Introduction of citrus to cocktails in the 19th Century

[martin]

Before this comes off as obsessively trivial, three big things happened to American mixed drinks in their history: the first was ice, the second was vermouth, and the third was citrus juice. This concerns the third.

We seem to be pointing at the Crusta as the watershed moment of citrus entering the cocktail, if only in vaguely small amounts. We’ve also got a somewhat elusively-documented earlier tradition of some Americans doctoring their slings and slugs with a little lemon juice (e.g., Sour Whiskey, Whisky Sweet and Sour… see Oxford p. 657). This latter business seems to me the mostly intuitively likely source of the name of the 1850s drink category “Sour”. Punch is always a factor, and the Fix looms large, being quick and small and containing a good amount of citrus juice.

By the 2oth century, we’ve got cocktail-y things under various names (Sour, Daisy) routinely containing the 3/4 oz–1 oz of lemon (or lime) juice that is common in today’s “sour” cocktails.

In between the 1850s and ~1910, it’s weird. A lot—but by no means all—recipes in this time period call for quantities of citrus juice that seem neither here nor there, today. I vaguely recall some notion that Golden Era bartender’s were employing dribs and drabs of acid to some sophisticated effect, but… this strikes me as fanciful. Another argument is that “dashes” could be very large, or were wildly discretionary. Today’s mixologists tend to look at recipes from the period that call for 2–3 dashes of lemon juice and instinctively crank that straight up to 3/4 oz or so. While that might result in a drink that is a better suited today’s tastes, when someone like Harry Johnson wrote 2–3 dashes of lemon juice, I’m pretty sure he didn’t mean 3/4 oz.

There’s a pretty well documented spread of sourness in drinks spanning about fifty years that still feels foreign.

• 1/2 lemon is ostensibly 1 oz, but varies
• 1/4 lemon is ostensibly 1/2 oz, but varies, and on average, seems closer to 1/3 oz in my experience
• 4 generous dashes is, at most, 3/4 tsp (~1/8 oz)

Is there an identifiable trend here? A muddle? Was citrus employed habitually sparingly in some situations due to availability/cost? Was it an evolution of taste? Did some people fancy slightly-sour drinks? (Lemonade was common, though.) Some sort of stigma to gradually overcome?

I don’t have answers, but here’s a survey, focusing on the Sour:

---

Thomas (1862) calls for “a small piece of lemon”, squeezed, for Sours, and “a little lemon juice” for the Crusta

Haney’s (1869) calls for “a portion of the juice of” 1/2 lemon for Sours (and all the juice of 1/2 lemon for fixes)

Simmons (1874) says the Sour is the same as the fix, which he says calls for 1/4 lemon

Thomas (1876) now specifies a 1/4 lemon for Sours

Engel (1881, London) calls for "a small piece of lemon”, squeezed, for Sours [probably just plagiarizing 1862 Thomas]

Johnson (1882) calls for only 2–3 dashes lemon juice for Sours

McDonough’s (1883) explicitly has it both ways for the Sour: you can either use 5–6 dashes lemon juice, OR you use 1/2 lemon

Winter (1884) calls for 3–5 dashes lemon juice across his Sours

Byron (1884) has most Sours calling for 2–5 dashes lemon juice, but a couple—namely the “Continental Sour”—calling for 1/2 lemon

Cordon Bleu (1885) has a Bourbon Sour recipe calling for 1/2 glass (1 oz) lemon juice

Paul (1887) calls for 1/2 lemon for Sours

Thomas (post., 1887) now explicitly calls for 1 dash of lemon juice in the Crusta, with some Sours calling for 2–3 dashes of lemon juice and others for the juice of 1/2 lemon

The ever-surprising Proulx (1888—Chicago) has a whole business on Whiskey Sours, calling for 1/4 lemon or 2 tsp—roughly the same quantity; also discusses the Sour Whisky, which is a Sling with “a little lemon juice”; he also has the Silver Sour and Golden Sour, which are the Silver Fizz and Golden Fizz sans seltzer, with those calling for “the amount of lemon you would for a Whisky Sour”(!)

Schmidt (1891) doesn’t offer a lot of straightforward Sours, but the ones he has call for the juice of a lime, the juice of 1/2 lime, or the juice of 1/2 lemon; his Crusta omits any citrus juice

Boothby (1891, San Francisco) describes (#290) two methods: a Western-style “sour” sling (see also Proulx, above), and an Eastern-style [Brandy] Sour with sugar and the juice of an entire lemon.

