0
Hydrogen: 6.6
Nitrogen: 10.0
Sulfur: 0.6
A little phosphorus
Oxygen: 35.8
[Total] 100.00
(1) Mitscherlich, “Comptes rendus de l’Académie de Berlin” and “Annales allemandes de Chimie et de Pharmacie, volume LVI”.
[Schlossberger's data, page 375]
(2) Schlossberger, Annales allemandes de Chimie et de Pharmacie, vol. LI; 1845.
In the analyses of Messrs. Dumas and Schlossberger, the ash content is deducted. Mr. Schlossberger's work on the composition of yeast is very thorough and forms an indispensable complement to Mr. Payen's analysis. By using potassium hydroxide to dissolve the nitrogenous matter in yeast, he manages to separate it almost entirely and thus isolate the cellulose, which he transforms with acids into fermentable sugar. Then, precipitating the matter removed by the potassium hydroxide with acids, he demonstrates through analysis that it is much closer to the average composition of so-called albuminoid substances than yeast; that it is richer than yeast in carbon, nitrogen, and hydrogen. I have verified the accuracy of Mr. Schlossberger's results.
Mr. Mitscherlich has provided very good analyses of yeast ash:
[Mitscherlich's data, page 376]
It is worth noting the excellent method Mr. Mitscherlich employed for burning yeast. The material was placed in a silver dish, which was then inserted into a glass tube and heated in a stream of oxygen. However, since silver oxidizes upon contact with glass, the silver dish rested on another dish made of platinum. As long as distillation products were being released, carbon dioxide gas was passed through, and finally, the oxygen stream was used.
I will now present the results of my own research on the nature of yeast and the transformations it undergoes during alcoholic fermentation.
§ II. — The nitrogen in yeast is never converted into ammonia during alcoholic fermentation. Far from ammonia being formed, any ammonia that is added may even disappear.
I recalled Mr. Thénard's observations on yeast. Among other results, he acknowledged:
1. That some of the yeast disappeared during fermentation, meaning that Mr. Thénard collected less yeast after fermentation than he had used. This fact is accurate under the conditions of his experiment, but its significance will only be understood and appreciated with the explanations I will provide shortly.
2. Mr. Thénard found no nitrogen in the yeast after fermentation; this is a serious error, and one might rightly be surprised that it would have escaped an experimenter of Mr. Thénard's skill. But one must remember the state of science at the time Mr. Thénard published his research. So, the presence or absence of nitrogen in organic matter was determined by distilling the substance, and the existence of this element was confirmed when distillation yielded ammonia. Exhausted yeast yielded none. The residue of the evaporated fermented liquid yielded no more, and Mr. Thénard, quite surprised by these results, wondered where the nitrogen in the yeast could be. He looked for it in the released carbonic acid; but, as Lavoisier had already observed, this gas is completely absorbable by potash. He looked for it in alcohol without any more success. What had become of it? And according to a passage I quoted earlier, these preoccupations of Mr. Thénard persisted even in the later editions of his Traité de Chimie.
The explanation seems quite simple to me. We will see in a later paragraph, through a detailed examination of the solid yeast and its soluble components after fermentation, that the elements of sugar bind in significant proportions to the yeast and the soluble components of the fermented liquid residue. Furthermore, we know that the extract of the fermented liquid contains glycerin and succinic acid. It is very likely that in Mr. Thénard's experiments, the acids formed by the glycerin, and the other substances derived from the sugar, during the distillation itself, saturated the ammonia and masked its presence.
Mr. Thénard came very close to solving his predicament. For he adds a sentence that betrayed his doubts and indicated how they could have been resolved. “However,” he adds, “I have conducted several other experiments which, so far, tend to prove that nitrogen can exist in a substance without being detectable by distillation; that consequently, it may be a constituent substance of plants, although they do not generally yield ammonia by distillation. But I have not yet repeated these experiments, and one cannot be too cautious in announcing their results.”
Döbereiner dismissed, but through inaccurate experiments, the difficulties raised by Mr. Thénard's paper. He announced that this chemist had erred on one point and that the soluble residue of the fermented liquid contained the nitrogen of the yeast in the form of ammonia. From this observation, all subsequent works repeated ad nauseam that the nitrogen of the yeast gradually disappeared during fermentation, that the yeast, as Mr. Thénard had stated, became similar to wood, and that the altered and soluble portion contained all the nitrogen in the form of ammonium salts, from which alkalis could easily separate it.
Döbereiner's results are erroneous. The soluble residue of the fermented liquid does not contain the slightest amount of ammonia originating from the nitrogen of the yeast. The nitrogen it contains is entirely in the form of nitrogenous organic matter, highly susceptible to alteration by alkalis, and even releasing ammonia when exposed to potash, lime, and baryte at room temperature. This is what misled Döbereiner.
My own experiments on the nitrogen in ferments only achieved the desired precision once I became aware of the excellent methods for determining ammonia described by Mr. Boussingault in his Memoirs and in his remarkable lectures at the Conservatoire des Arts et Métiers, where I was able to witness every detail of the procedures. Mr. Boussingault's most valuable information for the research I had undertaken is the following: calcined magnesium converts ammonium salts to free ammonia at boiling point, and does not release ammonia from the most nitrogenous organic matter that is altered by lime, potash or baryte. Under the conditions of my studies, this result is of irreproachable accuracy.
That being said, here are some analyses of the fermentation liquids. They clearly show that not the slightest amount of ammonia is formed during the action of the yeast. On January 18, 1858, I fermented 100 grams of sugar with one liter of water containing the soluble substances of brewer's yeast and a barely perceptible quantity of fresh yeast globules. A separate titration of another liter of this liquid revealed that it contained 0.038 grams of ammonia. On February 5, the fermentation was complete. The ammonia was titrated by testing the entire liquid; only 0.020 grams of ammonia remained, that is to say, less ammonia than at the beginning.
On April 30, I also fermented 100 grams of sugar, but this time with ordinary yeast used in very small quantities. I added only 1.037 grams of yeast (weight of dried matter at 100 degrees). On August 30, the fermentation was still noticeable. A sampling tube adapted to the fermentation flask was always immersed in water. The fermented liquid was analyzed on November 27. It was found that the entire liquid contained only 0.0008 grams of ammonia, and it is very probable that there was none at all, and that this minimal quantity of ammonia resulted from a dosage error or a weak reaction of magnesium with the albuminoid matter of the fermented liquid.
10 grams of sugar were fermented with 200 cubic centimeters of clear yeast water containing 0.0075 grams of ammonia and a trace of fresh yeast. Six days later, by testing the entire liquid, it was found to contain 0.0005 grams of ammonia. The consistency of the results of these experiments and many similar ones leaves no doubt as to the main fact. Not the slightest amount of ammonia is formed during alcoholic fermentation at the expense of the yeast. But these results go further: they indicate the disappearance of a portion of the ammonia present in the initial liquid. This latter circumstance led me to add ammonia directly in order to better study the phenomenon.
The following experiments were carried out in this direction: 100 grams of sugar was fermented with 10 grams of washed yeast paste and 0.200 grams of ammonium L-tartrate, containing 0.0185 grams of ammonia. The fermentation lasts a very long time. When all the sugar [...] has disappeared, the liquid is analyzed. It contains only 0.0015 g of ammonia. I also found free tartaric acid in the solution with all its characteristics. Thus, almost all the ammonia added in the form of ammonium tartrate has disappeared, as has that which was present in the 10 grams of yeast.
In the following experiment, I used ammonium D-tartrate:
19.575 g of sugar is fermented with
200.00 g of water,
0.523 g of yeast, containing 0.179 g of dry matter.
I also add 0.475 g of ammonium D-tartrate, containing 0.088 g of ammonia. One month later, the fermentation is complete. 0.071 g of ammonia remains; 0.017 g has therefore disappeared, that is to say one fifth of the total quantity used.
In summary, we see that, far from ammonia forming during alcoholic fermentation, the ammonia added can disappear, especially in cases where there is an insufficient amount of soluble albuminoid compounds due to the use of a small quantity of brewer's yeast. The studies in the following section will show us that the ammonia that disappears in this way becomes part of the yeast's composition in the form of albuminoid matter.
