Follow along: IG @birectifier

[Batavia arak is one of the two mother rums, the other being Jamaica rum such as what is produced by Hampden. Batavia arak and Jamaica rum both use fission yeasts to achieve legendary amounts of rum oil, but do it in different ways. Batavia arak uses a high pH strategy where the fission yeast produces the rum oil (carotene derived aroma) by maximizing enzyme activity. Jamaica takes an opposite tact and uses a high acid strategy (not necessarily low pH because it is still higher than a winery ferment). The fission yeast in Jamaica becomes the symbiotic partner for bacteria which produces the rum oil. Cory and I have demonstrated rum oil creation with both fission yeast and bacteria. We hold a Batavia arak fission yeast strain from 1923 and Jamaica fission yeast from 1912. Arroyo likely did not realize it, but his high pH work with heavy rum from fission yeasts was an abridged version of the Batavia arak process.

In addition to the ending bibliography, its worth pursuing Kervegant’s chapter on Arak as well as this blockbuster 1936 paper also investigating Jamaica rum.

The photos are not from this article, but are from this wonderful resource.]


In 1893 a publication of VORDERMAN (1) appeared, dealing with the preparation of various indigenous foodstuffs and stimulants, in which, among other things, a description is given of arak production, and some remarks about micro-organisms that he found in fermenting molasses.

In 1894 a short publication was published by EYKMAN (2) in Batavia about the various microorganisms that seemed to play a role in arak manufacture. Like VORDERMAN, E. mentions a microorganism, which is always present in fermenting arak, with cells, in the form of a threshing flail. [dorschvlegel, and this could possibly refer to cell division by fission which the French sometimes described as scissoring]

[(2) At one time this was translated by Pombenet but seems to be lost when they rebuilt the site. Eijkman, who also won the Nobel prize for discovering Vitamin B, identifies a fission yeast.]

In the same year a publication by Went and PRINSEN GEERLIGS (3) was published, in which the arak manufacture, as it is done on the coast, is extensively treated. It also provides a detailed description of microorganisms that occur in the fermenting molasses. W. and P. G. also point here, among other things, to the difference in quality between the Batavia arak and the coastal arak, which writers mainly attribute to the difference in working methods, followed at Batavia and on the coast. According to them, fermentation in Batavian distilleries is dominated by a yeast called Saccharomyces Vordermannii P.G. and W., while in the distilleries, which work according to the usual coastal methods, fermentation takes place mainly by Monilia javanica W and P.G., while the S. Vordermannii plays only a very minor role. [Those initials after the names of the organism refers to these authors who discovered them.]

[(3) I never actually translated this paper.]

Finally in 1905, a publication appeared from PRINSEN GEERLIGS (4) on the quality of arak, in which he gives a detailed overview of results obtained during chemical research on arak and spirits of different origins. He showed convincingly that coastal arak and Batavia arak also differ chemically in composition, namely that in Batavia arak, the combined content of ester, aldehyde, etc. is much higher than in coastal arak.

He now no longer attributes this difference in composition solely to the various organisms that fermented the molasses, but mainly to the difference in composition of the raw material.

Batavia arak distillers then purchased their molasses from the Tangerang sugar factories, and because the work was very primitive there, the molasses was much less exhausted than molasses used by coastal arak distillers, which were obtained from the large factories in Cheribon, etc.

In recent years, however, the situation in Tangerang has changed completely: the Chinese factories have been successively closed down and as a result, the arak distillers, now also without exception, source their molasses from the coastal towns.

Despite approximately the same procedure, and despite use of the same raw materials, difference in quality still exists between Batavia arak and arak distilled on the coast.

Nor has the use of the more exhausted molasses of the large factories had the slightest influence on the taste of Batavia arak.

Some time ago, this writer had the opportunity to conduct an extensive investigation into the microbiology of arak manufacture in Batavia. The microbiology of fermentation has been extensively investigated, with the intention of finding, if possible, the cause of the difference in quality between coastal arak and Batavia arak.


There are 5 arak distilleries in Batavia, all of which are operated by the Chinese. The secret of fabrication is inherited from father to son.

Almost all Batavia arak is exported to Europe; only the products of failed fermentations come on the market here, and that is, after it has been made a drug by adding all kinds of herbs. In Europe, Batavia arak is bought by liqueur distillers, and then serves as a raw material for preparation of finer liqueurs, punch, etc.

An arak distillery in Batavia consists of 3 sections, namely:
1°. the yeast factory;
2°. the room where the batter is prepared and where the fermentation takes place, and
3°, the distillery.

