High Fidelity Gin Distilling / Perceptual puzzles / Musings

Recently I’ve helped a few people start the journey of developing a gin. All I’ve really done is give people stuff to read so they were up on all the published literature. Sadly, gin has the most incomplete of all the distilling literature and it takes a lot of seeing for yourself to really get anywhere with it. One of the incomplete parts of the gin literature is how you make cuts, why you make them, and then what do you do with the cut fractions. Can they be recycled like in other spirits?

Gin is primarily made from neutral spirit on stills that we assume are free of fusel oils from the previous distillations of other types of spirits (remember the demisting test?). For atmospheric distillation (because gins are often distilled at vacuum or just partial vacuum), the heads cut is to separate unwanted terpenes which can have solubility issues when cut to bottling strength among other issues.

Terpenes are a very broad category of flavor compounds with differing volatilities and differing ethanol/water solubilities. I haven’t really figured them out to be honest. Think of them as what makes an expressed lemon peel so fresh, zesty, and angular. I think we could compare terpenes encountered in gin distillation to the esters and say some are more noble than others. Like the esters, many terpenes are formed (or transformed) in the still and more form at higher temperatures than lower which is why in both cases vacuum distillation produces less of each.

Terpenes are also more soluble in ethanol than they are in water which is why a Lemoncello has to be above a certain alcohol content to not cloud, the same with an Absinthe which louches when water is added, and the same with the first fractions of a gin when cut with water. But then at a sensory level, why do we want them in a lemoncello but not in a gin?

Strange sensory stuff happens with terpenes, perhaps just the generic ones, where they raise the threshold of perception of an essential oil therefore somewhat masking it. Sadly, I can’t find the why of this explained anywhere. All you really just see in perfume or flavoring literature is the rule of thumb that essential oils should undergo terpene separation to reduced the usage rate. But consider the Lemoncello, non removal of terpenes is not always a flaw, immediately we find a context where it is a significant feature.

Terpeneless and concentrated citrus oils from which only a part of the terpenes have been removed are widely used as flavoring materials as they have improved stability and a longer shelf life, a lower usage rate, and improved solubility making them of particular value in the flavoring of soft drinks and liqueurs. -Flavor Chemistry And Technology, Second Edition by Gary Reineccius.

This usage rate claim is wildly interesting and a surprise that it isn’t investigated further anywhere that I can find. Does some strange perceptual phenomenon happen when terpenes are removed, and is mastering this critical to creating high fidelity gins?

Cointreau was revealed to use a centrifuge in their production process but was very cryptic about why exactly they used it and Joseph Merory shared a wildly intricate recipe for a triple-sec that had multiple stages of terpene removal but no detailed explanation why.

What I’m wondering is if any of the heads fractions of gin distillations can be process by centrifuge for ignoble terpene removal then reintroduced into the hearts fraction to maximize fidelity because undesirable terpenes and desirable volatile oils are likely to overlap. Cointreau does staggering volumes and uses a continuous centrifuge, but batch style blood bank centrifuges are ubiquitous in culinary these days and the heads fraction of a gin distillation might practically be processed in one. But are any sophisticated producers doing this? It seems like the big guys are successfully private enough about their ways that they could be doing tons of things we don’t yet know about.

Hopefully I’ll be in a position to investigate this first hand soon. I do have a gorgeous three liter centrifuge ready to play with the fractions.

The other idea I’ve been curious about lately is how to teach the making of cuts by providing references from industry leaders and how this might be applied to working on the ultra small scale. When cuts are made to a distillation run of fermented spirit we are primarily concerned with tracking basic congeners like ethyl acetate, acetaldehyde and then fusel oil. We want these congeners as close to the recognition threshold as possible without going over. If they are over they will stick out and become a flaw (aka a regret or missed opportunity), but if they are just below they will support and bridge other aromas.

But this threshold line is for the overall spirit diluted with water and finished for the consumer. When we make the cut we are experiencing these congeners bunched up and over the recognition threshold so it can be tricky for new distillers to navigate. They don’t yet have the opportunity to see how things turn out down the road. Or what if they are making one offs and there isn’t much down the road? We need ways to trace references from a completed product back to when congeners are coming across the still.

Can we create references from the proliferation of white dogs on the market? The unaged white whiskeys could be taken and broken up in a vacuum still into segments. The critical segments would be at the beginning and the end where major generic congeners would be bunched up back over the recognition threshold. This would give new distillers targets to shoot for based on the decisions of major commercial producers.

What else am I missing?