§ III. — Production of yeast in a medium made up of sugar, an ammonia salt and phosphates.
Unprecedented experiments conducted at the outset of my research on gaseous fermentation products in grain and beet distilleries had demonstrated to me that the carbonic acid in the vats is almost completely absorbable by potassium hydroxide.
In several trials, where I collected 60 to 70 liters of gas each time in a few hours using a potassium hydroxide apparatus that dissolved the gas as it was released, I found that the carbon dioxide from these large-scale industrial fermentations, carried out in the presence of the ammonium salts naturally present in the liquors, contained approximately one ten-thousandth of its volume in nitrogen. 60 to 70 liters of the gas leave a residue of 7 to 8 cubic centimeters that is not absorbable by potassium hydroxide. These experiments would be worth repeating under the conditions described in the preceding paragraph. I nevertheless believed I could infer from their results, without further verification, that the nitrogen from the ammonia that disappears during alcoholic fermentation is not released in a gaseous state.
Guided by these indications, I wondered, however well-founded this assumption might seem, whether, under the conditions of fermentation, ammonia might not form albuminoid matter through a kind of coupling with sugar, thus becoming part of the yeast composition, which would explain its disappearance as ammonia. This is how I was led to the following results, which will demonstrate the full organizational power of yeast and will, it seems, put an end to discussions about its nature:
In a solution of pure candy sugar, I place, on the one hand, an ammonium salt, for example, ammonium tartrate, on the other hand, the mineral matter that constitutes brewer's yeast, and then a very small quantity of fresh yeast globules. Remarkably, the globules sown under these conditions develop, multiply, and the sugar ferments, while the mineral matter gradually dissolves and the ammonia disappears. In other words, the ammonia is transformed into the complex albuminoid matter that constitutes the yeast, at the same time as the phosphates provide the new globules with their mineral components. As for the carbon, it is obviously supplied by the sugar.
Here, for example, is the composition of one of the solutions used:
10 grams of sugar, ashes of 1 gram of yeast obtained in the muffle of a cup furnace, 0.1 grams of ammonium D-tartrate, traces of fresh, washed beer yeast, the size of a pinhead when fresh and moist, losing 80 parts water at 100 degrees.
In such a mixture, with the vessel filled to the neck and tightly sealed, or fitted with a gas tube immersed in pure water, fermentation begins. After twenty-four to thirty-six hours, the liquid begins to show noticeable signs of fermentation through the release of microscopic bubbles, indicating that the liquid is already saturated with carbonic acid. For I do not believe that fermentation manifests itself through the apparent release of gas before this saturation condition is met.
Over the following days, the liquid gradually becomes cloudier, as does the release of gas, which becomes noticeable enough for foam to fill the neck of the bottle. A deposit gradually covers the bottom of the container. Observed under a microscope, a drop of this deposit reveals a beautiful, highly branched yeast, extremely young in appearance; that is to say, the globules are swollen, translucent, and not granular, and one can distinguish, with surprising ease, each globule from the small quantity of yeast sown at the beginning. These globules have a thick shell, standing out in darker circles; their contents are yellowish and entirely granular; but the way in which they are sometimes surrounded by the young globules clearly indicates that they have given rise to those of the latter that form the heads of the strands.
It is in the first few days that these interesting observations must be made: in the evening, by the bright light of the gaslight, the old globules stand out from the infinitely more numerous young ones, just as one would distinguish a black marble among many white ones. Gradually, the differences disappear, and the newly separated globules lose all branched appearance; buds are no longer visible. The globules are very granular, like mature or exhausted brewer's yeast.
However, one should not assume that fermentation will ever become as active as if, instead of ammonia as the nitrogenous feed for the sown globules, one were to use a suitable albuminoid substance, such as grape juice, beet juice, or the soluble part of ordinary brewer's yeast. If fresh yeast globules are sown in sugar water mixed with a little of this albuminoid matter, the general phenomena will be in every respect the same as those I have just described, but the fermentation will be noticeably more active. For example, instead of appearing after thirty-six to forty-eight hours, the first small bubbles of carbonic acid appear after twelve to twenty-four hours. Moreover, the quantity of yeast formed and deposited in the same time is much greater; but, I repeat, everything is the same, only with greater energy, and the products formed are exactly the same.
Nothing is more curious than the influence of the nitrogenous and mineral nature of the environment on the activity of fermentation. I have conducted numerous experiments in this regard, some results of which I will report. One of the most interesting concerns the use of egg white albumin. I was quite surprised to find this substance entirely unsuitable for nourishing brewer's yeast globules. Whether one dissolves sugar in fresh egg albumin diluted in water and filtered, with or without making it very slightly acidic, and adds a very small quantity of brewer's yeast, the sown globules will not develop at all; there will be no trace of fermentation.
We know, however, from the experiments of Messrs. Collin and Thénard, that a decoction of sweetened albumin left to itself will ferment and that, according to Mr. Turpin, ordinary alcoholic yeast is formed within it; but, as Messrs. Thénard and Collin rightly observed, this effect only occurs after three weeks or a month at a temperature of 30 to 35°C, and from that point onward, fermentation is always very slow. Now, by studying the liquid during the interval preceding alcoholic fermentation, it is easy to recognize that various infusorial or mucedin products are formed within it, and I have no doubt that the albumin, somewhat modified in its nature by these substances, gradually becomes capable of nourishing brewer's yeast.
Things are quite different with blood serum or the fluids pressed out of muscles. Clear, colorless serum, when sugar and a few yeast globules are added, allows these to develop with remarkable ease, and the sugar ferments almost as readily as if one were using natural sugar juice or clear yeast water.
It is not, I think, that the albumin in the serum has a different nature from that of egg white; I believe it is due to other albuminoid substances that accompany albumin in the blood and that are, by their individual nature, suitable for nourishing yeast globules. This is what leads me to adopt this opinion. I coagulated some colorless serum, then boiled it with water. After filtering it to perfect clarity to separate the coagulated albumin, I dissolved sugar in the filtered liquid and added a few fresh globules. These multiplied, and a very well-characterized alcoholic fermentation occurred.
I performed the same experiment with boiled water on egg white and obtained no fermentation at all. These experiments were repeated many times and always yielded the same results.
In any case, isn't it quite remarkable to see an ammonium salt capable of nourishing yeast globules, providing them with their albuminoid principles, while pure egg white albumin is entirely unsuitable? It is thus understandable that the distance can be considerable between the various species of this generic group designated by the expression albuminoid or proteinaceous substances.
I have also observed that certain proteinaceous substances are much more favorable than others to the development of lactic acid yeast: for example, the soluble parts of gluten and casein, and the nitrogenous residue of liquids that have undergone alcoholic fermentation. Even when brewer's yeast globules are sown in aqueous solutions of these products, in which sugar has been previously dissolved, it is not uncommon to see alcoholic fermentation accompanied by lactic acid fermentation; that is to say, lactic acid yeast develops spontaneously (initially due to contact with air) and acts independently on the sugar, in parallel with the alcoholic yeast.
The influence of the environment, on the absorption of nitrogenous and mineral matter by yeast, manifests itself in yet another, no less demonstrative, way: I am referring to the spontaneous fermentation of sugary liquids without the prior addition of a specific yeast. Everyone knows that grape juice left undisturbed undergoes alcoholic fermentation after a few hours, and it is extremely rare for this to be complicated by another fermentation, for example, lactic acid fermentation.
The same thing happens to beetroot juice if it has been acidified, following Mr. Dubrunfaut's method, in the same way as the juices of acidic fruits. But here, one would already encounter quite frequently—and I have ample evidence of this—the production of parallel and simultaneous fermentations with their respective specific yeasts.
If yeast water is used—that is, the soluble part of brewer's yeast filtered to perfect clarity, then mixed with sugar and left to its own devices—alcoholic fermentation, that is, the spontaneous formation of brewer's yeast, will almost always occur if it has been exposed to air initially. Very rarely, although I have seen several examples in the course of my research, will only lactic, butyric, etc. yeasts be produced; but what is common under these conditions is the simultaneous formation of alcoholic and lactic yeasts, and one can even, to some extent, make one or the other of these yeasts predominate, depending on whether fresh yeast water or altered yeast water is used. The altered yeast water, although perfectly clear and obtained after boiling, will be much more suitable for the formation of lactic yeast.
