In a previous BNS Blog note the die numbered shillings of 1865 were used to validate the statistical methods used to predict the numbers of dies used for a coinage. The 78 different die numbers gathered there have now been supplemented to produce an almost complete set of die number images for 1865 presented here.
Analysing the data for the shillings for the whole of the die number period, including Royal Mint annual reports, reveals a systematic year-on-year variation in the numbers of coins struck per die. The production data for the obverse and reverse dies track each other almost exactly over the whole period. This can only be the result of a variation in the die manufacturing process or Mint practices – “The Die Number Experiment”. The 1870 Royal Mint report cites changes in the die hardening process resulting in increased durability, the result is visible in the data.
Further analysis of the data for other denominations and the Mint reports might reveal the process parameters that were being varied during the experiment.
BNS Research Blog 
Quick update: some doubt has been cast about the newly discovered die 50, as possibly being a worn and damaged 30. Investigations continue.