Apparently I’m not the first person to use fidelity in this context and a helpful patent from Pepsi does the same. The patent is titled: Increasing the Terpene Compounds in Liquids and gives a nice background on the challenges. Here I’ll extract some choice passages:

Consumers also demand fidelity of flavor in soft drinks and other liquids.
The water-insoluble compounds in flavors typically make a significant contribution to the perception of flavor as a complete, true, faithful representation of the flavor. As the skilled practitioner recognizes, the water-insoluble compounds often introduce haze, cloud, precipitation, or a phase separation in aqueous liquids, or may form a ring on the beverage container. These phenomena may cause consumers not to accept the liquid because these phenomena often are taken as an indication that the liquid is unfit for consumption, or that the beverage has spoiled.
Removal of water-insoluble components from flavoring compositions, referred to as “extraction” or “washing” in the trade, typically provides an incomplete flavor. Thus, even though the liquid may not be hazed or cloudy, the product is rejected because the flavor does not mimic fruit flavor found in nature. For example, lemonade that does not contain an appropriate concentration of water-insoluble compounds tastes objectionably ‘watered down,’ or candy-like as compared to fresh squeezed fruit.
The inventors have discovered that terpene compounds are solubilized by addition of flavor compounds more polar than the terpene compound. The inventors have discovered that solubiiizing the terpene compounds enables a higher concentration of terpene compounds in aqueous solution. Therefore, transparent liquids can be made with a flavor that reproduces the intended flavor more faithfully than known flavor compositions that have lower concentrations of terpene compounds.
The skilled practitioner recognizes that it is possible to increase the concentration of terpene compounds in a composition by increasing surfactant concentration. However, the skilled practitioner also recognizes high surfactant concentration may lead to beverage formulation difficulties, including adverse flavor effects, high cost, excessive foaming, and the possibility that regulatory limits would be exceeded at a surfactant concentration required to achieve the desired concentration of terpene compounds.
[end quote]

So the strange thing is that the information here works against the earlier ideas of fidelity. Many people have the idea that ethanol is a powerful enough solvent, but various surfactants are included in the likes of Angostura bitters, Fee brothers bitters, and even certain gin line extensions like Tanqueray Rangpur. Apparently they do it for fidelity. And even in the products that Pepsi makes, this idea of fidelity might trump lowering of the usage rate mentioned above when terpenes are removed which could be economically very significant.

So where are we now? Terpene removal simultaneously increases and decreases fidelity?

In the world of alcohol and even perfume do we want fidelity and the faithful reproduction of an orange peel or do we just want raw, extraordinary, attention sensations like I’ve mused about before? Maybe its a matter of metaphor? The terpene removed gin doesn’t exactly have fidelity which would make it ordinary but rather it has some sort of contrast enhancement trick with an extraordinary clarity and sharpness not found in the natural world. Perhaps its like applying a hipstamatic filter of sorts.

Some times we get nostalgic and we want high fidelity Sorrento lemoncello because that is rare and extraordinary relative to other lemon experiences while other times we want that sharpness and subtle contrast enhancement of terpene reduced gins because that is extraordinary relative to ubiquitous high fidelity botanical experiences in your unabstracted every day lives. Gin, because of its manipulation is like watching David Lynch give the noir treatment to wholesome rural American in The Straight Story. The use of light and contrast enhancement lifts it all up to be subtly, subversively, more attentional; realer than real.

Important Snippets from Joseph Merory’s Food Flavorings

Merory’s out of print texts have escalated in value and become increasingly hard to find so I thought I would type up a few important recipes to help someone out.

A few things about Merory to note. Firstly, I only have the first edition of the book and there were a few more editions years after so who knows if any liqueurs recipes were added or changed. Secondly, Merory sometimes engages in what I think is armchair speculation and sometimes wrote about ideas he pondered but never actually tried. So who knows if he actually tried these recipes. I’ve seen this behavior in other major texts about spirits especially in the context of chemical analysis procedures.

I typed up this first orange essence recipe because it was all in oil measures. I thought it would be useful as a starting point to give people an idea of the ratios of aromatic adjuncts like nutmeg and coriander as well as an idea of how much terpene is removed.

Orange-Curacao (Triple Sec) Essence MF 229

(a) Mix the following oils:
91.5000 gm. bitter orange
17.5000 gm. orange, cold pressed
04.2500 gm. lemon, cold pressed
00.1250 gm. nutmeg
00.0625 gm. neroli
00.0625 gm. coriander
113.5000 gm. or 4 fl. ozs.