[Arroyo used the word “batición” which also translates as batter like cake batter]


Raw rice, preferably as cheap as possible, is used as raw material for yeast production. The rice is boiled in large open pans under constant stirring over an open flame, and then spread on tetampas, to cool. [Tetampa may be a Malay word for embroidered cloth.] It is clear that the rice is strongly infected with all kinds of airborn germs during that cooling. When the rice has cooled, a certain number of raggi balls (1) are finely ground and intimately mixed with the rice. 10 raggi balls are used on 40 kattie rice [kattie = kilogram?]. The mixture is then poured into wooden tubs with a double bottom. On the top bottom, which is provided with holes, a mat is placed first, and then the rice is piled on it. When the tub is filled, the rice is covered from above with a second mat. The aim of the bottom mat is to prevent rice from falling through the holes, that of the top mat, to prevent exposure as much as possible.

The rice remains in the tub for 2 days, after which the cake is placed in the fermentation tray in its entirety.

In those 2 days a series of very interesting bacteriological processes take place in the rice, because different organisms that appeared in the raggi have developed.

As a result of this growth, a temperature increase takes place in the tubs, so that temperatures being reached can even exceed 50° C. The temperature in the tubs is regularly checked by the arak distillers; it must reach 50° C, otherwise, the “yeast” is not added to the batter.

I will not go into more detail here about the different processes that take place during those 2 days: as we will see below, the yeast thus prepared plays only a very minor role in the manufacture of arak.

I would just like to point out that due to the increase in temperature in the tubs, a large number of bacteria, which could exert a detrimental influence on the course of arak fermentation, are killed or at least weakened, so that they are otherwise harmless.



Molasses from cane sugar factories serves as raw material for preparation of the batter, that arrives in shiploads from Cheribon, etc. Molasses is stored at the distilleries in large shallow vats, open from above. Thanks to the preservative effect of the molasses, it does not ferment.

The batter is prepared by diluting molasses with potash water. Since molasses is difficult to mix with water, in some distilleries warm cooling water from the condenser vessels is used for dissolving. In European molasses distilleries dissolved rinsing water from the stills is used for this dissolution, a procedure which is never used here for reasons which can be explained.

Molasses is diluted in large wooden vats called yeast trays. These tubs are often partly sunk into the ground. The yeast trays are open again, of course, and sometimes barely sheltered from rain. Fortunately, in the dilutions used here, molasses is a material that is very unsuitable for development of microorganisms. Thanks to the high content of dissolved substances, bacteria and fungi already grow very poorly, while for yeasts this molasses is not an ideal breeding ground for a long time, mainly because of its alkaline reaction. In order to obtain a rapid fermentation, as is necessary, for example, in spirit plants, molasses therefore must be acidified first.

A slow fermentation is precisely what an arak distiller needs. The slower the fermentation proceeds, the more by-products are formed, and the amount, and the nature of those by-products, determine, as we have seen above, the quality of the arak.

In Batavia arak distilleries, concentrated batter is used from the outset, and as the fermentation continues, the concentration is gradually increased. The reason for this is twofold, namely: in the first place, because of the constant addition, there is no chance that fermentation will proceed too quickly, and in the second place, acids formed are neutralized again and again.

The first fermenter contains 40 liters of molasses and 160 liters of water, to which the rice from the tub is added. After 24 hours a clear fermentation has started and the batter is transferred to a second fermentation tank, with the addition of 224 liters of molasses and 672 liters of water. The batter remains there for another day, and is then poured into a second wooden container, after first having been sieved of the solids which float therein. Then 200 liters of molasses with 280 liters of water are added to the batter. After another whole day, another 200 liters of molasses are added, but no more water, and then the batter continues to ferment without further addition. After those 3 days, the main fermentation has ended.

For subsequent fermentation, the batter is poured into unglazed, round stone pots, which are filled up to the neck. Those jars have a capacity of about 20 liters. The filled jars are neatly stacked in rows. Of course, the fairly brittle pots break repeatedly. [Supposedly this area was littered with piles of broken pottery.]

Secondary fermentation in the jars takes 9 days. The whole process in a Batavia arak distillery therefore takes 18 days:
First and second day: preparation of the yeast.
Third, fourth and fifth days: fermentation in the fermentation tanks.
Sixth day: sieving of the batter.
Sixth, seventh and eighth day: main fermentation in the batter trays.
Ninth to eighteenth day: secondary fermentation in the pots.
Eighteenth day: distillation.

As discussed above, there is little chance of overgrowth by bacteria or other undesirable organisms taking place during the main fermentation. However, it is now completely different during secondary fermentation in the pots. No molasses is added during this secondary fermentation, and since the yeasts produce acid, the alkaline reaction of the batter gradually turns into an acidic one. Once this is the case, the batter forms a fairly suitable breeding ground for different groups of bacteria, for example for bacteria of the Granulobacter and Saccharobacter group.

If such an overgrowth by bacteria takes place, a batter is obtained which contains too little alcohol. Jars, which have had such a failed fermentation, can always be recognized immediately, because the top edge of the jar looks dry, and not moist, as with a well-running fermentation.

