On Sherry Molecules: Or, Sherry is Forever

James Bond: An unusually fine Sholera. Fifty wun, I believe.

M: There is no year for sherry 007.
James Bond: I was rufurrung to the orruginal vintuge on which the sherry is based,  shir. Eighteen fifty wun. Unmishtakable.

- Diamonds are Forever (1971)

It's a bit like asking how many angels can dance on a pin-head; how long can the original fill of sherry remain in the solera if you remove a third each year and top up with younger wine ?

The Wrong Answers

The schoolboy error is to say three years: a third each year.

An equally erroneous answer, in my view, is to say forever. This is called the asymptotic argument which says that even if you keep halving something there will always be at least a tiny bit left.

This works for infinitely-divisible things, but sherry molecules are indivisible, so eventually you'll come up against the last molecule and have to remove it from the solera.

The Right Question

So, the question is: take the number of sherry molecules in a solera, keep taking away a third and when the result you get is less than a whole number, then that's your answer. On average.

To put it another way: 2/3rds to the power x times the number of molecules in a solera equals just less than 1. What is x?

The Solera System Explained

Not to be confused with the solar system, the solera system is one of the key elements of sherry-making.

A third of the wine is bottled and the solera is topped up with younger wine, giving a consistency of style.

It also means, as M points out in Diamonds Are Forever, that all sherry is non-vintage.

The Romance of sherry - as Davy Kurniawan tells me - is that there is always a bit of the original sherry left in the solera.

However, if you keep taking away a third of the wine, eventually you will get to the point where no original wine is left - in my view at least.

A Worked Example

Let's say you put nine apples in a box; take out a third and replace them with bean bags.

Now take out a third of the things in the box - on average you will take 2 apples and 1 bean-bag (there are twice as many apples as bean bags).

Replace with bean bags and you have 4 apples and 5 bean bags.

Repeat and assume that you take 1 apple and two bean bags. You are left with 3 apples and 6 bean bags.

Keep going and eventually, you'll have no apples left - your original supply has run out.

So it must be with a solera - eventually, all the original sherry molecules must be removed.

The Asymptotic Argument

The asymptotic principle states that the distance between curve and axis approaches zero (but never reaches it) as it tends towards infinity.

As an example, Davy suggests imagining continually halving something; you never end up with zero because you always have the other half left. Same principle with sherry.

I get the point about continually halving; but sherry molecules are indivisible - you can't have half a molecule.

So eventually you get to a point where you cannot take a fraction of a molecule out - you have to take a whole molecule out and that molecule is the last one.

The Twittersphere Responds

I asked the question on twitter and got a number of different answers - sometimes from the same person:

Dr Jamie Goode, a science PhD, originally said the original sherry will last forever, you just keep depleting (the asymptotic argument). When I pointed out that sherry molecules are indivisible, he suggested it would take thousands of years.

Jeremy Rockett chipped in to say that Nobel Prize-winning economist Milton Friedman, on visiting the Tio Pepe bodega, calculated that there are 50-60 vintages in a glass of Tio Pepe Fino.

Of course, only an economist could spend time in a sherry bodgea working out the number of vintages in a glass.

But I'm not sure that the actual number of vintages in a glass is the same as the theoretical time it would take to deplete the original fill.

At this point Tio Pepe themselves contributed to the scientific discussion by sending me a picture of a solera with Milton Friedman's signature on from 1976 - the year he won the Nobel Prize. Nice

Tony Milanowski, an Australian winemaker and lecturer in Oenology at Plumpton College, then weighed in to say Milton got it wrong and the actual answer is 158 years.

If this is correct, then Bond's 1851 solera (actually, it was M's sherry, but Bond's observation) would, until a few years ago, still have had some molecules of the original fill. But as the new M seems to prefer whisky, where the 1851 sherry is now is anyone's guess.

Whether any of the character of the original fill would remain is another question entirely and perhaps a moot point - as Andy Hickey points out.

Throughout this discussion, I have been referring to sherry molecules as if they existed as a standardised, homogeneous compound.

Actually, sherry is of course about 80% water, 20% alcohol and some complex flavour compounds; a bit like using ping-pong balls, apples and footballs in my earlier example.

Godfrey Spence adds these comments:

First, only the cheapest Sherry will be run with a 1/3rd saca. Other than Pale Cream, even the big sellers are less than that (Tio Pepe I think is 20%, La Gitana a tad less). The age-dated wines are taking a tiny proportion from the solera each year: under 5% for VOS, and under 3 % for VORS.

Second, surely the issue is not how much of the oldest but the fact that there is also, in every bottle you buy, a tiny bit of the second oldest, and the third etc. So if a solera had been set up in, say, 1900 every bottle will have a percentage, however small, of every vintage of the last century.

All this assumes that the solera was set up from scratch and has never been added to from other soleras, which is a common, and inevitable, practice in the region

My final tweet on this subject was to the one person I know who has the intelligence, curiosity and knowledgeable contacts to work the answer out, the Principle of the Stephen Perse Foundation, Tricia Kelleher. She passed it on to her Head of Chemistry who had first to look up what a solera is, then gave this answer:

"If 1/3rd is removed each year then the amount of original remaining is 2/3 raised to the power of the number of years; a smaller and smaller amount approaching but never equalling zero [this is the asymptotic argument - TL].

One of the other tweets commented that eventually you must get down to a fraction of a molecule but performing a simpler calculation using a pure substance like water then a 1 litre bottle contains approximately approximately 6 x 10 to the power 25 molecules so it would take over 10 to the power 17 years to get down to only 1 molecule of the original water! Of course it's random chance as to which molecules are removed.

For now, I leave the last word to Davy:

I think you should always have the romance in your head, that there is an infinitessimal amount of the original sherry there - though perhaps in homeopathic quantities.

Quite.

With thanks to Dr Jamie Goode, Dr Davy Kurniawan, Jeremy Rockett, Tony Milanowski, Andy Hickey, Godfrey Spence and Tricia Kelleher for all their thoughts and contributions on this topic: I hope I have represented them accurately and credited appropriately.

Other related articles
On Sherry's Image
The Great Sherry Tasting


Image credit: El Pantera