These results would be even more striking if changes were made to the conditions of neutrality or alkalinity of the media, whether initial or permanent. One might have thought that in these phenomena the nature of the albuminoid matter, independently of its association or combination with mineral substances, played the principal role. But here are facts that clearly show that the presence and quality of mineral elements are no less essential than those concerning organic elements. Indeed, if the mineral matter is removed from the composition of the medium made of sugar water, an ammonium salt, and yeast ash, the sown globules do not multiply at all, and no fermentation occurs. Moreover, if the nature of the mineral components is altered, for example, by removing the alkali phosphates, the fermentation process is significantly modified and slowed.
Magnesium phosphate used alone does not produce the same results as raw yeast ash. Changes occur when yeast ash is used that has been melted to a white heat (which has partially expelled the alkalis) or simply sintered with moderate heat. It is under these latter conditions that the fermentation process is the fastest and most regular. Similar tests can confirm that an ammonium salt is essential. Yeast globules sown in sugar water mixed with yeast ash do not produce any noticeable fermentation. However, it is not entirely zero; it sometimes gives a fraction of a cubic centimeter of gas, which must be due to the ammonia in distilled water or to the infinitely small proportion of albuminoid matter present in the initial yeast.
The necessary intervention of sugar, which alone can provide the carbon for yeast globules, is sufficiently proven in these experiments that I need not dwell on it. Thus, everything contributes to the completion of the fermentation process: sugar, nitrogenous matter, and mineral matter.
The influence of the inoculum is no less certain. It is so significant that, if it is removed, fermentation still occurs, but I have never seen a single brewer's yeast globule develop, only infusoria, the smallest of all, and lactic acid yeast with the fermentation that is correlated with its development. Why this complete absence of brewer's yeast in these latter experimental setups? All the facts I have previously reported clearly explain it. It is because the medium is not sufficiently conducive to the propagation of this yeast. There is no physical impossibility of brewer's yeast forming, even if it is not sown. It does indeed appear spontaneously through contact with air in grape must, beet juice, etc., but the medium composed of sugar, phosphates, and ammonium salts is rather unsuitable for it, making its spontaneous production impossible, although this same medium can sustain the life and development of the mature yeast sown in it.
On the contrary, this particular, almost entirely mineral medium seems more appropriate for lactic acid yeast and infusoria, allowing the formation of these latter organisms, provided the liquid has been exposed to common air. If the mixture is boiled for a few minutes and heated air is introduced, no organism appears, no fermentation of any kind.
All these facts, which, it seems to me, shed new light on the phenomena of fermentation, will help us understand a very common feature of fermentations that take place in a medium composed of sugar water, ammonium salt, phosphates, and a small initial amount of brewer's yeast: the accidental emergence of lactic acid yeast and infusoria. The latter disappear quickly; they are only seen in the first few days, but the lactic acid yeast persists and multiplies, and often it even ends up acting almost alone, because the increasing acidity it brings to the liquid is very detrimental to the brewer's yeast.
The fact that the two yeasts were mixed, even though only brewer's yeast was sown, is due to the nature of the medium, which is more conducive to the development of lactic acid yeast than to that of alcoholic yeast, since, in the case of spontaneous fermentation, alcoholic yeast never appears.
Having established this, I will now give a detailed analysis of a fermentation carried out in a medium composed of water, pure candy sugar, ammonium tartrate, and white, melted, and pulverized yeast ash. On December 10, 1858, at noon, the following were placed in the oven:
10 grams of sugar.
100 cubic centimeters of water; 0.1 grams of ammonium D-tartrate.
Ash from 1 gram of yeast.
Traces of fresh yeast (the size of a pinhead).
On the 11th, at 4 PM, upon careful observation of the spot at the bottom of the vessel where the small fragment of added yeast had fallen, extremely fine gas bubbles were seen rising continuously. The phenomenon was continuous in this spot. Elsewhere, at the bottom of the flask, only occasionally did a very small and rare gas bubble rise. Furthermore, within the liquid, a few tiny flakes floated suspended at various heights by very small adhering gas bubbles.
At 7 PM that same day, fermentation was much more noticeable, though still very weak. There was already a little foam in the neck of the vessel, level with the liquid. Bubbles rose from various points at the bottom. On the 12th, fermentation was very noticeable; there was a lot of foam and already a noticeable sediment on the bottom: the liquid was cloudy from the suspended yeast, which was very fine and exhibited the characteristics I have already described. On the 13th, 14th, and 15th, fermentation was active, but in the following days it gradually slowed down, although it remained continuous.
By examining the liquid from time to time in January, it was evident that lactic acid yeast had developed and was increasing, as was the acidity of the liquor. Seeing that the lactic acid fermentation was hindering the alcoholic fermentation, I ended the experiment and studied the liquor.
One portion yielded a very significant amount of alcohol, which was not measured.
Analysis of the remaining sugar using copper sulfate [Fehling’s test solution] showed that 4.5 grams of sugar had fermented, meaning that 5.5 grams remained.
Saturating 10 cubic centimeters with standardized limewater revealed the formation of an amount of acid equivalent to 0.597 [gram?] of sulfuric acid, which is approximately 1 gram of organic acids; a considerable quantity, clearly indicating that the alcoholic fermentation had deviated from its initial direction.
I determined the amount of ammonia by analyzing 50 cubic centimeters of the fermented liquid. I found that 0.0062 grams of ammonia had been lost. The yeast, initially collected on a tared filter and dried at 100 degrees, weighed 0.043 grams. I ensured that all the yeast ash used had dissolved during the process. The weight 0.043 grams is therefore the actual weight of the yeast formed in its dry state.
To determine the nature of the acid, a portion of the fermented liquid was evaporated, re-evaporated several times with ether, evaporated again, saturated with limewater, evaporated once more, and treated with alcohol. This yielded a very small, crystalline, but unmistakable precipitate of calcium succinate, from which I extracted crystallized succinic acid, so easily recognizable even when working with a very small quantity of this product.
The alcoholic liquid produced abundant crystallization of calcium lactate mixed with a small amount of the calcium salt to which I have already referred. The volume of calcium lactate crystallization left no doubt that the greater part of the acid in the liquid was lactic acid. I transformed a portion of this salt into zinc lactate, easily characterized by its crystalline form.
Finally, in the residue treated with ether and left insoluble by this liquid, I clearly distinguished under the microscope a crystalline precipitate with the exact shape of mannitol and a sweet taste, thus eliminating any doubt that might have arisen from the presence of tartaric acid in the residue. As for glycerin, it was detected in this residue after treatment with the alcohol and ether mixture.
All these results, of the highest precision, although most were obtained by working with very small quantities of matter, establish the production of alcoholic and lactic acid yeasts and the specific fermentations that correspond to them in a medium composed solely of sugar, an ammonium salt, and mineral elements.
In this paper, I have only wished to highlight this result. I will later publish a separate work on alcoholic fermentation carried out under these conditions, and I will then study in particular the nature of the albuminoid substances of yeast formed with the help of sugar and ammonia (1).
< Footnote: (1) Mr. Dumas, when I had the honor of communicating to him orally the first announcement of the facts discussed in this chapter, was very struck by the individually necessary role of ammonia salts, phosphates, and hydrocarbons for the life and multiplication of yeast globules. And comparing yeast to the young tissues of plants:
“I understand,” he told me, “that sugar, ammonia salts, and phosphates are always found together in plant sap. It must be with their help that the cell is formed.”
During my research, I had the opportunity to reread Mr. Payen’s remarkable work on plant composition, and I must admit that the numerous connections that can be established between yeast and the cells of young plant organs constantly came to mind.
The reader would do well to also compare Mr. Mirbel’s earlier research on the cambium with more recent writings on the role of phosphates in plant life. End footnote >
§ IV. — Study of the relationship between yeast and sugar. What happens to the nitrogen in yeast during alcoholic fermentation.