Mixture of:
04.0 fl oz. mixture of (a)
12.0 fl oz. alcohol, 95 per cent
18.0 fl oz. water
34.0 fl. oz.
mix well, and let stand in a terpene separator (Fig. 14) for 24 hours for separation of terpenes;
-3.5 fl. oz. separated terpenes
30.5 fl. oz. taken from below; then add:
+1.5 fl. oz. alcohol, 95 per cent
32.0 fl. oz. finished curacao (triple sec) essence, filter if necessary.

Besides the very significant amount of aromatic adjuncts, notice how this Grand Marnier knockoff uses a combination of infusion and distillation to create the final product. This is a very different idea than the clear Grand M’s on the market now.

Grand “M” Type Flavor MF 257 (Continental Formula)

(a) Extract the following comminuted botanical ingredients:
4750 gm. orange peels, bitter
2500 gm. peppermint herb
2250 gm. orange peels, sweet
1750 gm. lemon peels
1500 gm. coriander seed
1500 gm. ginger
1500 gm. orange blossoms
0875 gm. cinnamon
1075 gm. cloves
0875 gm. angelica seed
0250 gm. cardamom
0100 gm. tonka beans
0110 gm. saffron
20525 gm.
with menstruum consisting of:
72 li. alcohol 95 per cent.
50 li. water
Extract for four days.
Then take off:
5 kg. extract
(b) Add to remaining botanical ingredients and menstruum:
50 li. water,
and distill slowly at atmospheric pressure to obtain:
90 li. flavor distillate
(c) Finished flavor mixture:
90 li. distillate (b)
5 kg. extract (a)
5 li. wine distillate
100 li. Grand “M” type flavor

Full Aromatic Liqueurs.–Full aromatic liqueurs are made entirely from flavor distillates. The procedure of the full aromatic flavor distillation yields a product with sufficient alcoholic content to make the addition of alcohol to the required proof strength for liqueur unnecessary. The alcoholic content of the finished liqueur is thus made up entirely from the alcohol contained in the flavor distillate. The full aromatic flavor distillation requires that the quantities of botanical ingredients, alcohol, and water be exactly determined to yield the quantity of alcoholic flavor distillate which is necessary both for flavor and alcohol content in the manufacture of the intended volume of liqueur.

The distillation procedure is performed at atmospheric pressure under the same conditions as described in the flavor distillation of botanical ingredients. Comprehensive knowledge of aromatic yield assists in determining the quantity of botanical ingredients from which to obtain the required flavor by distillation [emphasis mine!]. Experience in distillation and fractionation make it easy to calculate the necessary quantities of alcohol and water which are needed in the menstruum to yield a flavor distillate.

A liqueur made from the flavor distillate alone, containing sufficient alcohol content for its required strength, is a full aromatic product of unsurpassed quality. Formula MF 262 is the best example of full aromatic cordial production.

Full Aromatic Triple Sec Cordial Flavor MF 262 (Original French Recipe)
(Made from the peels of Curacao Oranges and sweet oranges)
First production:
(a) Put the following ingredients into a 200 gal. still with a perforated stainless steel plate above the edge of the steam jacket:
125.0 lbs coarsely ground peels of ripe sweet oranges
425.0 gm. orris root pulverized
170.0 gm. orange blossoms; add the menstruum of about 60 per cent alcohol content, consisting of:
249.0 lbs. or 30 gal. water
353.6 lbs. or 52 gal. alcohol 95 per cent.
(b) Procedure: After 24 hours extraction, distill at atmospheric pressure, slowly, without dephlegmation up to 78 per cent alcohol content of the condensate, then turn on dephlegmation to retain a high proof alcohol content of the distillate. The yield of the first fraction is:
40.0 gal. flavor distillate, of about 82 per cent alcohol content. It is used in (d)
(c) Procedure: The distillation of procedure (b) continues until all the alcohol is recovered. It yields a second fraction of approximately:
30.00 gal. distillate of about 45 per cent alcohol content. It is used in (f).
(d) Procedure: The 40 gal. flavor distillate first fraction of (b) is mixed with 40 gal. water. It is allowed to stand a few hours for separation of terpenes which are removed by decantation and the aqueous solution is then filtered. The terpene-free flavor is redistilled at atmospheric pressure, slowly, and in the same manner as in procedure (b), to obtain a first fraction:
20.00 gal. flavor distillate of about 80-84 per cetn alcohol content. It is then used in (m).
(e) The distillation of the terpene-free flavor of (d) continues unchanged, slowly, with dephlegmation, to recover all the alcohol and to yield a second fraction of approximately:
30.0 gal. distillate of about 50 per cent alcohol content. It is used in (f).
(f) Procedure: mixture and distillation of:
30. gal. second fraction distillate, 45 per cent alcohol content, of (c) and:
30. gal. second fraction distillate, 50 per cent alcohol content, of (e) and:
40. gal. water, to yield total of:
100.0 gal. mixture of about 28.5 per cent alcohol content. The mixture is left to stand a few hours for separation of terpenes. After the separation of terpenes it is filtered and then redistilled at atmospheric pressure, slowly, with dephlegmation applied to retain a high proof alcohol content in the distillate and yields approximately:
40.0 gal. distillate of about 64 per cent alcohol content. It is used in the second production batch and distillation of curacao peels of procedure (g) of second production.

