Microorganisms Causing Fermentation in Cane Syrups, Especially Barbados “Molasses”

Follow along: IG @birectifier

HALL (HH), JAMES (L.H.) & NELSON (EK) — Micro-organisms that flavor Barbados molasses. J. Bacteriology XXXIII, 577-585, 1937.


I’ve sat on this paper from Kervegant’s bibliography for quite a while even though I did share it around with industry colleagues. The researchers analyze an odd form of very high quality Barbados molasses that cures, ferments, and produces gas; a rum-like flavor develops. You would think this would be a document attributed to the U.K., but it was conducted by the U.S. Department of Agriculture.

During recent years a considerable quantity of a high grade cane sirup has been shipped into this country from the Island of Barbados, British West Indies. This sirup, commonly known as “Barbados Molasses,” is sold at a slightly higher price than domestic sirups and is generally distributed in barrels, a comparatively small quantity being sold in retail packages. Upon receipt in this country, the molasses is stored or “cured” in barrels for some months prior to distribution. During this curing period several reactions take place. First there is an active yeast fermentation accompanied by vigorous gassing, during which most of the alcohol is produced. A distinct odor and flavor of alcohol develop. After the fermentation stage, a rum-like flavor develops, and increases in intensity upon prolonged storage.

What is this stuff relative to common molasses and what extremophile organisms could be living in there? We get three citations and a paragraph that describes this fancy molasses.

Cane juice, clarified with milk of lime, is boiled down in open tayches. Factories making this type of product maintain a supply of “sour” juice, i.e., cane juice which has undergone natural alcoholic and then acetic fermentations. During the process of concentration sour cane juice is added to invert sufficient of the sucrose to prevent subsequent crystallization, the excess of the volatile acid being boiled away in the tayches. The sirup is evaporated, with continuous skimming, to about 36° Baume hot, or about 42° Baume cold. When cooled this sirup is run into puncheons, hauled to port, and stored in underground cisterns, pending shipment. Thus the output from many small factories is blended, overcoming individual variations. (2) Being free from sulphur dioxide, hydro-sulphites, and an excess of lime, these sirups possess a characteristic taste and aroma.

(2) Information from importer.

So we are not dealing with molasses so much as a clarified cane juice product. This high end product was also likely invented at the beginning of the 20th century, as evidence by pursuing their citations, and a big feature is the lack crystalization. Being free from mineral acid was also important.

42° Baume is a specific gravity of 1.402 which is also 75.6° if we allow 1° Baume to equal 1.8° Brix (Another formula is [42*1.905]-1.6= 78.41). This stuff is quite concentrated, but does have more than 20% free water so it is possible that microorganisms can grow. Somehow, possibly in that underground cistern, extra water likely gets absorbed.

What is extremely cool is that we see acknowledgment of cane vinegar being used in Barbados. Here, it is added to invert the sucrose and then the volatile acid is mostly evaporated as the juice concentrates. No sulfuric acid enters the process which would not evaporate and would influence the ability for microorganisms to grow later.

We think they are producing a culinary sugar product but they immediately reference Jamaica rum production:

Allen (1906), discussing the manufacture of Jamaica rum from sugar cane products, pointed to the importance of Bacillus Butyricus and Bacillus amylobacter and allied forms in the production of organic acids which are essential in the flavoring of rum. He also described the part played by yeasts in the production of alcohol, a forerunner of the aromatic esters. During the earlier part of the fermentation of the cane juice, members of the Saccharomycetes predominated, but as the acidity developed members of the Schizosaccharomycetes became prevalent. At the end of the fermentation only bacteria could be isolated from the “dead” liquor. Allen contends that at the end of the yeast fermentation the bacteria utilize the dead yeast cells as a source of food. Bacillus mesentericus was thought to produce butyl alcohol in the fermenting mixture. Minute quantities of higher alcohols, furfural, and aldehydes were present, presumably the products of microorganisms.

The cane juice is allowed to sour as it ferments, producing volatile acid, so the dominant yeasts species eventually changes from a budding yeast to a fission yeast. In Jamaica rum fermentation conducted with percentages of cane juice, the fresh juice fraction may act as an informal starter. After as much as 5% ABV is produced, a bacterial souring phase takes place. Muck is not mentioned here, but when it is added towards the end, it may increase nitrogen to a point it can create super blooms of bacteria.

The authors want to further develop this rum like aroma so they examine the microbial flora of the molasses as well as conduct chemical analysis on two entire cured barrels.

The centrifuge molasses samples to collect a sediment but then have some issues which they think they overcome. In a recently translated paper, A contribution to the ecology of fission yeasts on grapes (1977), the authors feel many surveys of yeasts have failed because of isolation techniques. Slow growing yeasts, can take considerable time to appear. They isolate stuff, but they might have not isolated everything.