We have now arrived at a very delicate point in this research: the relationship between sugar and yeast. Here, we will deal only with the physical aspects, but these are obviously the ones that must be considered before addressing the more intimate, physiological relationships.
I will begin by giving some details on the structure of brewer's yeast globules. Yeast globules are undoubtedly formed of small vesicles with elastic walls, filled with a liquid associated with a soft, more or less granular and vacuolar substance located primarily immediately below the wall, but this substance gradually moves towards the center as the globule ages.
The cell wall is elastic. Indeed, when a drop of water filled with young yeast cells dries on a glass slide placed on the microscope slide, the removal of the drop, which divides upon the introduction of air, causes pressure between the cells, and they can then be seen to deform and become more or less polyhedral.
The contents of the cells, especially the central contents, are liquid: this is evidenced by the presence in most mature cells of one or more internal granules, agitated by the characteristic twitching of Brownian motion. It would be quite difficult to say whether this is a true Brownian motion. The cause of this movement, probably entirely physical, is still too poorly understood to know whether or not it can act through the cell wall on the freest granules in the center of these cells. In any case, the fact I am pointing out leaves no room for doubt regarding the more or less liquid state of the interior of the globules.
The budding of the globules, which constitutes the important discovery of Mr. Cagniard de Latour, occurs, according to Mr. Mitscherlich, as illustrated by the passage from Fig. 3 to Fig. 4, that is to say, the new globule begins with a simple protrusion. My own observations have convinced me of the accuracy of this opinion of Mr. Mitscherlich. I have seen this many times in the clearest possible way. Soon the small bud, while remaining attached and fused to the larger one, appears to have its own envelope and to constitute a true globule in itself. The movements of the liquid can only detach it when it has reached approximately the volume of the parent globule. Until then, its fusion is quite intimate and resistant.
Is the bud formed, as some people claim, by the effect of contact, of pressure against the inner wall of one of the granulations of the globule? I have seen nothing to support this opinion, and I believe it to be incorrect. On the one hand, translucent globules, without apparent granulations, are of all globules the most prone to budding; on the other hand, the development of granules seems linked to the more or less advanced age of the globules, and there are all the more of them the older, less active, and less capable the globule is of budding.
I do not believe either in a fact accepted by Mr. Mitscherlich, and already put forward by Messrs. Cagniard de Latour and Turpin, namely, that yeast globules often burst and release their granular contents, which disperse into the liquid as seeds, which then enlarge and become ordinary yeast globules. I can affirm that never, in the course of three years of the most assiduous and numerous observations on brewer's yeast, under the most extreme conditions of its development, have I once encountered this phenomenon.
There is one decisive circumstance against it: the uniform size of the globules of a yeast acting on sugar. Those smaller than average are not free, but attached to larger globules in the form of buds. Now, it is clear that if the yeast reproduced through the released granules of the large globules, we would find all sizes of globules among those that are free.
German authors, including Mitscherlich, distinguish two types of yeast. They call one superior yeast and the other inferior yeast, the latter being used in the brewing of Bavarian beer, as it is produced and active at a lower temperature than the former. Superior yeast is said to be more active. German chemists say that this second type of yeast buds, and that this is how it reproduces; but inferior yeast, according to them, develops through the release of granules from mature globules (1).
< Footnote: (1) Mitscherlich, Sur la fermentation (Annales de Poggendorf, Volume LIX; 1843.) End footnote >
I would not dare to go too far in my opinion on facts that I may not have been able to encounter in my research, with the particular characteristics under which they were studied in Germany. However, I cannot help but express doubts about the existence of these two types of brewer's yeast. Yeast is sometimes lighter, sometimes heavier. It is lighter when it is young. Older yeast is heavier and more prone to compaction. It contains almost no liquid inside its globules, which are entirely filled with granules.
The difference has always seemed very marked to me between the density of young yeast and old yeast, more or less exhausted, and there is no doubt that by the gaseous movement of the vats there is on the surface young yeast, active, suitable for budding, and at the bottom old yeast, of slower action on the sugar; but these are two identical yeasts, at two different periods of their life and their action.
All the studies I have conducted compel me to reject the opinion of the existence of two yeasts with different modes of reproduction, and I am confident that by resuming careful examination of superior and inferior yeasts, my opinion will be shared. Moreover, I do not deny any of the specific effects attributed to superior and inferior yeasts. My research serves rather to explain them (1).
< Footnote: (1) I have already noted that it resulted from analyses of alcoholic fermentations that aged, granular brewer's yeast, which has lost some of its activity through the beginning of depletion, either by breaking down sugar or by transforming its own tissues as I will explain later, yields more glycerin and succinic acid and less alcohol than young, lurgid, translucent, and less granular yeast. It is easy to understand that at different ages, yeast has slightly different ways of living, independently of the modifications that the nature of the environment and external conditions may cause. End footnote >
Let us now follow the yeast from the moment it is exposed to sugar. When placed in pure water or sugar water, the yeast releases some of its liquid or soluble components, which disperse into the liquid. This is easily demonstrated. If the yeast is diluted with cold or hot water, which is then filtered, there will be albuminoid and mineral matter in the filtered liquid. Now, we know that, taken separately, these substances, when mixed with sugar and a trace of fresh yeast, cause the yeast to develop and multiply, and that the sugar ferments. The same effect will obviously occur with greater intensity and vigor when all the yeast is left in the liquid.
The experiment of producing yeast and fermenting sugar in a medium composed of sugar, ammonium salt, and phosphates clearly showed us that yeast lives and ferments sugar as soon as it is in the presence of:
1. Sugar,
2. A nitrogenous substance,
3. Phosphate minerals.
Now, yeast carries within itself these nitrogenous and mineral substances that are soluble, at least partially, so that when it is added to sugar water, it has exactly what it needs to live. Sugar never undergoes alcoholic fermentation without the presence of living yeast globules, and conversely, brewer's yeast globules do not form without the presence of sugar or a hydrocarbon substance and without the fermentation of these substances. What has been written contrary to these principles is based on inaccurate or incomplete experiments.
All chemistry textbooks present alcoholic fermentation as occurring under two very distinct circumstances, depending on whether the yeast is added to pure sugar water or to sugar water mixed with albuminoid substances: in the first case, it is said, the yeast acts [causes fermentation] but does not reproduce; in the second, it acts and reproduces: this is the case in beer making.
“If fermentation,” says Mr. Liebig, “were a consequence of the development and multiplication of the globules, they would not excite fermentation in pure sugar water, which lacks the essential conditions for the manifestation of vital activity. This water does not contain the nitrogenous matter necessary for the production of the nitrogenous part of the globules.” The globules cause fermentation, not because they continue to develop, but as a result of their nitrogenous internal part which decomposes into ammonia and other products, that is to say, as a result of a chemical decomposition which is the complete opposite of an organic process."
The facts I just reported are in clear opposition to this view, and I am certain that the phenomena are almost the same, whether the yeast is used in pure sugar water or in sugar water mixed with albuminoid substances. In both cases, the yeast organizes and multiplies; only in the first case, when fermentation is complete, all the globules, young and old, are deprived of soluble nitrogen. What nitrogenous food was present in this form has become insoluble on the newly formed globules. Therefore, all of these globules have no possible action on pure sugar water. There is no longer any nitrogenous food for the globules that would be young enough to act and multiply further.
On the contrary, in the case of fermentation in the presence of an albuminoid substance, while some globules do die, most of the new ones emerge from the liquid filled with nitrogenous and mineral matter, all capable of surviving with these nutrients in fresh sugar water.
Other observations, whose importance has been exaggerated without their accuracy being verified, have contributed to a great deal of obscurity surrounding these studies. In Mr. Thénard's Memoir, we find that 20 parts of brewer's yeast and 100 parts of sugar, after complete fermentation, left 13.0 parts of an insoluble residue which, when brought into contact with a new quantity of sugar, was reduced to 10 parts. We will see shortly how incomplete and erroneous this observation is when taken in these exclusive terms.