Second Production:
(g) Put into 200 gal. still with perforated stainless steel plate above heat line, the following ingredients:
125.0 lbs. curacao peels, expulpated or coarsely ground
425.0 gm. mace, pulverized. Add to it a menstruum of 64 per cent alcohol content, consisting of:
141.1 lbs. or 17.0 gal. water, and
238.0 lbs. or 35.0 gal. alcohol, 95 per cent, and
40.0 gal. distillate, 64 per cent alcohol content, of (f).
(h) Procedure: After 24 hours extraction, distill at atmospheric pressure, slowly, without dephlegmation, up to 78 per cent alcohol content of the condensate, then turn on dephlegmation to retain a high proof alcohol content in the distillate. The yield of the first fraction is approximately:
60.0 gal. flavor distillate, of about 80 per cent alcohol content. It is used in (j)
(i) Procedure: The distillation of (h) continues until all the alcohol is recovered and yields a second fraction of approximately:
30.0 gal. distillate of about 30 percent alcohol content; it is used in (l).
(j) 60.0 gal. Flavor distillate of the first fraction of (h), of 80 per cent alcohol content, is mixed with:
60.0 gal. water, and left to stand a few hours for separation of terpenes. The terpene-free flavor is then filtered and redistilled at atmospheric pressure, slowly, and with dephlegmation turn on, to obtain a yield of approximately:
30 gal. flavor distillate (first fraction) of about 80-84 per cent alcohol content; it is then used in (m).
(k) The distillation of the flavor distillate of (j) procedure continues to recover all the alcohol and to yield a second fraction:
40.0 gal. distillate of about 50 per cent alcohol content. It is used in (l).
(l) Mix and distill
40.0 gal. distillate (second fraction), of 50 per cent alcohol content, of (k) procedure, and
30.0 gal. distillate (second fraction), of 30 per cent alcohol content, of (i) procedure, and
30.0 gal. water, to yield a total of:
100.0 gal. mixture of about 29 per cent alcohol content; the mixture is allowed to stand a few hours to separate terpenes. It is then filtered and redistilled at atmospheric pressure, slowly, and dephlegmation is applied to yield approximately:
40.0 gal. distillate of 64 per cent alcohol content. It is used in the next production batch of orange peels.
(m) finished flavor mixture consisting of:
20.0 gal. flavor distillate of 80-84 per cent of (d) procedure, first fraction, and
30.0 gal. flavor distillate of 80-84 per cent of (j) procedure, first fraction. Total:
50.0 gal. full aromatic flavor distillate, of about 80-84 per cent alcohol content.
Remarks:–If the entire quantity of the flavor mixture of (m) is used in the manufacture of 100 gal. Triple Sec cordial it yields a beverage of finest quality.

I boldened Merory’s remarks relating to oil yield analysis but nowhere in the text does he explain any methods for determining yield. It would also probably be really helpful to rewrite this recipe in the style of an infographic so the movements of the fraction recycling are much clearer.

Distiller’s Workbook exercise 15 of 15

Hopped Gin

Hops have one of the most seductive aromas known to mankind with a spectrum that is staggeringly broad so it is amazing that hops have never been widely explored in distillates. Unfortunately, a notorious louching problem is presented by hops which this exercise will explore. The recipe is less explicit than some of the previous exercises because hops vary so significantly. It is recommended to try the many proposed options and enjoy both your successes and cloudy failures in cocktails.

