Bixley

This is what the postie delivered today

This is what the postie delivered today

Is that a time lapse picture of one coin moving further and further away? 😁
Nice pickup  

This is what the postie delivered today

This is what the postie delivered today

This is what the postie delivered today

This is what the postie delivered today

This is what the postie delivered today

Thank goodness the coins are not made of Fe, otherwise they truly would be Ironic 😉

https://digital.csic.es/bitstream/10261/136506/4/Palomar_A_Comparative_2016_ Journal of Cultural Heritage.pdf
I've found this article when browsing around. The authors compared different ways of conserving silverware, which applies equally to coins.
I will try to summarise the article in a way that is friendly to people with little background in chemistry.
The article tests three kinds of methods of silver conservation:
Physical (wiping, rubbing, polishing, etc) which are straight out for coins.
Chemical, which the article defines as dipping, to differentiate from "electrochemical" methods.
Electrochemical, in plain English the baking soda+hot water+aluminium method.
Below I state as facts:
1) you should never try to meddle with a coin physically however hopeless it looks
2) toning is metals (primarily silver and copper) reacting with chemicals in the environment. Primarily, with hydrogen sulphide aided by oxygen, but many other substances also react with silver with ease: cyanides (if you smoke actually), chlorine, and many other stuff
3) toning is the same as corrosion and is irreversible. Silver sulphide (and other toning products) have different crystal structures from the pure metals.
4) dipping is removal of silver sulphide from the coin using acidified thiourea, whereas soda removes sulphide and leaves silver behind, with permanently changed crystal structure.
5) lustre is light scattered by porous surface of a coin. An ideally polished coin (the perfect proof) should have no lustre at all.
6) according to the article and my own experience, both dipping and soda remove tarnish from .999 silver almost completely, but soda works less efficiently for .925 or .900 silver. Some thalers have .8xx purity and basically cannot be helped with soda. Both methods work well on reddish-purple gold coins (if you don't like the reddish toning on old gold), but for very deep copper spots or greenish spots on some old .900 gold, only dipping works.
Below I add some context in terms of chemistry (you don't have to understand this bit)
The most important rule governing our lives and the entire universe is the second law of thermodynamics, and in very simple language inaccurately reads: things will spontaneously go to the less ordered state. This means coins start to corrupt the second they come off the press. At the early stage of corruption the appearance is minimal. Some lucky coins corrupt in a way that enhances their eye appeal. 
The main suspect of corruption is oxygen and hydrogen sulphide (which smells like bad eggs). They work together to give silver a darkened look (from a layer of silver sulphide); colourful toning results in different depths of said darkened layer (light diffraction like rainbows). The process is analogous to aqua regia dissolving gold, and has something in common with the dipping process (what a surprise)
Where oxygen acts as the oxidant, and hydrogen sulphide provides a ligand. In the dipping process, thiourea is the ligand. In the aqua regia process, nitric acid is the oxidant, itself becoming reduced giving the brown fume, and hydrochloric acid provides chloride which acts as the ligand. Metal ions that are insoluble in water forms a complex that is soluble, and therefore some metal is removed.
The baking soda method, on the other hand, is essentially a battery. Silver sulphide is reduced and aluminium is oxidised. The sulphide ions leave the surface of a coin to carry charge and combine with hydrogen in water to become hydrogen sulphide, a gas, which leaves the solution forever (2nd law in action).
To summarise, dipping does remove some metal from a coin's surface, and the baking soda does not. Neither methods restores the coin to a state prior to toning.
Below I comment on some common misconception in the numismatic hobby. This section may contain some personal opinions, and some viewers may find it offensive. 
1) in theory, over-dipping does not remove precious metals from a coin and cause a loss of lustre, because thiourea is not an oxidant. You need an oxidant to work with the dip to remove gold or silver from a coin. In reality, however, oxygen in air is sufficient in helping thiourea remove gold and silver. It is therefore important to dip very briefly if at all.
2) toning is but a marketing whitewash for corrosion. Attractively toned coins will not stay attractive forever, and your best hope is removing oxygen and sulphur from the coin to "freeze" the coin if you like it. Honourary mention to Lighthouse Intercept.
3) a coin is not "original" the second it comes off the press. It is untrue to say a toned coin is "original" and a dipped coin is not. If toning is light, dipping will make it look "original" without harming the coin too much. If toning is deep, then dipping will leave behind an ugly cloudy look, but it is unfair to say dipping has damaged the coin: toning has.
4) there is no difference between natural and artificial toning in a chemical sense. Physically though, artificial toning tend to be shallow. Because of this, artificially toned coins tend to be more colourful. Natural toning, on the other hand, will develop much deeper into a coin's surface and look darker. Colour often appear only at certain angles. This is why some NGC or PCGS pictures look beautiful with the actual coin ugly in hand, but those are most likely natural. Beautiful AT can be removed without ruining the coin, and ugly NT has ruined the coin already, whatever TPGs and CoinTalk puritans say.
I'd like to conclude with some very personal opinions on what coins you should or should not try to conserve.
Dipping or soda method effectively remove hazy toning on old proof coins. You'd better conserve old proof coins before it's too late. If you're uncomfortable doing that yourself try NCG or PCGS Conservation. This is because, as said above, toning permanently disrupts a coin's surface, rendering a proof coin no more. This is perhaps truer for old proof gold coins than for proof silver, and given the expensive nature of such coins, make sure you know what you're doing or hand into TPG conservation.
Copper spots on gold coins will eat into the coin. These should be removed when it's not too late.
Red-to-purple toning on gold coins is beautiful. You may leave as is or soda the coin to make it shiny gold, and it's a matter of personal preference. Don't dip if soda alone works efficiently.
Thou shalt not remove beautiful natural toning from circulation-type silver coins.
If the toning on an old silver coin is ugly, the chances are it's hopeless already. Don't pay too much for those coins even if they're MS63. But sometimes an ugly coin can be successfully restored. You should assume an ugly coin is hopeless and only conserve to try your luck.
I would happily listen to opposite opinions, and I gladly accept correction on my chemistry. I hope at least some people will gain better understanding of the nature of toning and conservation. It's very important we know what goes on before we decide what to do (and not to do). There is actually quite a lot of myth among the numismatic community, and it's important we understand the hobby.