In any case, everyone thought that, in fermentation with yeast and sugar water, some of the yeast is destroyed, and they added: How different this is from what happens when yeast is placed in sugar water mixed with nitrogenous matter! For example, in beer making, "the ferment, far from being destroyed, develops by budding and increases considerably in proportion. This is because the albuminoid matter in the liquor is suitable for the nutrition of the globules, and the brewer finds six to seven times more yeast than he put in."
All of this has been quite misinterpreted. As for the quantity of yeast the brewer recovers at the end of the process, one could just as easily say that he collects a certain number of times more than he used; for if the brewer added only a few globules of yeast, there would be fermentation and yeast sediment just as much as if he added a fairly large proportion. Only the fermentation would be slower and could become lactic, which led to the practice of adding 1/4 of the yeast that may form.
But let us examine things more closely, and we will be convinced that in the fermentation of sugars in the presence of albuminoid substances, no more, and even less, yeast is produced than when fermentation is carried out with pure sugar water. I will begin by citing some observations on fermentations with yeast and pure sugar water, in which the yeast used, the yeast collected, and the soluble portion of the yeast remaining after the operation were weighed. I will then report results relating to fermentations in the presence of albuminoid substances.
It follows from the figures in this table that if one uses a quantity of yeast paste amounting, for example, to 15 to 20 percent of the weight of the sugar, one collects less yeast after fermentation than was used: A, B, C. Now, it was precisely under these conditions that Mr. Thenard had placed himself; he had used 20 parts of yeast paste per 100 of sugar. These are also the proportions he recommends in his Treatise on Chemistry.
But when the weight of yeast paste is reduced to 10 percent or less than the weight of the sugar, more yeast is collected than was used: D, E, F, G, H, I. [From the table on page 401 I cannot see this: D, F, I match what Pasteur is saying, but E, G, H does not. For example in row H he uses 4 grams of sugar and 1.474 grams of fresh washed yeast paste, where the weight of the latter is about 37% of the former.]
And, in all cases, if care is taken to determine the weight of nitrogenous extractive matter from the yeast, which is dissolved in the fermented liquid, it is found that, added to the weight of yeast after fermentation, it significantly exceeds the total weight of the original yeast. The increase amounts to approximately 1.2 to 1.5 percent of the sugar's weight. The disappearance of the yeast in Mr. Thénard's experiment is therefore only apparent. Less yeast was collected because a large quantity had been used for fermentation, and the amount that dissolved exceeded the weight of the newly formed globules. Therefore, and disregarding the dissolved matter, there was an apparent decrease in the weight of the yeast.
The substances that the yeast releases into the liquid depend both on its natural solubility and on the solubility of the substances isolated within it by the fermentation process. Let us now examine what happens when there is a presence and excess of foreign albuminoid matter.
I dissolved 9.899 grams of pure candy sugar and added to the liquor 20 cubic centimeters of a clear liquid prepared by boiling fresh brewer's yeast with distilled water and filtering the liquid. These 20 cubic centimeters of liquid contained 0.334 grams of albuminoid and mineral matter, very suitable for the life of yeast globules.
Then, after bringing the volume to 100 cubic centimeters, I added a trace of fresh yeast, which multiplied and caused the complete fermentation of the sugar, which was complete eleven days later.
I then collected all the yeast that had formed. It weighed 0.152 grams dry. The fermented liquid, evaporated and treated with a mixture of alcohol and ether to remove glycerin and succinic acid, leaves an insoluble, nitrogenous residue suitable for further fermentation. It weighs 0.260 g. Thus, 0.334 g of albuminoid and mineral substances were used, and 0.152 g of yeast and 0.260 g of nitrogenous and mineral matter not used by or released by fermentation was collected, giving a total of 0.412 g, the difference from 0.334 g being 0.078 g.
For a fermentation of 100 grams [of sugar], this would represent an increase of approximately 0.8 grams [of yeast, soluble + insoluble matter], to which should be added a small quantity of extractive matter removed by the ether alcoholic mixture, which, as we have seen, is found to an insignificant extent in the glycerin, and for the rest fixed on the calcium succinate in the method of analysis I have given for fermented liquids.
I have repeated the preceding experiment many times and always with essentially the same results; that is to say, in the fermentation of sugar in the presence of albuminoid matter, approximately 1 percent (of the weight of the sugar) of yeast and soluble products is formed, a little less, therefore, than when one operates with fully formed yeast and pure sugar water: further proof that things happen in the same way in fermentations with yeast and pure sugar water, or with yeast, sugar water, and albuminoid matter.
The slight difference observed in these two circumstances is undoubtedly due to the fact that globules formed in an environment rich in nitrogenous nutrients are more active and, for the same weight, decompose more sugar than those formed in an environment depleted of minerals or nitrogenous nutrients. It could also be due to variations in the assimilation conditions for perfect globules and for bud globules.
The importance of the results I am trying to highlight, considered in their relation to the general theory of fermentation phenomena, has led me to seek further confirmation of their accuracy in experiments that seem to me as decisive as those preceding them. I have just tried to prove that brewer's yeast placed in sugar water lives at the expense of sugar and its nitrogenous and mineral matter, soluble, or capable of becoming so due to the mutations that take place between the substances it contains during fermentation.
If this is the case, and to verify it as thoroughly as possible, let us determine the weights of sugar that can be decomposed by: 1. a known weight of brewer's yeast; 2. the soluble portion of an equal weight of the same yeast.
I took two equal weights of yeast, each 6.254 grams. One portion was fermented with 100 grams of sugar, the other portion was boiled with pure water for one hour, then filtered, and the clear liquid, completely free of globules, was mixed with 100 grams of sugar and a tiny trace of fresh yeast as seed. A similar test performed on a third portion of the same yeast showed that the 6.254 grams of yeast yielded 0.325 grams of nitrogenous and mineral matter to boiling water, with a remainder of 0.873 grams of insoluble products [weighted after the insoluble matter had been dried to remove the water weight].
The fermentations were of virtually unlimited duration. Begun on June 1, 1858, they continued into the first days of September, after which I was unable to monitor them further. I have already drawn the reader's attention to this indefinite duration of fermentations when there is an excess of sugar, a remarkable fact and, it seems to me, quite easy to explain today. We know, in fact, that the nitrogenous matter in yeast does not break down. There are only displacements or modifications of these substances, but they remain in the complex state (as we are used to finding in these products). Those which are soluble do, it is true, partly fix themselves in an insoluble state on the globules, but the organizing power of the globules is such that it is understandable that the old globules can give up their solid nitrogenous substances in a soluble state, to serve as nourishment for the new globules.
The seemingly endless duration of these fermentations with excess sugar makes experiments of the kind I am describing extremely difficult and delicate. In any case, after several months I studied the liquids and found that in the one where the yeast had been used in its natural state, the fermentation was almost complete, and that in the other, with soluble components of the yeast, nearly 70 grams of sugar had disappeared. But nitrogenous matter still remained in the liquid, and there is no doubt that it could have exceeded 70 grams, especially by collecting it separately for use in a new fermentation of the same kind. One can easily understand the significant disruption that a food diluted to such an extent that it is contained in 150 cubic centimeters of water, weighing only a few centigrams, must cause to the life of the globules.
I would not, however, argue that a given weight of the soluble portion of brewer's yeast could ferment exactly as much sugar as the gross weight of yeast used to provide that soluble matter. Everything leads me to believe, in fact, that a certain quantity of the portion of the globules insoluble in boiling water is capable of releasing a significant amount of its nitrogenous matter into the liquid during the very act of fermentation.
Nevertheless, I see from the experiment I have just described, and which I have often repeated with the same result, what an enormous proportion of the sugar ferments simply by the organization of the immediately soluble nitrogenous and mineral matter, an organization brought about by the globules themselves. No one, I believe, will be able to doubt any longer that the fermentation of pure sugar water is truly a fermentation that takes place in the presence of albinoid and mineral matter.
I will conclude this chapter by giving the nitrogen content of the various nitrogenous materials from an alcoholic fermentation in which the yeast was as depleted as possible. In a fermentation where 1.198 grams of washed yeast (dry weight, the nitrogen content of this yeast being 9.77%) were used to ferment 100 grams of sugar, 1.745 grams of yeast were collected after fermentation. This yeast contained 5.5% nitrogen. The extractive residue of the fermented liquid, insoluble in the alcohol and ether mixture, weighed 0.600 grams and contained 3.8% nitrogen. The extractive matter contaminating the calcium succinate weighed 0.500 grams and contained only 0.5% nitrogen. Purified, it contains none.