The inspiration for the hopped gin exercise came from the brilliant but seldom imported Japanese product Kiuchi No Shizuku which is produced by the Hitachino brewery. The famous Hitachino white ale is reportedly distilled once to a low proof, briefly mellowed in barrels, and then re-distilled with more hops, coriander, and orange peel. Other similar distillates might be the seldom imported beer schnapps of Bavaria.

Hopped spirits may not be common to the market because of how hops behave when distilled. A clear distillate is very challenging to achieve when working with hops and clarity is something consumers expect out of un-aged distillates. Unlike hopped distillates, most typical distillates become cloudy when certain compounds are included that are volatile only at high temperatures. These compounds can be avoided by making tales cuts below a certain temperature (which corresponds to a certain alcohol level). Whatever clouds hopped distillates, curiously, comes through at low temperatures and is therefore very difficult to avoid because these compounds overlap with (or even are) much of the hop defining aroma. The poorly soluble compound is likely a terpene and could possibly be separated through terpene separation instead of a heads cut where removal could separate other aroma compounds with it.

Hitachino mellows their distillate in barrels at a low proof before re-distilling which presents a clue. Wonderful research on limoncello, to which terpenes are important in defining flavor, state that if the alcohol content goes below 30%, terpenes are at risk of separating as either a louche or possibly an insoluble oil floating on the top. In the Hitachino production process, terpenes could be separated through mellowing at low proof before re-distillation. Stability tests could then be conducted to determine what percentage of terpenes could be reintroduced to the distillate while maintaining clarity at a range of proofs. A practical test for consumers would be making a chilled shot while maintaining crystal clarity.

Terpenes, which have a piney character, are often separated during the production of citrus essences and the process is described well by Joseph Merory in his text, Food Flavorings. The essential oils spend time in a conical separator where the terpenes accumulate on the surface as an insoluble oil that is easy to separate. Terpenes are also reportedly separated from commercial orange liqueurs due to concerns with either their solubility or possible concerns with their stability as an aroma. Conversely, as previously mentioned, terpenes are critically important to limoncellos where they contribute unique timbre and terroir, but also from the limoncello literature, terpenes reportedly can change detrimentally due to hydrolysis.

A concern about hopped distillates which can be explored (with self control not to drink it all!) is whether they change markedly over time which might be a reason they have never been common to the market. Hops are so magical an aroma that you would think every major gin distiller has explored them but perhaps taken a pass due to consumer notions of the stability of distillates. Certain hop distillates explored during development of this exercise have sat around longer than a year and some varietals seem to have lost aroma though age-ability trials were never set up systematically so no conclusions can be drawn for sure. There is also unexplored concerns whether the compound responsible for skunking beer is volatile and could effect a hop distillate stored in clear bottles. Supply chain management is much different than it used to be and with a more educated consumer base, there might now be room for a hopped distillate with a best by date.

The recipe tries to present an elegant starting point for hop aroma that works for nearly every varietal, but if the goal is to learn more about terpene management it may make sense to distill a concentrate, lower the proof, patiently separate the terpenes, re-distill, blend down with plain gin, then re-introduce the terpenes while doing stability tests.


500 mL dry gin (Seagram’s)

8 g hops (Cascade is a good place to start, but experiment with numerous types)

Mix and re-distill together on high reflux until the thermometer on the still reads 93.33°C. Going past 93.33°C may result in a cloudy distillate for other reasons.

Using your hydrometer re-cut the distillate to your desired proof (recommended 80-90).

Hops vary in the potency of their aroma. If the hop distillate is too aromatic it can always be diluted with more plain gin.

It may also be rewarding to collect the product one ounce at a time in small canning jars then dilute them one at a time from the beginning into the first half of the hearts to reduce the proof by 50%. This will illustrate how the first fractions louche. The distillate can slowly be married from the back to the front exploring the sensory contribution of each re-introduced fraction.


I.P.A. → I.P.C. (Imperial Pegu Club)

1.5 oz. Cascade hopped gin

.75 oz. triple-sec

.75 oz. lime juice

dash Angostura bitters


hopped Negroni

1 oz. Pacific Jade hopped gin

1 oz. sweet vermouth

1 oz. Campari

expressed oil of grapefruit peel


hopped Gin Fizz

1.5 oz. Chinook hopped gin

.75 oz. lemon juice

.75 oz. simple syrup (1:1)

top with soda water

Fruit Brandy Distillate and Brandy Flavor Essence

This is a small excerpt from a book I bought years ago that has become impossibly rare and expensive. Joseph Merory’s Food Flavorings (1960).