Kappeler (1895) calls for 1/4 lemon in his Sours

Herbert Green (1895, Indianapolis) has a couple Sour recipes that variably call for 1/4 lemon or 1/2 lemon

Lawlor (1895, Cincinnati) calls variably for 1/4 lemon or 1/2 lemon in his Sours

Frank Newman (1900, France) Brandy Sour calls for 1/2 lemon and notes that some prefer no sugar; his Whiskey Sling (#315 & #316) calls for 1/4 lemon(!), his Gin Sour (#133) calls for 3 dashes of lemon juice

Johnson (1900) still calls for drops of lemon juice in the crusta, and 2–4 dashes of lemon juice across his Sours

Mahoney (1905) has sours that call for 2–3 dashes of lemon juice and others than call for 1/2 lemon

Boothby (1908, San Francisco) (#404) has replaced previous with gum syrup or sugar, and the juice of TWO lemons!

[CraigE]

It does seem to me like the cocktail recipe world has largely bifurcated into the margarita/daiquiri/sour shaken world expecting 3/4–1 oz citrus in a drink and the boozy stirred world that eschews citrus entirely. I think this binary conceit misses the “sharpening” benefits of citrus in smaller quantities. Here are three of my inventions that use between a barspoon and a quarter-ounce of lemon:

• Trampoline Casualty
• Forgotten Leap Year
• Feneon

[martin]

Your choice of the word “sharpening” would seem to place you in a modern culinary camp, perhaps in the same spirit as those who season their cocktails with saline? (I don’t discount either technique.) No question that a small measure of citrus has an effect, and “sharpening” seems like a reasonable word for it.

I may be overlooking something, but I haven’t picked up on 19th Century bartenders talking about citrus in a comparable way—or really (directly) talking about it at all.

Perhaps Santini was intuitively “sharpening” the Brandy Cocktail when he made the Crusta? Seems possible. Or was he countering his elaborately sugared rim? Are there some other old drinks that might shed light on the matter?

Given there was a longstanding habit amongst some drinkers to add a touch of lemon juice to their whiskey, presumably they liked the tanginess better than a merely sweetened whiskey? Was a Sour to the Harry Johnsons just the old Whiskey Sweet and Sour, whereas to many others, the Sour was a strained Fix without the decorations?

[CraigE]

The other variable I didn’t see you address is: has the fruit itself changed? Have there been changes in agricultural practices, varietal breeding, changing sources, etc. that made the fruit used in the century before last quite different from what we work with now (in volume and/or flavor)?

[martin]

I have never read anything that suggested regular lemons or their juice changed markedly during the last couple centuries.

So, this is Harry Johnson’s Gin Sour recipe:

And this is about as straight as I can make it, given the clunky verbiage:

I dialed back the sugar a touch and probably would earn an D-minus (at best) from Harry for my garnishing efforts.

As you can see, it is a slightly hazy drink, being nothing more than Gin Sling with a light touch of citrus. It bears absolutely no resemblance to a Daiquiri or anything else we would typically call a “sour”, today. Compared to a regular Sling or an Old Fashioned, it is, nevertheless, sour.

It is also quite nice.

[CraigE]

This 1967 article touting California-grown lemon growers says “Their share of lemon sales in the United States rose from a few per cent in the 1880’s to 18 per cent in 1900. By 1920 their share had increased to 75 per cent. After 1930 only small numbers of foreign lemons were being imported and American lemons were entering West European markets in large numbers.”
So even setting aside “terroir” or variety differences, I would think the shorter transportation times of domestic lemons that took over in the twentieth century would likely change how American drink-makers might have used them compared to the import-dominated days of the nineteenth century.

[EvanD]

I’ve read through this a few times, and I’m still not sure what your question is. Given your survey, it looks like 1/4 lemon (or 1/3 oz, or 2 tsp) is pretty standard for a “sour cocktail,” but also 2~3 dashes, so I have to assume those are the same. What I would want to see is graphic evidence of how lemon juice was dispensed in this time period. There were probably places that had quarter or sixth wedges which were finger squeezed, and there were probably places that pre-squeezed batches of juice, but I wouldn’t know if dasher tops were used in these instances.

What I really want to know is how you were supposed to ornament a crusta with multiple fruits when it already had the lemon peel. Pineapple matchsticks? Fine-chopped strawberries?

[martin]

@EvanD

Given your survey, it looks like 1/4 lemon (or 1/3 oz, or 2 tsp) is pretty standard for a “sour cocktail,” but also 2~3 dashes, so I have to assume those are the same.

If a dash was as much as 1/8 oz, then Johnson’s Martini recipe would contain a whopping 1/3 oz Boker’s bitters and 1/3 oz of gum syrup.

Regardless, 1/3 oz of lemon juice is less than half what winds up in the vast majority of citrus cocktails, today. My hunch is that we’re actually looking at a “lost style” of drink that went extinct once we could have all the fresh citrus we wanted and we got used to it.