These nitrogen analyses were carried out on the fermentation products where I had most thoroughly depleted the yeast, and it can be seen that the yeast, after fermentation, still contained a high proportion of nitrogen. Moreover, its decrease, compared to that of the raw yeast used, is immediately explained by noting, on the one hand, that 1.198 were obtained from 1.700 at the expense of sugar, a substance devoid of nitrogen; and on the other hand, that this same weight of yeast, 1.198, decreased by yielding soluble nitrogenous matter to the liquid, since 0.600 of soluble products are found; if they do not come entirely from the yeast, they are largely supplied by it, since they are nitrogenous.
This is the twofold cause of the decrease in the percentage of nitrogen in the yeast during fermentation: 1. An increase in yeast weight due to consumption of sugar; 2. Weight loss due to the solubility of certain nitrogenous substances of yeast.
§ V. — In alcoholic fermentation, part of the sugar is fixed on the yeast in the form of cellulose.
[According to modern sources, the cell wall of yeast contains a material similar but not identical to cellulose.]
The fact that alcoholic yeast is produced in a medium composed of pure sugar and phosphates demonstrates the significant role sugar plays in yeast formation, at least under these specific conditions. There can be no doubt: 1. That the cellulose in yeast is composed of sugar components, 2. That ammonia combines with another portion of the sugar to form the soluble and insoluble albuminoid substances of the globules.
Do things happen in a similar way in the case of sugar fermentation in the presence of albuminoid substances? The experiments, which I will describe in detail shortly, will leave no doubt on this point. They establish, in fact, that there is more cellulose in the yeast after fermentation than before. So that by comparing them with those whose results I have just mentioned, they make it extremely probable, if not certain, that all the cellulose in yeast globules is made up of sugar elements.
What, on the other hand, happens between sugar and albuminoid substances when sugar ferments in the presence of these substances? Is there still a fixation of sugar as undoubtedly occurred with ammonia? Are the soluble albuminoid substances of the globules not, by themselves, capable of entering the constitution of the globules, of fixing themselves there in a solid state, solely through a kind of isomeric change? Must sugar, as in the case of ammonia, intervene, modifying them by transferring all or part of its elements? This is one of the most delicate points of these studies. It would be rash to equate a priori the two modes of fermentation with ammonia and with complex nitrogenous substances, and to assert by analogy that sugar must also intervene in the second case to modify the albuminoid substances and make them suitable for nourishing the globules.
We will see, however, that the increase in the weight of the yeast due to the fixed cellulose is not sufficient to account for the increase in the total weight of the yeast during fermentation; that is to say, if we add to the weight of the yeast used all the weight of the cellulose fixed during fermentation, we do not reach the total weight of the yeast and its soluble part as found after fermentation is completed.
It is therefore impossible to doubt that, apart from the formation of cellulose, a significant portion of the sugar is bound to the yeast. But does it bind by modifying the nitrogenous matter, as in the first case it binds to ammonia to make it suitable for entering the cells? I am inclined to believe so, but it would be imprudent to assert it definitively. I have no conclusive proof. Thus, the difference between the total increase in the weight of the yeast and the weight of bound cellulose could be explained by the production of one or more specific substances, none of which would interfere with the modifications undergone by the nitrogenous matter.
This study of the influence of hydrocarbons [carbohydrates] in the formation of albuminoid substances is of great interest; but it is quite delicate, and the facts require interpretation with great caution. Having established these general conditions, let us examine the proportion of cellulose in the yeast before and after fermentation. I fermented:
100 grams of sugar.
Approximately 750 cubic centimeters of water.
2.626 g of yeast (dry weight).
After fermentation, which lasted twenty days, I collected 2.965 g of yeast (dry weight). I then boiled two determined weights of fermented yeast and the same yeast before fermentation with sulfuric acid diluted with 20 times its weight in water for several hours (six to eight hours). The weight of the fermented yeast was 1.707 g and the weight of the unfermented yeast was 1.730 g, measured at 100 degrees Celsius [i.e. dry weight].
The residues insoluble in sulfuric acid were collected on fine filters and dried at 100 degrees Celsius. As for the liquors, after saturating their acid with barium carbonate at boiling, I determined the proportion of sugar they contained, both by Fehling's solution and by the amount of carbonic acid supplied by fermentation.
By reducing the results obtained to the yeast weights of 2.626 g and 2.965 g, it is found that 2.626 g of crude yeast yields an insoluble nitrogenous residue of 0.391 g (i.e., 14.8 percent), and 0.532 g of fermentable sugar, while 2.965 g yields a nitrogenous residue of 0.634 g (or 21.4 percent), and 0.918 g of fermentable sugar, it follows from this:
1. That in the fermentation of 100 grams of sugar by 2.626 g of brewer's yeast, approximately 0.4 g of hydrocarbon matter, transformable by dilute sulfuric acid into fermentable sugar, was fixed;
2. That there is a significant increase in the nitrogenous matter that is insoluble in dilute sulfuric acid.
This latter result is further proof that during fermentation, the soluble albuminoid substances contained in the active yeast are fixed in an insoluble state.
It remained to be determined whether boiling with dilute sulfuric acid had completely dissolved all the cellulose. In order to identify it, I determined the proportion of cellulose in the crude yeast using Mr. Schlossberger's method. I stated that Mr. Schlossberger, using potash, a reagent so frequently employed by Mr. Payen in his research to dissolve albuminoid substances, had succeeded in removing these from the yeast almost completely, leaving as a residue a substance of cellulose composition that can be transformed into fermentable sugar by sulfuric acid.
I treated three portions of washed yeast, each weighing 4.757 grams after drying at 100 degrees Celsius, with potash. The first sample was digested with concentrated potassium hydroxide of density 1.25; the second with a 10% solution; and the third with a 5% caustic potassium hydroxide solution. The contact period lasted eight days, with the flasks placed in a water bath for two hours each day. The mixture was then filtered and washed with acetic acid, and the filters were dried at 100 degrees Celsius.
The insoluble residues, composed of hydrocarbon matter that can be transformed into sugar by boiling with dilute sulfuric acid and leaving only a barely perceptible amount after this treatment, were found in the three trials to be 17.77%, 19.29%, and 19.21% by weight of dry yeast.
Now, the 0.532 grams of sugar obtained without the intervention of potassium hydroxide from 2.626 grams of yeast corresponds to 20.2 percent of the yeast's weight. Boiling with dilute sulfuric acid had therefore indeed removed all the cellulose.
The preceding results indicate a lower proportion of cellulose in the yeast than that given by Mr. Payen, but I would point out, firstly, that the titration carried out by Mr. Payen appears to have been indirect, and secondly, it is quite evident from everything we have just discussed that the proportion of cellulose in the yeast varies with its age and the duration of its action on the sugar.
Thus, we have just determined that the yeast collected after fermentation yielded 0.918 grams of sugar for a total weight of 2.965 grams, which indicates 31.9 percent cellulose, a figure 11 percent higher than it was before fermentation. I need not point out that this considerable increase in the weight of cellulose in the yeast while it acts on the sugar is yet another piece of evidence to add to all the others I have given regarding the life of the yeast during alcoholic fermentation.
§ VI. — In all alcoholic fermentation, a portion of the sugar is fixed in yeast in the form of fats.
Braconnot has long noted the presence of fats in the lees [the sediment in unfiltered wine], which are, to a large extent, nothing other than wine yeast. The previously cited analysis by Mr. Payen gives 2 percent fat in brewer's yeast.
It is generally believed that the fatty substances in yeast are derived from the fatty substances of barley or other substances used to prepare the yeast. I have determined through a direct experiment, very easy to reproduce, that during fermentation the yeast itself forms its own fat using elements from sugar. I mix with sugar water, prepared with very pure candy sugar, a clear yeast water extract treated several times with alcohol and ether. To the mixed solution, I add, as seed, a tiny quantity of fresh yeast globules. They multiply, the sugar ferments, and in this way I manage to prepare a few grams of yeast using substances containing not the slightest amount of fat.