For yeasts, they report finding:

Group I. Zygosaccharomyces nussbaumeri Lochhead and Heron. This group includes yeasts numbered 32-1, 32-3, 33-1,
33-5 and 33-6.

The characteristics of this yeast agree essentially with those of Z. nussbaumeri Lochhead and Heron (1929). A point of difference is the formation of scum by cultures 32-1, 32-3, and 33-5, but this single difference is not considered sufficient for disagreement with the type species. This yeast was first isolated from honey by Fabian and Quinet (1928) who named it Z. priorianus Locker. Later, however, Lochhead and Heron, discovering an error in Guillermond (1920), renamed it Z. nussbaumeri. It has since been isolated from fermented maple sirup by Fabian and Hall (1933).

Group II. Zygosaccharomyces major Takahashi and Yukawa. This group includes yeasts numbered 32-2, 32-4, 32-6, 33-2,
33-3 and 334.

Since the morphological and cultural characteristics of this group correspond essentially with those of the yeast isolated by Takahashi and Yukawa from ripening “Shoju” (1912); it is felt that it should be called Z. major Takahashi and Yukawa.

It is very cool that nearly a century ago, these Western investigators are drawing comparisons to Soju which is produced in numerous Asian countries and relates to Chinese Baiju which is known for incredible microbial diversity.

Zygosaccharomyces are an extremophile yeast that turns up in diverse concentrated syrups around the world.

The authors move on to bacteria:

A variety of colonies developed under practically all the test conditions in agar media. No one type of organism predominated, and subsequently none could be found to play a part in the flavoring of sirups. Mold colonies were frequent on plates incubated at lower temperatures under aerobic conditions. The molds proved to be mostly Penicillium and Mucor. Vigorous gassing and clouding of the broth in all the meat tubes occurred within 24 to 48 hours, so that anaerobic organisms appeared to be most prevalent. Later these were found to play a part in the flavoring of laboratory samples of sirups.

A modern version of the meat tubes may be Robertson’s Cooked Meat (RCM) Medium.

The growth in the meat-medium tubes was vigorous, accompanied by gassing and a highly putrefactive odor.

Intriguing. They try a variety of growth parameters to observe morphology.

This culture is identified as Clostridium saccharolyticum Bergey et al. Variations from the type species such as motility and spore
size were observed. Since all other characteristics correspond so closely with those described for, the type culture it does not seem that these differences warrant naming a new species.

They name a suspect!

When samples of sirup were fermented in the absence of C. saccharolyticum the characteristic rum flavor did not develop. That this organism utilizes the residual yeast cells and other organic matter in molasses and produces substances that contribute to the formation of rum flavors seems highly probable.

I interpret this as the Zygosaccharomyces grow, possible contribute little aroma and this bacteria feeds on them to support its own growth, producing aroma. Can they add nutrients and cut out the middle man?

The volatile constituents were distilled with steam from a barrel of cured Barbados molasses. The distillate was fractionated with a Glinsky column, and 2444 cc. of ethyl alcohol boiling below 80°C. were obtained. This alcohol appeared to carry most of the characteristic Barbados flavor. Careful fractionation of it gave a small amount of highly flavored distillate in which furfural was identified by its semi-carbazone. After the removal of the furfural an aldehyde was obtained by treatment with sodium bisulfite. It had an unmistakable odor of vanillin, but chemical proof of the presence of vanillin was lacking. In the nonaldehyde fraction butyl alcohol was recognized by its odor. The constituents forming the true Barbados flavor were not isolated.

There is a lot going on here! They use the volume of an entire barrel to accumulate to accumulate 2444 ml of absolute alcohol. This likely means the ABV of the molasses was less than 1%. A Glinsky column is a type of proto-birectifier with a weaker fractioning power. They are likely collecting multiple fractions similar to Micko’s 8 fraction method so isolate high value aroma. Something that is like vanilla but not clearly vanilla may be a rose ketone like damascenone, but we cannot be sure. It does sound an awful lot like rum oil. When they cannot observe what they think is the “true Barbados flavor”, that may be because that flavor relies on radiance and different constituents being smelt together. They may require esters to make their vanillin-like aroma smell like rum and vice versa.

True Barbados flavor is a beautiful phrase and its interesting that we are talking about culinary sugar and not even real rum. So far these researchers have not references any aspect of Barbados rum production and instead only Jamaican.

Don’t forget about that second barrel!