@EvanD

What I would want to see is graphic evidence of how lemon juice was dispensed in this time period.

I think a dash is an inherently subjective unit (dashers being inconsistent in design and inconsistent in yield, depending on fill level and vigorousness). “A dash” is seasoning, like salt in cooking (often not measured, but added in pinches based on experience). If these guys are writing down instructions in dashes, as opposed to measured quantities, then it seems to me to imply the use of dasher bottles and disretion. My interpretation is that Johnson treated lemon juice much like he treated bitters in many cocktail situations. Schmidt is another a-dash-of-this and a-dash-of-that mixologist.

[EvanD]

I think there’s room to hypothesize that dashes of different ingredients were incommensurate. A dash of bitters dispensed from a cruet might have approximated 1 mL, but a “dash” of highly viscous gum syrup might have been closer to 5 mL, or a “glug.”

[Splificator]

That’s my working hypothesis: some dashes came out of needle-type dasher tops and others out of regular, uncooked bottles. IYKYK.

[martin]

To toss an ember on these proceedings, I just came across this statement in the 1871 The Gentleman’s Table Guide on page 40:

“Dashes” are half a teaspoonful.

[GoodSpiritsNews]

That seems extreme. Like too much salt in a food recipe.

[martin]

Yes, if you take them at their word, their Gin Cocktail contains:

1-2 tsp gum syrup
1-1.5 tsp Angostura bitters
1 tsp curaçao
1 tsp sugar

So, almost a full oz of these additions.

Interestingly, the book only uses “dashes” in four recipes, relying on teaspoons and other units in the rest.

[armin]

I am currently working on the topic of lemons from a historical perspective. A multi-part post on this will appear on bar-vademecum.eu. I wanted to share an important insight with you in advance: there is a big misunderstanding about lemons in the 19th century. When we read the recipes, we have today’s lemons in mind. But they are not the same as in the past. Anselm Josti provides an important clue in the third edition of his book “Die Bereitung warmer und kalter Bowlen” (The preparation of hot and cold punch) from 1885, in which he tells us: “One lemon provides an average of 24 g of juice.”

Once you have understood this, you can see that Jerry Thomas’ specification of half a lemon corresponds to the 2 to 3 dashes given by Harry Johnson. The amount of sugar must be chosen according to the sweetness of the lemons - or, as Harry Johnson does, he defines the amount of sugar and lets us know that we must balance the acidity of the lemon by adding more or less of it, i.e. 2 or 3 dashes.

Prepared in this way, the sour is something completely different from today’s sours, and as the guests of a befriended bar unanimously confirm: much better. So try this recipe from the Mirabeau Bar in Freiburg im Breisgau, Germany:

60 ml Eagle Rare 10 Years Kentucky Straight Bourbon

12.5 cl lemon juice, freshly squeezed

12.5-15 ml sugar syrup (2:1), depending on lemon

3 full ice cubes

Preparation: stir all ingredients in a mixing glass for 15 seconds. Strain into a very well chilled Nick&Nora glass.

[martin]

I’m excited, because—obviously–something has long just seemed “off” to me about current assumptions. I’m daunted, because if there’s a shift in assumptions, I’m going to have to go back and correct an awful lot of recipes.

[Splificator]

Unfortunately, I would take Josti’s pronouncement with a large grain of salt. As the United States Department of Agriculture noted in 1909, there was a great deal of variability in the size and juiciness of lemons, so that “the sizes vary from 180 to 540 to the box” and “the juice constitutes from 30 to 55 per cent of the total weight of the lemon.” And that was just American lemons. Add in the wide variations in skin thickness and moisture content between different growing regions and I don’t think one can use a one-size-fits-all determination for a measure such as “the juice of half a lemon.”

See R. H. True and A. F. Sievers, “Some Factors Affecting the Keeping Qualities of American Lemons,” U.S. DDept of Agriculture, May 26, 1909.

[armin]

For most pre-1900 recipes, this is “the magic formula” that just keeps coming up and works incredibly well, whether it’s sour, fix, daisy, etc., even if there are exceptions: 6:1.25:1.25
And because it’s often said that only poor quality spirits were available back then: that can’t be right, because you can’t cover up poor quality with this magic formula. On the contrary: you need excellent spirits. However, these are then also raised to a new level.

[armin]

I found a study from the 1880s in California that goes into detail about the lemons, and the results not only from here confirm that Josti’s information is basically correct, even if there are of course deviations upwards and downwards, but there are only small differences; and we are also talking about the average quantity. This can also be confirmed indirectly from other sources (by extrapolating from the weight of the case and the number of lemons it contains). It seems that large, juicy lemons, as we know them today, only became the standard at a later date. I can’t say exactly when, but it could have been around 1900 and there were certainly local differences depending on whether you were looking at Californian or French lemons, for example. If you are lucky enough to be able to buy “old-style lemons” (namely Ligurian lemons from a small citrus grower), you can see that they still correspond to the lemons of the 1880s in terms of quality and juice quantity.

