This is turning out to be a great
thread (I can already see it being stickied..), thanks to Nikos for starting it. So, for the sake of completeness, I'll throw in other flan-manufacture aspects as I think of them
First the "
Stannard Al-Marco Gouge". Rough gouges appear on many
Roman silver coins (example below). What is not generally understood for the new ancient coin collector, is that these were
applied for
weight control to the FLANS before striking, and are not the result of later damage. What is not often understand by even experienced collectors is that the flans were not weighed individually, but rather a batch of say 84 flans was made from, say, 13 ounces of silver, pouring them by hand into open moulds. Once the last few flans were reached, depending on how much silver was left, the later flans might be randomly a little heavy or light. Then the whole lot of 84 flans were put on a
scales, and random flans were taken out, one by one, gouged, and thrown back in the
scale until the batch weighed precisely 12 ounces (1 pound). Why do it this way rather than weigh individual flans? Because it's much much quicker, and it ensures that the random process of making individual flans doesn't result of you being short of the required number of
denarii once you reach 12 ounces, or having a lump of silver left over after you've already made 84 flans. It's a brilliant control mechanism.
In Clive
Stannard’s own words: Ancient mints sometimes adjusted
weights by gouging a sliver, occasionally slivers, of metal from the
face of a
flan, before striking the coin. The results are characteristic and easily recognizable. Examples in silver are known from
Lycia,
Paeonia (
King Adoleon), Velia and the
Roman Republic, and there is a gold example in the coinage of
Constantine I. This coin below is a clear example of a gouged piece.
The frequency of the use of gouging in the
Roman Republic makes it possible to study whether
weight adjustment was carried out al peso (which means that each individual
flan was brought within the tolerances of the
weight standard for the issue), or whether it was done al marco (which means not paying too much attention to the
weights of individual coins, but ensuring that a fixed number of flans were made from a fixed
weight of metal). This question can be investigated by looking at the histograms of large number of
denarii, in issues known to be use gouging. In al marco, adjustment, a block of flans is
cast a little heavy. The right number of flans for the desired
weight of coins is counted out (and the overall
weight will, of course be, too heavy). Flans that look heavy are successively picked out one by one, without too much attention to the
weight, and a sliver of metal is gouged off. The gouged flans are tossed back into the block, until the overall
weight is
reduced to the right overall
weight. The following figure models this process. As a result of adjustment al marco, the histogram is negatively skewed (the size of the upper
leg has been
reduced), and has high kurtosis (the centre of the histogram is higher than a normal distribution.)
8,649
denarii from between 144 and 43 BC were checked, to identify issues with gouging. 1.34% of all the coins looked at were gouged. The
weight histograms of 4,587
Roman Republican denarii in the issues known to be gouged was negatively skewed, with high kurtosis, showing that they
had been adjusted al marco. In these issues, 2.53% of the coins showed signs of gouging.
from Clive
Stannard, "The adjustment al marco of the
weight of
Roman Republican denarii blanks by gouging, in M.M. Archibald and M.R. Cowell (eds), Metallurgy in
Numismatics, Volume 3, pp. 45-70,
London, 1993