A second barrel of Barbados molasses was later distilled, and 15 gallons of distillate were collected. Acetaldehyde came over in
the first distillate and was identified by its semicarbazone. Ethyl alcohol boiling at 78.4 to 79.0°C. (2200 cc.) was also obtained from this distillate. The characteristic rum flavor was obtained on fractionating the alcohol and was also obtained from the watery distillates with ether. Hydrolysis or polymerization of the flavor fraction caused great loss in the essential oil which is responsible for the flavor, and only a very small amount was recovered. This had a powerful rum odor. From the flavor fraction and the last runnings of the alcohol, furfural was isolated as its semicarbazone.

I think they collect more distillate here because they think what they want is steam volatile. Their flavor fraction is likely analogous to the 5th fraction of the birectifier method. Ethanol is dropping off, and right before it ends, there is a band of volatility that contains the most high value aroma. I’m not confident I know how to define polymerization as they observe it, but I think the fragrant aroma compound links to something else making it no longer fragrant. This is different than how an aromatic acid may form a salt with an alkaline and become no longer fragrant. It may best parallel the formation of a glycoside where an aromatic molecule is bound to a sugar.

The isolation and identification of chemical compounds, which usually result from the metabolism of yeast and bacteria, add additional support to the belief that the flavoring of Barbados molasses is the result of a fermentation similar to that occurring in the manufacture of rum. The chief organism concerned is necessarily a yeast, upon which reliance must be placed for the production of alcohol. The volatile esters, higher alcohols, furfural and aldehydes probably result from the metabolism of bacteria.

Further evidence to support the theory of rum fermentation in Barbados molasses has been obtained by the development of
fermentation flavors in domestic cane sirups. Cultures of the yeast and bacteria previously described have been propagated in samples of high grade cane sirups and have resulted in the production of flavors not unlike those of Barbados molasses.

Remember, this team is from the U.S. Department of agriculture. They are bringing these ideas to domestic U.S. produced cane syrups. But what ever happened to it? Did the ideas vanish?


1. A microbiological examination was made of two samples of Barbados molasses.
2. The yeasts Zygosaccharomyces nussbaumeri Lochhead and Heron and Zygosaccharomyce8 major Takahashi and Yukawa were isolated and their r6le determined in the flavoring of cane sirups.
3. Of the bacteria Clostridium saccharolyticum Bergey et al., was isolated and probably aids in the flavoring of cane sirups.
4. The volatile substances ethyl alcohol, furfural, acetaldehyde, and butyl alcohol were obtained by distillation.
5. The relationship of yeast and bacteria to the production of rum flavors in domestic cane sirups is suggested.

So what happened? Did they produce rum oil in fancy cane syrup? Is there any way that any of these ideas can be harnessed for modern fine rum production? Is it viable to produce a fancy syrup, stimulate this kind of microbiol activity during a curing process before moving on to typical rum production? Do any producers using cane syrup, sometimes referred to as drum syrup such as in Haiti, ever experience similar curing in their own rum making process?


ALLEN, C. 1906 The manufacture of Jamaica rum. West. Indian Bull. 7, 141.

BERGEY, D. H. 1934 Bergey’s Manual of Determinative Bacteriology. Williams & Wilkins, Baltimore, Md.

BROWNE, C. A. 1919 Impressions of the sugar and syrup industries in Barbados. La. Planter, 63, 170. [Well worth a read]

FABIAN, F. W., AND HALL, H. H. 1933 Yeasts found in fermented maple sirup. Cent. Bakt., II Abt., Bd. 89, 31.

FABIAN, F. W., AND QUINET, R. I. 1928 A study of the cause of honey fermentation. Mich. Agr. Exp. Sta. Tech. Bull., 92

GUILLEBMOND, A. 1920 The Yeasts. Trans. by F. W. Tanner. John Wiley & Sons, New York, N. Y.

HALL, H. H., AND LOTHROP, R. E. 1934 The use of clarified honey in culture media. Jour. Bact., 27, 349-355.

LOCKHEAD, A. G., AND HERON, D. A. 1929 Microbiological studies of honey. I. Honey fermentation and its cause. II. Infection of honey by sugartolerant yeasts. Bull. No. 116 (N.S.) Canadian Dept. of Agr.

TAKAHASHI AND YUKAWA, M. 1912 On the budding fungi of “Shoju-Moranie.” Original communications: Eighth International Congress of Applied Chemistry, vol. 14-15, p. 162.

TEMPANY, H. A. 1913 Cane syrups or fancy molasses. West. Indian Bull. 13, 324.

WATT, F., AND TEMPANY, H. A. 1905 The fermentation of cane juice. West. Indian Bull. 6, 387.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

search previous next tag category expand menu location phone mail time cart zoom edit close