Now, I find that the yeast formed under these conditions nevertheless contains 1 to 2 percent of its weight in easily saponifiable fats with crystallizable fatty acids. This fat can only come from the elements of sugar or from the elements of albuminoid matter; but I have also observed that yeast prepared in a medium composed of water, sugar, ammonia, and phosphates also contains fat. It is therefore from the elements of sugar that the fat in yeast is derived.
These experiments, in their methods, are reminiscent of those that Messrs. Dumas and Milne Edwards carried out together to verify Huber's observations on the origin of beeswax. They provide confirmation of the views that Mr. Dumas has long held on the possible formation of fats using sugars (1).
< Footnote: (1) See on this subject Annales de Chimie et de Physique, 3rd series, volume VIII, pages 70 and following; 1843. End footnote >
§ VII. — Permanent vitality of yeast globules.
When, following the explanation I gave in one of the previous chapters, I attempted to confirm the results I had obtained on the presence of succinic acid and glycerin in fermented liquors by precisely measuring the carbonic acid content, I encountered some peculiar difficulties. Sometimes I arrived at satisfactory values, while at other times the volume of carbonic acid exceeded the calculated result, without any legitimate explanation being found.
Gradually, I realized that the cause of the discrepancies in the results might well be due to the use of excessively high proportions of yeast, which one is always inclined to overuse in these operations, with small quantities of sugar. I was led to the truth by the following fact:
For a particular purpose, which it is unnecessary to state here, I had fermented 5 grams of sugar with 10 grams of yeast paste, equivalent to 2.155 grams of dry matter, a weight of yeast far greater than what is necessary to transform 5 grams of sugar. Now, I was very surprised to see that this fermentation did not end very clearly, and that it had a tendency to continue with a very slight release of gas, as happens when the sugar is, on the contrary, in great excess relative to the yeast. Furthermore, the liquid, analyzed with Fehling's copper reagent, did not contain the slightest quantity of sugar, despite the contrary indication given by the release of carbon dioxide.
I then placed the following fermentations in inverted flasks over mercury:
I. Candy sugar: 1.313 g
Wine yeast (sediment from racked barrels): 6.95 g
Pure water: 9.336 g
II. Candy sugar: 1.4425 g
Brewer's yeast (containing 2.15 g dry weight): 10.00 g
Pure water: 9.21 g
Two days later, gas release was still noticeable in the flasks, and yet the first contained 360 cubic centimeters of gas and the second 387.5 cubic centimeters at 0°C and 760 millimeters of mercury [normal atmospheric pressure]. The theoretical quantities, even without succinic acid and glycerin, are 341.8 cm³ and 375.5 cm³. Despite this excess volume, the gas was completely absorbable by the potassium hydroxide. It was pure carbonic acid. Therefore, there was no doubt that using a high proportion of yeast would yield a volume of pure carbon dioxide far exceeding that indicated by Lavoisier's theoretical equation, which is already overestimated.
To better study this very unexpected phenomenon, I further increased the weight of the yeast. Since the results of the experiments I conducted in this direction were always the same, I will report only one experiment: 0.424 grams of candy sugar were fermented with 10 grams of yeast (dry weight). Two days later, the total volume of carbonic acid gas reached 300 cubic centimeters, nearly three times the theoretical volume, which is only 110 cubic centimeters.
It was of great interest to determine whether this excess carbonic acid was the result of a true alcoholic fermentation acting on the hydrocarbons of the yeast. To this end, I collected and distilled all the liquid mixed with the yeast and a certain quantity of barium water, which had been used to confirm that the gas formed was carbonic acid without any mixture of hydrogen or other gases.
50 cubic centimeters of alcoholic liquid were collected and distilled a second time, and the new liquid was studied using a very small alcoholmeter. To be certain of the alcoholmeter's readings, I prepared several liquids with mixtures of water and absolute alcohol by known weights, and I compared these various mixtures with the distilled liquid in order to understand the latter's relationship between two liquids of known composition.
In this case, I found slightly more than 0.6 grams of absolute alcohol. Therefore, much more alcohol was formed than the sugar used could provide, and in a quantity proportional to the total volume of carbon dioxide. We are indeed dealing with a true alcoholic fermentation.
Thus, when a proportionally very small weight of sugar is mixed with yeast, after it has been decomposed, the yeast's activity continues, acting on its own tissues with extraordinary energy and rapidity, which gradually slow down. This phenomenon presents several very interesting features:
1. If the fermentation is stopped when the volume of carbonic acid formed is equal to, or only slightly greater than, that corresponding to the weight of sugar used, no sugar is found in the liquor. It follows from this that yeast exerts its action on sugar before exerting it on itself. It feeds on sugar as long as there is any, and when it is exhausted, its life cannot stop suddenly and it continues at the expense of the materials it finds in its own tissues.
2. It is important to note the considerable activity of the secondary fermentation that follows that of the sugar.
3. The effect produced by the yeast on itself is not at all proportional to the weight of the yeast. It increases much more rapidly. I will go into some detail on this point, which is of great importance. Mr. Thénard and all the authors with him recommend, for obtaining good alcoholic fermentation, the use of a quantity of yeast equal to 20 percent of the weight of the sugar, or about 5 percent if the yeast is taken in its dry state. This proportion of yeast can be increased or decreased without any particular effect, except for a shorter or longer duration of the process. I have often fermented sugar with 1% of its weight in yeast (dry weight), and even then, this yeast had been washed, which reduces its activity. One can even go below this amount.
It would seem, therefore, that in all cases where these minimum doses of yeast are exceeded, a portion of it, remaining active and capable of acting on new sugar, should exert this remaining energy on itself, and that the proportion of carbonic acid and alcohol would be increased. This is not the case. I increased the weight of yeast to 8 percent (dry weight) or 40 percent (weight of paste), and I did not obtain a volume of carbonic acid greater than that indicated by rigorous calculation, that is, taking into account the succinic acid and glycerin formed during fermentation.
I have not conducted enough experiments to know the limit beyond which the yeast would act on itself; I believe that it would be necessary to use as much yeast paste as sugar, and that even with this proportion of yeast there would be no appreciable difference in the volume of carbonic acid (if indeed there was any), compared to what occurs in ordinary cases.
It follows from these observations that even a slight depletion of the yeast in the presence of sugar deprives it of this activity capable of acting on the hydrocarbon matter of its own tissues; and it can still transform sugar even though it can no longer make soluble and transform the cellulose of its envelopes.
I am very inclined to interpret the preceding facts as follows: Brewer's yeast, composed almost exclusively of globules that have reached their normal development—adult, so to speak—is placed in the presence of sugar. Its life begins anew. It produces buds. This is a proven fact. If there is enough sugar in the liquor, the buds develop, assimilate sugar and the albuminoid matter of the mother globules. They thus gradually reach the volume we know them to have. This is a true picture of ordinary slow fermentations.
Let us suppose, on the contrary, a weight of sugar far insufficient to bring the initial developments to the state of complete globules, or even of formed and visible globules; we will then find ourselves in the case of the experiments I have just described, and we will be dealing, in a way, with mother globules all bearing very young offspring. When external nutrients become scarce, the very young buds then live off the mother globules.
The preceding facts have drawn my attention to a long-recognized phenomenon concerning yeast, which, due to its connection with what we have just discussed, takes on a completely different importance than that attributed to it until now. Everyone knows that yeast in pure water, especially in summer, releases gas bubbles that rise from various points within the mass and burst at the surface of the liquid, lifting with them some of the deposited yeast. This phenomenon, like all those of the same kind, was explained by saying that, through contact with air, a decomposition process began in the yeast, that the yeast deteriorated. This judgment seemed all the more justified because, after a few days, the yeast gives off a very unpleasant odor; that the gases released contain a lot of hydrogen, hydrogen sulfide in small quantities, and that under the microscope a drop of the liquid or deposit shows itself to be filled with infusoria, vibrions, etc.; that finally the yeast loses much of its energy.