As the secondary fermentation progresses, and as a result of this the alcohol concentration increases, a slow death of the yeasts takes place: on the eighteenth day there is therefore almost no trace of a fermentation.


Distillation is started eighteen days after the rice, which is used for the yeast factory, has been cooked. The pots are emptied for this purpose in a sloping bamboo gutter, that flows into the top of a still. This boiler is cylindrical, and has a slightly curved bottom, above which a tap is fitted, in order to be able to drain the spent batter. The boiler has been bricked in as far as possible and heated over an open fire. The helmet is also made of copper and is attached airtight to the shell by means of clay. The helmet ends in a long spiral-wound tube, which passes through a vessel of water, which serves as a cooler. The boiler, which bears the name of crude boiler, is therefore as simple as possible; dephlegmation and rectification are not applied.

The firing is done with wood, and requires a certain routine to prevent foaming.

Distillate is collected in wooden barrels in various fractions. The first fraction has a strength of about 50%. The following fractions, of course, become increasingly weak until finally a distillate is obtained, which contains about 8% alcohol and which is often cloudy. Fractions, which are less than 50% strong, are collected and redistilled in the so-called still kettle. This still is designed in exactly the same way as the crude stills, and only serves to distill a high-percentage alcohol from the weak arak. The 50% arak, which was obtained as the first fraction from the crude kettle, is not yet strong enough; trade demands an arak of 60-66%. The spirit, which is obtained from the still, now serves to bring the arak to that strength.

Arak is stored on wooden shelves and undergoes a maturing process, which increases it significantly in smell and taste.


Nutrient agar was used to isolate the micro-organisms that play a role in the various phases of the fermentation process, to which, in addition to the necessary nitrogen compounds and inorganic salts, about 5% molasses was added. After sterilization, acidification was carried out with lactic acid to a weakly acidic reaction. Samples from the various trays and pots were seeded on culture media of this composition, and the organisms obtained thereby were further investigated.

Microscopic examination already showed that we were dealing with 2 fermentations, caused by 2 completely different organisms. One fermentation is caused by a type of yeast, which is very similar in shape to some wine yeasts and which I have named Torula indicum, and the other fermentation by an organism of very strange type, which already distinguishes itself from the ordinary yeasts, because multiplication takes place in 2 ways, namely: by budding as with the yeasts, and by dividing as with the bacteria, etc. Especially when examining the batter from the jars, the large cells, 2 by 2 connected by means of a threshing flail [dorschvlegel]. I gave this organism the name of Dematium arakii [This yeast possibly exists in a museum collection in Europe. It may also be the 1923 example I have.]. A detailed description of this Dematium follows below.

When I had isolated these two organisms from the fermenting batter, I naturally also tried to isolate them from the raggi and from the rice cake, which is added to the first fermenter at the beginning of the third day. As far as the Dematium is concerned, this proved impossible; I could not detect this organism in either raggi or rice cake. (1)

(1) As I later found out, this is not at all surprising; the Dematlum is not resistant to such high temperatures.

After many unsuccessful attempts this was successful for Torula indicum; the number of cells in this yeast was, however, much lower than that of the other Saccharomycetes occurring in the rice cake, such as, for example, Saccharomyces Vordermannii.

If one carefully inspects the terrain in an arak distillery, it is obvious that adding the relatively small amount of microorganisms from the rice cake can hardly affect the flora in the vats.

In an arak distillery, one always works with the same vats, which are never cleaned, but to which more or less diluted molasses are added time and time again. Once there is a flora in the vats, which under the prevailing conditions are the most favorable conditions for its growth and development, then that flora, despite all additions, will always retain the upper hand. Of course, such a situation has long since occurred in the arak distilleries, which have been in operation for decades. Whether or not to add the rice cake from the yeast factory will certainly not have any influence. The only influence that the cake can exert is the faster development of the Torula, because the easily absorbable nitrogen compounds from the rice cake diffuse into the molasses. Furthermore, the solid particles (rice grains, etc.) will exert a favorable influence on the fermentation, in that they prevent an oversaturation of the batter with carbon dioxide.

[Fascinating insights and I think I accurately translate that as decades which means this evolution of arak may not be older than mid 19th century.]

Thanks to the repeated transfer of one tub into the other, the batter is sufficiently aerated, which is particularly advantageous for the growth of the Torula.

Torula indicum is poorly resistant to higher alcohol concentrations, so that if that concentration gradually increases in the batter, it will be supplanted by organisms that are resistant to it, in this case Dematium arakii.

Besides Dematium arakii, some other organisms also developed on the molasses plates such as, for example, some colonies of Saccharomyces Vordermannii and some bacteria colonies. The number of these colonies, however, was so small in proportion to that of our Dematium that it is practically safe to disregard them.

