CJJ Berry's First Steps in Winemaking has been a most useful reference & recipe book - if at times a bit hard going with the theory.
For example, I've always thought that he spends forever banging on about whys & wherefores of testing the specific gravity (SG) of a wine - & I equally always thought that it was easy peasy just to test a sample of wine at whatever stage, & read the number off the hydrometer, which in turn allows you to measure the alcohol content of the wine.
Then I had a bit of a lightbulb moment as to why what I'm doing isn't exactly accurate...
When you start off with your 'must' in a bucket, it comprises fruit (which contains some natural sugar), extra sugar (&/or grape concentrate) to bump the sugar level up to give the yeast something to feed on & water making a total of about a gallon (or so), & then you measure a sample of this which gives you the starting SG.
What you are measuring is how sugary the 'must' is - the more syrupy/sugary/gloopy the 'must', the higher the SG reading will be. Water measures 1.ooo - a nice sugary must will be somewhere in the region of 1.100.
As the yeast scoffs the sugar it converts it into lighter-than-water alcohol meaning that the wine because less heavy & sweet, & more light & alcoholy, & the SG falls. The yeast eats sugar until it runs out & starves, or it has produced so much alcohol that it poisons itself, & dies (a lesson to us all, I think!).
At this point the fermentation has stopped, & the alcohol level in the resultant wine is the difference between the start & finish SG divided by 7.36.
But - & here's the rub - the volume of liquid that you measure the SG of must be the same at the first & last readings or else you have changed the goalposts & the final reading will not be accurate. I always have less wine at the end than I had 'must' at the beginning because every time the wine is racked you leave some behind (with the settled out 'yuk'), so the overall volume falls.
So where does that leave me? I have settled on making each batch of wine = 4.75litres - I've measured and marked the demijohn with the level. When I siphon it off each time & lose a bit, I'll make the bit I've lost back up to this volume with water, then measure the SG (you should keep some of the original must to one side to make up any shortfall, but life's a bit short).
The wine should taste the same - slightly diluted will just mean it's less lethal - & this way I'll always be able to produce six bottles per batch - with a bit left over as 'perks' - but the label on the front showing the strength will be rather more accurate...