The explanation given for these phenomena is far from accurate. It is indeed true that yeast left in water eventually begins to putrefy. But the initial gaseous movement is by no means the beginning of this subsequent putrefaction. The highly active yeast in water transfers its activity to its own tissues, and a true fermentation takes place. Alcohol is produced in extremely significant quantities, as well as perfectly pure carbon dioxide, and the globules are seen to change completely, just as in ordinary alcoholic fermentations.
That the carbon dioxide is pure is easy enough to verify by placing the mixture in an inverted tube filled with mercury; as for the alcohol, it suffices to determine the amount of alcohol in the yeast on the first, second, third, and fourth days. The proportion of alcohol, almost negligible on the first day if the yeast has been washed and the washing water discarded, gradually increases, and as long as the microscope only shows the presence of brewer's yeast, a true alcoholic fermentation is taking place. Hydrogen gas does not appear, and alcohol formation does not cease until after the first few days, when the microscope reveals various yeasts, notably lactic yeast and infusoria.
This is precisely the phenomenon we studied earlier, but under conditions where it was exaggerated by the effect of the fermentation of the small quantity of sugar that preceded it. These observations seem to me to have real utility. Not only do they give us the key to the spontaneous alterations of yeast, independent of any putrefaction, but they also show that the life of this type of mycoderma begins as soon as the temperature and humidity conditions are suitable. Like a seed always ready to germinate, yeast, if it has the necessary temperature and water, lives at the expense of its own substance, and its life is manifested by the physiological process that characterizes it: the formation of carbonic acid, alcohol, succinic acid, and glycerin (1). When this yeast is placed in the presence of sugar, it simply continues its life, which is never suspended, but then it completes all its phases with a much greater apparent energy, because at the same time the sum of its life and organization is incomparably increased.
< Footnote: (1) If a very small glass cuvette containing young, translucent yeast, with little or no internal granulation, diluted in pure water without the slightest addition of sugar, is placed on the microscope slide, it will be seen, within a few hours, to gradually become yellowish and release pure carbonic acid gas. Internal vacuoles and granules form within it, similar to those that occur when it acts directly on sugars. End footnote >
§ VIII. — Applications of some of the results of this Memoir to the composition of fermented liquids. Specific studies on wine.
We have recognized that glycerin and succinic acid are constant products of alcoholic fermentation. They must therefore be present in all liquids that have undergone this fermentation, such as wine, beer, cider, etc., and indeed they are found there, although numerous analyses of these fermented liquids have not yet detected them. The non-crystallizable extractive substances that these liquids always contain have masked their presence.
I have only conducted specific studies on wine, the most useful and widespread of alcoholic beverages in France. Here is the process I used to extract glycerin and succinic acid from wine: 250 cubic centimeters of wine are decolorized with 20 grams of charcoal; the mixture is filtered, and the charcoal is thoroughly washed on the filter. The filtered liquid is evaporated slowly, and when its volume is reduced to approximately 100 cubic centimeters, it is treated with a few grams of slaked lime to saturate all the acids. Evaporation is completed under dry vacuum.
The remaining mass is treated in the capsule in which it is found, or better yet in a mortar, with the mixture of alcohol and ether discussed in this memoir, consisting of 1 part 90% or 92% alcohol and 1.5 parts 62% rectified ether. This treatment is rather difficult because some of the residue is plastic; but little by little it clumps together and can be ground. Each portion of the wash is poured onto the filter, and the ether is evaporated with the precautions I have indicated, in a weighted capsule whose drying is completed under a dry vacuum. The glycerin thus prepared is almost pure, as its analysis and taste prove (it contains no more than 1 to 1.5 parts per 100 of foreign matter).
Here is a table of some analyses carried out following these instructions. The figures given for succinic acid were deduced by calculating those for glycerin, taking as the ratio of these two products the one that was most common in all the fermentations I analyzed: approximately 3.5 parts glycerin to 0.7 parts succinic acid. I extracted pure, crystallized succinic acid from each of the wines listed in the table; however, in the precise determination of this acid, I encountered difficulties that I have not yet entirely resolved in separating it without crystallization from all the other components of the wine. I will have occasion to return to this interesting detail.
Wines from which all the sugar has been removed by fermentation yield an extract quantity that varies, according to different authors, from 15 to 25 grams per liter. More than a third, often nearly half, of the solid components of wine were therefore unknown until now, and undoubtedly the most important. Everyone will be inclined to attribute a useful role to glycerin, the essential substance of fats, in the beneficial properties of wine. The presence of glycerin in wine, where it is associated with albinoids and phosphates, deserves serious attention from physiologists. Succinic acid, despite its relatively small proportion, is far from negligible. The flavor of this acid has something unusual about it, and by mixing water, alcohol, glycerin, and succinic acid in the proportions used in fermentation, one is surprised to perceive how closely these mixtures resemble wine. We thus acquire the conviction that the flavor specific to this drink, in what is most "sui generis", is due in essential part to succinic acid.
The proportion of glycerin found in these various types of wine, when compared to the alcohol content from which the original sugar content of the grape must can be approximately deduced, seems to indicate that much more glycerin and succinic acid are formed during the fermentation of grape must than in ordinary fermentations. This circumstance may be due to various causes. One might have thought, for example, that grape yeast was a particular variety of alcoholic yeast, with its own special action, somewhat different from varieties of other origins.
In order to verify this assumption, I had wine yeast from the 1858 harvest brought from a vineyard in the Jura region. This yeast was nothing other than the sediment remaining in a barrel of white wine after racking in the first days of March. I washed this sediment through a filter. Under the microscope, it had the appearance of brewer's yeast. Its globules were very granular. It differed slightly from ordinary brewer's yeast in the uneven size of its grains. But there were no very small globules at all, and no trace of lactic acid yeast or foreign matter.
The activity of this yeast was of medium strength. 50 grams of candy sugar were fermented with 10 grams of this yeast, which contained 3.785 grams of matter in its dry state. On April 7th and 9th, fermentation was complete. All the sugar in the liquor was gone. Under the microscope, neither lactic acid yeast nor any foreign yeast was found. The fermentation was therefore entirely and completely alcoholic. After fermentation, I collected:
1. 2.750 grams of yeast dried at 100 degrees.
2. 1.576 grams of extractive materials, insoluble in the ether-alcohol mixture, including that which accompanies the lime succinate.
The sum 2.750 + 1.576 equals 4.326. If we subtract the 3.875 of dry yeast used, we find a difference of 0.541. Thus, for 50 grams of sugar, 0.541 of cellulose and other matter was fixed on this wine yeast, or 1.082 per 100 of sugar. This result agrees very well with those of fermentations with ordinary brewer's yeast, although there are significant differences in the proportion of soluble and insoluble matter in these brewer's and wine yeasts. This is evident from the figures I have given.
As for succinic acid, I found 0.433 grams, calculated from the weight of lime necessary for saturation. Finally, the weight of glycerin, determined with great care, was equal to 1.833 g. These figures, when doubled, correspond to 0.866 g of succinic acid and 3.666 g of glycerin per 100 g of sugar. They closely resemble those found in slow fermentations with brewery yeast, and are, conversely, slightly higher than those of rapid fermentations.
Thus, wine yeast yields the same results as brewer's yeast. It would therefore be more likely that the high proportion of glycerin and succinic acid found in the wine should be attributed to the particular conditions of the environment during the fermentation of grape must, or to some other unknown circumstance. In any case, the study we have just conducted on certain wine varieties and the products of fermentation by grape yeast provides valuable confirmation of the veracity of some of the principal results of this paper.
As for the interpretation of all the new facts I have encountered in the course of this research, I am confident that anyone who judges them impartially will recognize that alcoholic fermentation is an act correlated with life, with the organization of globules, not with the death or putrefaction of these globules, any more than it appears as a contact phenomenon where the transformation of sugar takes place in the presence of the ferment without giving or taking anything from it. It was with these lines that I concluded my Memoir on the discovery of lactic yeast, an organized product entirely comparable to alcoholic yeast, and to whose presence all the phenomena of lactic fermentation are due, just as the various phenomena of alcoholic fermentation must be attributed to ordinary yeast.