[Splificator]

I do agree that lemons have gotten larger, on the whole, but they’ve also gotten more uniform. Sorrento, for example, used to grow three quite different types of lemon, with different juice yields and flavor; now they mostly grow one.

The beautiful thing about magic formulae is that everyone has a different one. I do agree, though, that current assumptions about how much citrus should go in a drink are off, or at least the common default of .75 oz (22 ml) is off. It’s just too much juice in a drink, and tends to dominate the other flavors, and is very hard to square with old recipes.

For what it’s worth, my current “magic formula” (since at least 2020, as one can see over and over in the old Lo-Fi Lush Hour drinks) is based on .5 oz (15 ml) lemon or lime–not much more than Armin’s 12.5 ml. Or I should say “formulae,” as from there things deviate depending on precisely what kind of Sour I’m making.

For a plain old Sour, I will use .25 oz (7 ml) rich simple and 1.75 oz (52 ml) spirit --or, better, I will stir .25 oz granulated sugar into the juice before adding the spirit.

For something like a Sidecar/White Lady/Silent Third/Margarita, I will add .75 oz liqueur and 1.5 oz (45 ml) spirit.

For me, those are magic.

[Pixy]

This is an interesting thread - if we’re discussing any sort of magic ratio of citrus to sugar - i have a heuristic myself but will say that it’s probably not the most accessible. I am happy to move this post elsewhere if it’s just not quite on point for the topic but i’ll post it here anyway.

When the drinks are of the more rudimentary type then my consideration for citrus is based more on how fresh it is and how acidic it appears on the day. if it’s a particularly sharp citrus that day then 15 ml will suffice and then i’ll follow the same general rule of thumb as Splificator, applying 15ml Citrus juice to roughly 7 ish of rich simple syrup.

That’s the easy part. If a drink has a liqueur then I prefer to know how much sugar is in that liqueur. for this i generally check SystemBolaget for the Bottle and check the per/100ml amount if it’s available, having worked in distilleries and having made my own liqueurs i may also just do a quick taste test to check and see whereabouts I think the sugar and abv is but this is not as exact as just making my own or referencing the system site.

If you’re making a Sidecar this becomes a particularly important thing to note as the sugar quantity of the triple-sec is going to either make or break that drink if you’re averse to adding in 10ml independantly.

this is where my heuristics set in; In a daiquiri you might choose the below recipe, booze forward, but balanced.

• 60ml, 15, 15

but a sidecar might be something like

• 45, 22.5 (3/4)triple sec , 15 lemon

The sidecar I mentioned with 10ml more sugar has only about 1.4~g more of sugar if you’re using cointreau (21.5g/100ml) different to the daiquiri if you’re using 10ml of a 1:1 syrup.

The difference of the sugar at the low end of this drink spectrum, the original sidecar and the modified sidecar is going to be enough to bring it back into a sort of more mystical “balanced” alignment.

If the liqueur is something like 26brix then 5ml of a rich simple syrup (2:1) will also get you into that ballpark of 10~g of sugar per cocktail. if you’re scaling the juice upwards to a daiquiri that is 2oz, 3/4oz, 3/4oz, you might consider that just one tbs of caster sugar is roughly 15g and that’s roughly 14.6g of sugar.

All of this is to say that these components and the amount of sugar in them are as vital to understanding how to build that historical drink as the amount of citrus, because if older Orange liqueuers before the triple-dry kind had more sugar in them, it would make sense that the sidecar was more “balanced” without the addition of any sugar, assuming the lemons were smaller and that it was closer to a 15ml amount.

If I was on a forum with any drinks history experts maybe they’d be able to find out how much sugar was in older styles of these liqueurs. It might be possible (though maybe not reasonably worth while) to map a path of specific liqueur use in drinks and listed juice and compare it to the more modern reccipes and see if the old and modern drinks comparative total sugar has a similar ratio?

If i had to bet any money - it’s that the “liqueuers” were sweeter back then and the ratios of juice to total sugar per drink were probably similar. I’m also ready to assume that there’s an “ideal” sweetness and sourness balance that is appropriate for a majority of people. if that’s 50:50 sugar to acid component that wouldn’t be surprising either and why it’s such a mainstay of drink ratio formulation. i’m sure large companies already have that ratio in use in all their soft drinks, so historical drink only need to scale and ratio the total drink sugar to the amount of core spirit and that could easily be used to inform the amount of juice to use to make an acceptable (palatable) replica.