The brewing logs tell what the brewer intended his beer to be. The gravity books what was actually served in pubs. The two rarely match.
Frederick Accum did something similar at the beginning of the 19th century. He sampled beers in the great Porter breweries and then in pubs. One average, by the time they were served, beers were 10 -15% weaker. A century or so later, not much seemed to have changed.
Samples of brown stout with which I have been obligingly favoured, whilst writing this Treatise, by Messrs. Barclay, Perkins, & Co.—Messrs. Truman, Hanbury, and Co.—Messrs. Henry Meux and Co.—and other eminent brewers of this capital—afforded, upon an average, 7,25 per cent. of alcohol, of 0,833 specific gravity; and porter, from the same houses, yielded upon an average 5,25 per cent. of alcohol, of the same specific gravity*; this beer received from the brewers was taken from the same store from which the publicans are supplied.
* The average specific gravity of different samples of brown stout, obtained direct from the breweries of Messrs. Barclay, Perkins, and Co. Messrs. Truman, Hanbury, and Co. Messrs. Henry Meux and Co. and from several other eminent London brewers, amounted to 1,022; and the average specific gravity of porter, from the same breweries, 1,108 [I think this should be 1,018].
It is nevertheless singular to observe, that from fifteen samples of beer of the same denominations, procured from different retailers, the proportions of spirit fell considerably short of the above quantities. Samples of brown stout, procured from the retailers, afforded, upon an average, 6,50 per cent. of alcohol; and the average strength of the porter was 4,50 per cent. Whence can this difference between the beer furnished by the brewer, and that retailed by the publican, arise? We shall not be at a loss to answer this question, when we find that so many retailers of porter have been prosecuted and convicted for mixing table beer with their strong beer; this is prohibited by law, as becomes obvious by the following words of the Act.
"If any common or other brewer, innkeeper, victualler, or retailer of beer or ale, shall mix or suffer to be mixed any strong beer, ale, or worts, with table beer, worts, or water, in any tub or measure, he shall forfeit £50." The difference between strong and table beer, is thus settled by Parliament.
"A Treatise on the Adulterations of Food" by Frederick Accum, 1820, pages 172 - 175.
The first thing that tipped me off to this possibility was the variation in gravity for different samples the same beer in the Whitbread and Truman Gravity Books. With the exception of traumatic periods like wartime, brewers didn't fiddle with the gravity of a beer from month to month. It stayed constant for years on end. I know that from the brewing logs. And they were professional enough to be able to hit the target gravity pretty much 100% or the time.
Sorry, I should have said variation in the gravity of draught beer. Bottled beers are much more constant. Which is a real giveaway as to the underlying cause: adulteration. The landlord can't really fiddle with bottled beer. Not so with draught beer.
Let's look at an example from Barclay Perkins. Here are five examples of their KK, or Burton, all from 1936. Four are from the Truman Gravity Book, one from the Barclay Perkins brewing logs.
The beer as brewed was a nice round 1056 gravity. The pub samples vary from 1048.8 to 1055.9. Only one matches the brewery gravity. As these have been calculated, we need to allow from some margin of error. So I'm willing to give the 1053.9 sample the benefit of the doubt. The other two are far too low. 10% and 13% too low to be precise. Something has clearly happened to the beer. But what?
My guess would be slops of weaker beer being put into the KK cask.
But take a look at the figures for Barclay Perkins ordinary Mild, X Ale:
Here all the pub samples have a higher gravity than the brewing log. How can this be? When I was younger, a reason people often gave for not drinking Mild was that the slops from all the other beers were put into it. Being dark, it was hard to spot. I was sceptical about this theory at the time. But that's exactly what these numbers suggest. As X would usually be the weakest draught beer, slops from anything else would raise its gravity.
This begs the question. Adulteration seems to have been common between 1800 and 1940. Does it still occur now?