{edited for corrections, so early readers should re-read - hard to compose and do the math correctly at the same time
][further edited to make a bit more readable.]
For those that might be buying 40% (versus those that have been successful and roll-searching wild ones, with no cash exchange), keep in mind the wear.
First recall wear on 90% silver.
The "normal" tolerable wear on 90% by accepted standard is 1.1% (715/723). Based on a standard 90% half which left the mint at nominal weight, this amount of wear would result in a coin that has worn during circulation to 12.36 grams. A .14 gram loss of 90% silver, or a .126 gram net weight loss of silver. Not really a lot of wear, but it usually results in a noticable change in appearance without being "slick". Coins that weigh less than this are considered below weight for trade. Unless you are focusing on AU/BU coin, it's always best to trade circulated coin based on weight because of wear, even with 90%.
Now, 40%.
I just pulled out a worn 40% cull, which didn't satisfy my eye enough to stack in a roll, and weighed it - 11.29 grams. 40% halves are nominal at 11.5, so this one is 0.21 grams below nominal weight. It's hard to tell how much of it resulted from wear versus an underweight planchet at the start, because (as weighing experiments show) there was a lot of weight variatino in the planchets. The ultimate determination of minor wear - on an individual coin - is visual appearance rather than weight. The weight difference from nominal on this one coin is definitely below the 90% wear threshold, but if this was wear alone the net affect, percentagewise, is far worse for 40% because of how these coins were constructed. Recall that while they are 40% overall, they are a clad laminated coin consisting of two thin layers of 80/20 silver-dominated Ag/Cu alloy surrounding a thick layer (two-thirds of the total mass) of copper-dominated 20.9/79.1 Ag/Cu alloy (which is often represented as a 20/80 alloy). While the total silver content is 4.6 grams, almost all of the silver loss is located in the high-silver layers, so the effect is multiplied.
Using my 11.29 gram coin as an example, the 0.21 grams of weight loss contained 0.168 grams of silver, which proportional to the 4.6 gram nominal silver content is 3.65% silver loss.
Comparing the 90% criterion to 40%, to have the same proportional level of assurance of silver content, the equivalent weight loss limit for 40% would have to be 11.436 g per coin or on a roll basis 228.73 grams for 20 coins versus a 230 gram nominal roll. Anything below that would (on average) have more wear than the "standard" amount.
I just weighed a random stack of low-wear 40% coins and they weighed 229.88. This weight confirms they are low wear coins, even though the individual coins in this roll have variation in them that are both below and above nominal.
So a 90% roll can tolerate a 2.76 gram loss and remain at the limit of tolerable wear while a 40% roll - if your desire is to maintain the same percentage of silver content retention - can only tolerate a 1.27 gram loss.
- 40 percent silver loss.jpg (69.57 KiB) Viewed 1094 times
Edit 9/27/2013 and with more emphasis on 12/15/2013: To emphasize, the above is based on nominal weight planchets. However the US mint didn't control the blanks of common circulating coin for the purpose of weight to the same standards that are more common today for bullion silver. Based on measurements of actual low wear coin of various types, face values, years, there was obviously a tolerance allowed on an individual blank as long as a sufficiently sized population of them averaged to some total value. It is therefore common for nominal roll weights of coin to be very close to the value that would result if the roll weight was calculated from a nominal coin weight, but any individual coin can see a substantial variation from nominal. I had previously evaluated this informally for 90% 64 JFKs and saw a surprising level of variation in that coin. Having just looked at a semi-random population of 65-69 (but most from 66-68) halves that same variation exists but at an even higher level. Even though the random draw rolls are near nominal in weight, I found that individual "underweight" coin (11.42 g or less) is on the order of 30% of the total population, even though most of this coin is of low wear based both on visual evidence and random roll weights. Unfortunately it is not possible to determine from these basic measurements whether or not the planchet variation is in the high silver layers or the high copper core layer (or both). Out of this population of >120 coins most were in the 11.40-11.59 range, but there were substantial high (11.60+) and low (11.3x for the most part, but some in the 11.2x range) outliers. The two high outliers weighed at 11.76 and (incredibly) 11.95 - both from 1967. The three low outliers were 11.18 (a 1968), and two at 11.20 (both 1966). Because of the above variation - apparently present in the raw blank and not due purely to wear - I would be cautious about presuming that a low weight coin is always a high wear coin. The best approach to evaluate wear for 40% coin appears to be based on visual evidence rather than weight. However, obviously weights below 11.20 would be very suspect.
9/28/2013 edit
I just weighed nearly 800 coins in the 65-69 years, sorting into high and low weight piles by year, using 11.43 grams (used because it would be the discrimating value for wear based on the nominal weight) as the fulcrum.
- Code: Select all
Year Light Heavier Percentage Light
1965 10 51 16.4
1966 34 57 37.4
1967 1 202 28.6
1968 79 146 35.1
1969 40 79 33.6
31.3 Overall
My overall "feel" from yesterday was correct; right about 30% coins are below the 11.43 gram mark and 70% are above. Taken as a population these are collectively very near the nominal weight, on average, but there is a fairly broad spread of planchet weight. The low of the population was a 1965 with an amazingly low weight of 10.81 grams (0.69 grams light), and the high was the heavy 1967 (noted above) at 11.95 grams (0.45 grams heavy).