It is quite strange that the Monilia javanica, isolated by WENT and PRINSEN GEERTIGS (I.c.), who plays a leading role in the arak distilleries on the coast, only occasionally appeared. It appeared in both the raggi and the rice cake.

If the alcohol concentration in the pots in which the refermentation takes place exceeds a certain percentage, then no increase of the Dematium takes place.

That working with pure cultures in an arak distillery, thanks in the first place to the composition of the raw material, not only is absolutely unnecessary, but also cannot yield the least results, I certainly do not need to demonstrate further after what has already been stated on the previous pages. [Awkward phrasing, but I think I translated it faithfully. Basically, he thinks you don’t need pure culture.]

As I discussed in detail at the beginning of this essay, different causes are given for the difference in quality between Batavia arak, and those distilled on the coast.

However, it follows from the above inquiries that none of those causes is correct, but that the difference in quality is due only to the various micro-organisms which effect fermentation on the coast and in the Batavian distilleries. [This is only half correct, because pH, which they were not aware of at the time, is also very key. These are high pH ferments similar to what Arroyo explored.]

In the Batavia distilleries, fermentation is caused by 2 different micro-organisms: Dematium arakii (d.K.) and Turula indicum (d. K.). [don’t forget, Dematium is a fission yeast, but these were still the early days where there wasn’t a concrete classification system.]

The role of the fast-growing Torula indicum is mainly to displace other unwanted organisms from the yeast trays, and to form small amounts of alcohol, making the batter more suitable for the development of Dematium. It does not influence arak quality.

Dematium arakii, on the other hand, causes the main fermentation of the batter and secondary fermentation in the jars, and determines arak quality by the quantity and nature of the by-products resulting from fermentation.

In the coastal arak distilleries, fermentation is caused by 2 completely different organisms, namely: Monilia javanica (W. and P. G.) and Saccharomyces Vorderininnii (W. and P. G.)


Torula indicum (d.K.)

On malt agar forms fast growing, flat colonies of regular shape. A fleece forms on malt extract. The yeast cells are ovoid or round, 2-2.5μ long and 1-1.2μ wide. Ferments glucose and levulose. Forms 4.5 vol. % alcohol in dilute molasses (1 to 3). Temperature optimum for growth is about 40° C.

Dematium arakii (d.K.).

In sugar-containing liquids, it forms elongated and elliptical cells up to 16 μ long and up to 6 μ diameter.

Multiplication, as described by LOEW (4) and DE BARY (5), takes place in 2 ways, namely by division (mycelial formation) and bud formation. In the liquid, as well as in the agar cultures, both forms always coexist.

The cells are surrounded by a clear wall, and the cell content contains, in addition to vacuoles, several globules of fat. In older cells or when cultivating in less suitable culture media, the cell contents become granular.

Spores are not formed, I have not been able to get the Dematium to form cells with thickened, colored cell wall (the so-called fumago shape). It is not surprising that Dematium has finally lost this quality due to the continuous culture in molasses yeast.

Dematium arakii fermented glucose, levulose, sucrose and maltose. Lactose is not fermented.

Molasses solutions are only slowly fermented in the first days: only after the fourth day does the fermentation proceed normally.

Concentrated molasses solutions (1 molasses in 1 water) are still fermented.

In dilute molasses solutions (1 molasses in 3 water), 12.3 vol. % alcohol formed.

The characteristic arak aroma is clearly visible on cultures in malt extract and diluted molasses. [Diageo intermittantly produces a single malt with a fission yeast.]


(1). Analecta op bromatologisch gebied 1 door A.G. VORDERMAN, (Geneeskundig tijdschrift v. Ned.-Indië Deel 33, pag. 369).
(2). Mikrobiologisches ueber die Arrakfabrikation in Batavia von DR. EYKMAN. (Centralblatt für Bacteriologie u.s.w. Band 16, pag. 97). [At one time this was translated by Pombenet but seems to be lost when they rebuilt the site.]
(3). Over suiker- en alcoholvorming door organismen in verband met de verwerking der naproducten in de suikerfabrieken door WENT en PRINSEN GEERLIGS. (Archief voor de Java-suikerindustrie 1894, pag. 529).
(4). De kwaliteit der Java-arak door PRINSEN GEERLIGS (Archief voor de Java-suikerindustrie 1905, pag. 379).
(5). Ueber Dematium pullulans de Bary von Loew (Pringsheims Jahrbücher 6. pag. 467).

BUITENZORG, December 1908.

Micro-biological Laboratory
v/h Dep. of Agriculture.



  1. Very interesting article from so many years prior to Arroyo, which points in the right direction for us! Thanks Stephen for this article!

Leave a Reply

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

Discover more from Boston Apothecary

Subscribe now to keep reading and get access to the full archive.

Continue reading

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