First a great thank you to Susan Coppin and the team of archivists at the University of Adelaide who made recovering these documents possible.
All these documents from Australia were contained in Series 600 of the Roseworthy Agricultural College archives. We should probably figure out how to digitize all of the papers.
This is Graham’s original thesis whose summary was published in an Australian Brewing Journal as part of the Karl Weiderhofer prize. Maynard Amerine included the journal article in his seminal look at brandy making, Commercial Production of Brandies, from 1941. For some reason the original thesis was never collected by Amerine nor other important papers that I’m republishing in this series. They might have been lost due to the war.
People can download the original paper for themselves but I’m going to re-type some choice quotes so they are better indexed by search engines.
“Various modification in the operation of the still and in the separation of the ‘spirit’ running are adopted commercially in Australia, but there are no recognized standards for the distiller to work upon. The rate of distillation is of prime importance, and a section of this paper is devoted to the study of it and its effects on the distribution of impurities in the fractions of the distillate.”
“Much importance has been attached to the ester content by some countries importing Australian brandy, and methods of increasing the amount of esters have been attempted by some distillers even as far as allowing the wine to become slightly ‘pricked’ to raise its volatile acidity before distillation. The volatile acidity of a wine was thought to have considerable influence on its ester development, and experiments were carried out to verify this. The ester content of a brandy is subject also to changes during storage, and changes during storage in ester content of freshly-distilled spirit were studied.”
What they might have found out is that not all esters are desirable esters. You want only so many esters of acetic acid. You want to have a base wine that is higher in other kinds of fatty acids to develop more ester diversity.
“At the present time there appears neither any relation between quality and composition, nor any methods of securing a spirit of desired composition (at least with a pot still) and it was for these reasons mainly that this subject was undertaken for investigation.”
“Vizetell – (1) says that the nearer brandies are distilled to their drinkable degree, the more perfect are their vinous flavour and aroma, and the more distinguishable their taste and smell.”
“Distillation at a low strength may be quite satisfactory for clean sound wines, but those unsound or of poor quality should be distilled at a higher strength to eliminate undesirable impurities and off-flavours.”
“Rate of Distillation.
Young (3) slow and regular distillation will give a more selective separation of impurities. Different rates of distillation will give different distribution of impurities in the distillate. Elliot (2) In brandy-making in France the distillation of the spirit fraction which amounts to 80-85 gallons for a still holding about 220 gallons, occupies from 10 to 11 hours, while the distillation occupies from 16-18 hours. This is a very slow rate of distillation. Schidrowitz (4) By vigorous distillation certain impurities as furfural may be formed.”
A bunch of new names here. I only recognize Young.
“The distribution of impurities in the fractions of the distillate.
It is well-known fact that in distillation, those substances with a higher vapour pressure (i.e. lower boiling point) will tend to distill over first. Consequently the more volatile substances as aldehydes will accumulate in the foreshots to a greater extent, and the slower the distillation the more marked should this separation be – Young (3). Young (3) The distribution of the impurities is, however also affected by their relative miscibility with the alcohol-water mixture. The rectifying action of the neck of the still will cause differences in the distribution also, and the longer the neck the greater will be the degree of victification and the more marked should be the separation of various constituents.
G. Buttner and A. Miermeister (5) studied the course of distillation and found that acetaldehyde distills early, furfuraldehyde at the middle of the run, while esters and higher alcohols distill more or less regularly and no appreciable partition is effected….”
A better explanation of this can be found in Roberte Léauté’s 1989 James Guymon lecture.
“…L.W. Cornell and R.E. Montanna (6) showed that acetic acid (principle volatile acid in wine) distilled more or less regularly from start to finish.”
Again, more great names that I haven’t come across before.
“The chief chemical changes (14) consist of an increased aldehyde, acid (both fixed and volatile) and ester content, the latter being formed by the esterification of the alcohols and acids present, this change being subject to an equilibrium constant which is only obtained after long periods at ordinary temperatures, and occurs when two thirds of the equilibrium mixture is esterified. Whatever the modifications in composition of the brandy occurring during maturation, there must be in the newly distilled spirit some optimum amount of impurities which will give a final product of highest quality for the type of wine from which it was distilled.”
So what we are dealing with is a moving target. A spirit straight of the still may not be instantly enjoyable so we have to know how it will evolve so that it can have the right starting point. I suspect many white dogs sold to be drunk immediately are not exactly the same as what is worth putting into barrels, or they are held to come to equilibrium.
“A considerable amount of work has been done on the ‘judgement’ of spirits. Most authors state that spirits cannot be judged on results of analyses alone, but by combining analyses with testing. Hehner (7) Analyst – 1905 – says that the quality of brandy cannot be judged by the amount of impurities present, and also that esters fluctuate as widely as other impurities. He says that too much importance is attached to the ester content. Quoting Nessler and Barth 1883, he says that the noses and palate of a brandy are better reagents for quality and purity than are chemical means. This was also recommended by Fresenius 1890, Sell 1892, Bersch 1895, and Koenig 1904. (Analyst 1905) (7). Hehner says that, taking all circumstances into consideration, it follows that the amount of impurities (coefficient of impurities) is largely an accidental quantity, applicable under proper restrictions, to products made in the same way from approximately uniform crude products, but not applicable to all brandies alike, not useful for fixing a standard.