I was hoping to draw attention to the technique of making a fruit brandy by distilling first and fermenting second which might come across to some as kind of crazy.  To me, the idea of distilling unfermented fruit should be widely known as a tool for the amateur home distiller.  Often people have backyard fruit trees but not the time or space or expertise to ferment the fruit.  Passable brandies can still be made in short amounts of time by pulping the fruit then fortifying it with a non-neutral spirit such as rum or whiskey before distillation to collect the wonderful aroma.  A non-neutral spirit should be used as opposed to vodka because it will contain valuable, almost requisite congeners that end up missing by skipping fermentation.

[edited eventually to add: I suspect Merory wrote about techniques he never really practiced or never widely explored.  Some seem like arm chair ideas.  From what I’ve learned lately if I tried to capture the aroma of unfermented fruit pulp I wouldn’t do it with vodka but rather a “living” spirit that has has all the necessary generic congeners to support the fruit aroma; ethyl acetate and acetaldehyde.  Robert Leaute called these congeners “aroma fixatives”.  Without fermentation the fruit probably wouldn’t have enough of these.  The best place to get them is probably a cheaper rhum agricole.]

p. 298

Property of Fruit Brandies

Fruit brandies take their names from the fruit from which they are manufactured. They are reduced in alcoholic strength by the addition of water after distillation to make them drinkable.

Any fruit may be used for the production of fruit brandies. There are four ways in which to produce fruit brandies.

Crush the fruits, allow the pulp to ferment, and then distill the fermented mash.

Crush the fruits, allow the pulp to ferment, express the juice, continuing fermentation, and then proceed with fractional distillation of the cleared juice.  The middle fraction of the distillation can be used to fortify fruit flavors.

In this method the fermented and cleared juice of the second procedure is concentrated by freezing and centrifuging to an alcohol content of 30 per cent and then distilled.  The middle fraction of the distillation yields the finest fruit flavor essence.  It is useful as a natural fortifier of dessert wine, champagne, fruit flavored brandies, and fruit flavors.

[Merory doesn’t spell it out but freeze concentration here increases total acidity which catalyzes esterification which is a process of aroma creation in the still.]

In full flavor brandy, the evolution of carbon dioxide during fermentation tends to volatilize some of the aromatic, volatile flavor constituents.  In order to produce a full flavored brandy the aromatic fraction may be removed by distillation before fermentation and is then added to the distilled brandy.  To obtain the best possible yield, fruit should be subjected to slow fermentation.

[I bet this is one of Merory’s arm chair ideas but it seems like a fun things to try and possibly valuable to beginning distillers who are new to fermentation.  If the fermentation gets botched whatever alcohol and simple congeners would have been produced can be replaced in the next batch by a substitute like rhum agricole. I still have yet to try it]


p. 24
cherry and benzaldehyde flavor
wild cherry bark.
“… the chief constituent of the bark is d-mandelonitrile glucoside or prunasin, which has properties similar to the amygdalin of the almond seed.  the other constituents are benzoic, trimethylgallic, and p-coumaric acids, tannin, and volatile oil.  if ground or pulvergized wild cherry bark is treated with warm water of about 131F, enzyme emulsin will hydrolyze prunasin to benzaldehyde, glucose, and hydrocyanic acid.  the latter is removed chemically or is lost during the distillation.  distillation yields a flavor similar to kirschwasser…” [I suspect the use of the bark might be how Hiram Walker can make such a fun Kirschwasser so affordably.]
bitter almond
“… after removal of the fixed oil, the cake of the bitter almonds is mixed with warm water (131F) allowed to hydrolize and is then subjected to steam distillation.  the distilled oil contains more than 80 per cent benzaldehyde, party in free state but mainly in combination in a small amount of hydrocyanic acid.  the latter is removed chemically by heating the distilled oil with a sulfite, or a slaked lime and an iron salt, and then the mass is redistilled.  oil of bitter almond is heavier than water.”
“oil of bitter almond is also derived from the seeds of the apricot.  the oil derived from the seeds of the almond tree is imported mainly from southern france, spain, and italy.  it is obtainable in two varieties; of containing hydrocyanic acid besides benzaldehyde, the other being free from Prussic (hydrocyanic) acid, often labeled in abbreviated form– FFPA.”
“kirschwasser.–Kirschwasser is made by fractional distillation of a fermented mash of cherries and crushed seeds.  the increased temperature of the mash during the fermentation hydrolyzes the amygdalin of the seeds to benzaldehyde.  a 50 to 55 per cent alcohol content of the distillate yields the best aroma of kirschwasser.”