They used master tools to stamp the dies then hand carved some details. I guess this could be the reason. 
It's actually easier to carve things raised than incused, because they'd know the relief of different parts of the design. The master tools were handmade and hence all the differences.
https://www.royalmintmuseum.org.uk/collection/master-tools-and-dies/
These will need to be reworked from time to time, but far less frequent than the dies were remade. Say a punch lasted to make 100 dies, and each die was used to make 10,000 coins or more, for every million coins you remade the die 100 times and the punch only once.
The master tools were also used to counterstamp the spanish silver in the 1790s and until 1804.

Automatically validating sovereigns was a fairly mature technology in the 19th century - they were allowed about 60mg of wear by law and then they were returned to the Bank of England, which would exchange them for new ones and send the old ones back to the mint for recycling.  By the latter part of the 19th century the banks all had machines that could automatically weigh sovereigns by their thousands and sort out underweight ones, and could do so very accurately.  I dare say they would reject most fakes as well, unless it had been done with great fidelity.  Most fake sovereigns would fail on dimensions and/or weight as this was well before the days of tungsten cores. However, I have no idea how accurate changing machines like the one depicted above really were.
Here's a 19th century engraving of the machines at the Bank of England.  I think it's worth noting that a substantial amount of R&D went into the sovereign - the technology to mint them by the million to precise specifications, and all the infrastructure that went around circulating them.  They were quite a major innovation in their day.
 

The pickling process noted above was first used by the Romans. It allowed them to debase the silver denarius yet still produce perfectly silver-looking coinage. The 22k blank is blanched in acid, which results in the surface copper leaching out and only the gold remaining.  Then when it is struck, the surface will look gold, but the coin still is made of an alloy. If the coin remains in pristine condition, it will continue to look gold, but if it is sufficiently circulated, then the wear will rub past the more pure external gold surface and into the alloy, where the coin can look more rose gold in color. The pickling described explains 100% of this phenomenon. The Royal Mint could do 18K Russian gold and the pickling would still yield a yellow gold coin. The imperial Romans got a 20% silver Denarius to look just like the 95% Denarius of the Republic with this method. The acid removes surface copper and leaves the gold.

With all this in mind, I'm ordering my William V coins now... 😁

Oh hello, what's this then?!

The article no longer exists, but Google's cache reveals...

The Seymour Unicorn joins the rest, courtesy of Chard’s.

Parrot? Sulphur-crested cockatoo. 😀

The Seymour Unicorn joins the rest, courtesy of Chard’s.

Another unexpected pickup.

These are surprisingly nice examples. I never realised some were packaged like this but they certainly seem to fare better like this than in the black foam capsules.

The Seymour Unicorn joins the rest, courtesy of Chard’s.

The Seymour Unicorn joins the rest, courtesy of Chard’s.

The Seymour Unicorn joins the rest, courtesy of Chard’s.

The Seymour Unicorn joins the rest, courtesy of Chard’s.

The Seymour Unicorn joins the rest, courtesy of Chard’s.