Dr. Schidrowitz (7) did not agree that it was an impossible matter to practically substantiate the genuine character of a spirit. He said of the analysis done, the samples did not represent those coming from the still-head. There was practical conditions under which samples might be taken so as to enable it to be shown within what limits the different qualities are types of brandy did vary. It is evident that there is no definite relationship between the quality of a mature brandy and its composition., but there is possibly come relationship between the composition of a newly-distilled spirit and its potential quality.”
Lots of new names. I only recognize Konig. I had always wondered what these people were reading back then and apparently it was some serious stuff. I suspect most of the 19th century literature is not in English.
One problem Graham seems to have by looking at quality from an analytical measure is that his measurements are over simplified. He can only find total esters as acetic. He cannot further subdivide this ester content into the various individual esters like you can with new techniques though the new techniques cannot even find patterns in quality.
Understanding quality seems to be limited by our understanding of perception and neuroscience. Subjectivity seems to come from our tendency to generate aroma illusions from taking in just aroma fragments in a distillate.
Graham ends up distilling fast and slow with his experimental distillate and finds it “inexplicable” how they differed in alcohol content when the same amount of distillate was collected. I was hoping he was able to explain this phenomenon. I can’t remember which other paper it gets explained in but apparently distilling fast challenges the subtle reflux provided by the neck of the pot still and can lead to significant changes in fractions.
“The ester content of the distillate obtained by low speed distillation is significantly higher than that of the distillate obtained by high speed. This may be caused by a loss of esters due to incomplete condensation when distilling rapidly, or to the formation of more esters during the longer distillation.”
“The difference between the aldehyde contents is not significant.”
“Alcohol. By rapid distillation the strength of the distillate remains fairly constant for the first part of the distillation but after the 5th fraction the strength drops gradually in increasing amounts. When distilling slowly the strength falls gradually from beginning to end in increasing amounts. Thus when distilling slowly the temperature in the still will increase gradually, and impurities will distill over more uniformally, and their separation should be more marked. The lower strength of the fraction by the slow distillations is probably due to the difference in strengths of the low wines.
Esters. By rapid distillation there is a sharp decrease in esters from the first to the fifth fractions, but from the 5th to the 10th fractions the change is not appreciable.”
“According to Australian methods of separating the ‘heads’ and ‘tails’ fractions, 1 to 3 and possibly 4 would be separated as ‘heads’ while fraction 10 and 11, owing to their low strength, would most probably be included in the ‘tails’. This leaves fractions 5 to 9 (inclusive) as representing the spirits fraction. The average ester content would then be about 37, which is not high enough to satisfy the customs authorities of certain importing countries. On consulting the results of analysis of commercial brandies it is noted that the average ester content of the sample (excluding the samples 8,11, and 12, none of which are really typical commercial brandies in Australia) is 39, which agrees remarkably well with the above figure. The above method of separation is probably a fair indication of Australian method. The rate of distillation has considerable influence on the distribution of the esters throughout the distillate. When distilling slowly the ester content of the first fractions is not so high, and the decrease is not so sharp. The value for the middle fractions is not so low, and consequently the ester content of the spirit as obtained by the aforesaid method of separation will be higher. The actual value is 46 as compared with 37 obtained by rapid distillation.”
“Greatest differences due to varying the speed of distillation are noted in the distribution of the volatile acids. By rapid distillation the first few fractions comparatively high in volatile acids, there being a sharp decrease from the first to the fourth fraction, when variations become relatively small. The amount gradually decreases up to the seventh fraction when a slight increase in each fraction becomes noticeable. Marked increases are noticeable towards the end of the distillate….
…The value for the tails for the slow distillation is very much greater than that for the rapid distillation, and it must be concluded that by slowing down the rate of distillation the volatile acids will tend more so to concentrate in the tail fractions. The explains the difference in the volatile acid values of the low wines.”
At the end Graham gives an acknowledgement to Allen R. Hickinbotham for advice throughout the work. Someone named W.J. Hickinbotham did lots of theoretical work on distillation in the early 20th century that is referenced by Amerine. Apparently there was even a book written about the collaboration of Allen Hickinbotham and Roseworthy Agricultural College. I thought they might be one in the same person but it appears to be a common English name.
“References.
1. Vizell “Wines of France.”
2. Elliot, C. “Distillation in practice”.
3. Young, S. [Sydney] “Distillation Principles and Processes.”
4. Allen’s Commercial Organic Analysis – Vol. I Fifth Edition.
5. Buttner, G. and Miermeister, A. Brit. Chem.Abs. B. June 1930.
6. Cornell, L. and Montanna, R. Ind, Eng. Chem. Dec. 1933.
7. Hehner, O. Analyst 1905. “Brandy”. [this paper privdes a great list of names and titles we can scour to find analysis of early commercial products similar to joseph Konig which I discovered years ago.]
8. Methods of the Assocn. Official Agric. Chemists 1935.
9. Hossack, J. Analyst 1935. “Determination of Esters in Alcoholic Liquids.”
10. Kelly, C.W. Unpublished data.
11. Espezal, P. and Jaulmes, P. Analyst 1935. “Determination of Aldehydes in Wines and Spirits.”
12. Joslin, M.A. and Comar, C.L. Ind. Eng. Chem. July 1938. “The determination of Acetaldehyde in Wines”.
13. Sutton, F. “Volumetric Analysis”.
14. Simmonds, C. “Alcohol, Its Production, Properties and Applications”.
2 thoughts on “W.O. Graham. 1939. A comparison of the composition of successive fractions obtained during distillation and their relation to the composition of commercial brandies.”