Weight variation in older U.S. coins

Quick question: does anyone know how much weights tended to vary amongst genuine specimens of early 19th century coins (say from 1800-1830 or so)?

The reason is that I just got started weighing some of my old pieces with a scale and I was surprised by how much variation there was from the "official weights" published on the PCGS website. These are mostly low-grade, common date copper coins and varieties that I have no reason to suspect are fake, and many were 10% or more underweight!

They seem to be the right diameter so I doubt this is due to shaving of the edges -- maybe they were made of impure copper? If an old copper coin weighs less than 90% of what PCGS says it ought to, should I be worried? Or less than 80%?

Has anyone compiled statistics such as: "for this type of coin, the average weight is 5.44 grams, but 95% of genuine specimens weigh between 5.2 and 5.49 grams"? (For instance, I just made up the ranges, but 5.44 g is what an old Draped Bust Half Cent should weigh.) If so, this would be really useful to know!

 

Log in or Sign up to hide this ad.

I think this is in the Coin World Almanac. Someone with a copy at hand will have to verify.

 

The chart has been posted here from time to time -- here's one such post:

https://www.cointalk.com/threads/coin-weight-tolerances.380149/#post-7524440

Note, though, that there are some inaccuracies in the chart. It claims a weight of 2.527g for Barber/Mercury/Roosevelt silver dimes, but every other reference I've seen claims 2.500g target weight.

It looks like the tolerance for large cents and half cents was more than 4% in either direction. In my experience, coins worn down to AG or lower can lose significant weight, although it's more pronounced for smaller coins. (I've weighed slick Barber dimes that were under 2g.)

 

Thanks for the chart! But as someone else commented on that other thread, I'm a little perplexed about where this data came from. Are these the maximum weight variances that the U.S. Mint itself "tolerated"?

Particularly for the earlier coins I'm interested in, color me skeptical that in 1796 they were weighing each minted Liberty Cap Half Cent with balances accurate to the milligram, as this chart would suggest, and throwing out the "bad" ones. Weighing my own coins with a modern electronic scale several times successively, I tend to get random variations of around 4-5 milligrams.

 

Remember that the more a coin is worn the lower the weight will become.

 

why skeptical? The ability to do precisely that is as old as coins themselves.

 

The chart is surely guilty of displaying excessive significant figures, and the original weights and tolerances were most likely in grains or something of the like.

It doesn't take much technology to build a balance that's sensitive to differences of a few milligrams. Once you have that:

Find or make two coins that match in weight to the limit of your balance's sensitivity.

Take one and file it down until it's as light as you're willing to accept. That's your lower limit.

Take another and add metal to it until it's as heavy as you're willing to let go. That's your upper limit.

Put the coin under test onto one side of your scale, and the light one on the other. If the light one goes down, your coin's too light. Now, put the heavy one on the other. If the heavy one goes up, your coin's too heavy. Neither happened? Your coin is within tolerance. Right?

 

Exactly! To be clear my annoyances with the chart posted previously are (a) that it is unsourced and hence hard to verify, and (b) that the tolerances seem anachronistically precise (like if I said "the population of Boston in 1700 was approximately 6,713.41," which is technically true but misleading).

Thanks for the replies, which inspired me to look up the old U.S. coinage laws. No need to just speculate since the U.S. Mint has helpfully posted the Coinage Act of April 2, 1792 here:

https://www.usmint.gov/learn/history/historical-documents/coinage-act-of-april-2-1792

Relevant clauses:

(From Section 9) "Cents—each to be of the value of the one hundredth part of a dollar, and to contain eleven penny-weights of copper. Half Cents—each to be of the value of half a cent, and to contain five penny-weights and half a penny-weight of copper."

(From Section 18) "Be it further enacted, That from every separate mass of standard gold or silver, which shall be made into coins at the said Mint, there shall be taken, set apart by the Treasurer and reserved in his custody a certain number of pieces, not less than three, and that once in every year the pieces so set apart and reserved, shall be assayed under the inspection of the Chief Justice of the United States, the Secretary for the Department of State, [etc.]"

"... it shall be found that the gold and silver so assayed, shall not be inferior to their respective standards herein before declared more than one part in one hundred and forty-four parts, the officer or officers of the said Mint whom it may concern shall be held excusable; but if any greater inferiority shall appear, it shall be certified to the President of the United States, and the said officer or officers shall be deemed disqualified to hold their respective offices."

This confirms what other commenters above have said, that they'd pick three coins to test each year (at least amongst the gold and silver pieces) and weigh them, and that they took minting underweight coins to be a fireable offense.

In the case of, say, a Half Disme [sic], for example, the Coinage Act didn't really specify the exact weight of the coin but rather how much silver it should contain, set at "eighteen grains and nine-sixteenth parts of a grain of pure of pure [silver]," or about 1.2028 grams. The unsourced chart linked above lists the specified weight as 1.350 grams, which is approximately equal the "twenty grains and four fifth parts of a grain of standard [i.e. slightly adulterated] silver" of the Coinage Act. The Coinage Act gives a tolerance of 0.694 %, which would be about 9 milligrams for a 1.35-gram Half Dime.

The weight tolerances for copper coins (Cents and Half Cents) were not specified in the Coinage Act of 1792, nor even in the "Act to provide for a Copper Coinage" signed by Washington a month later (see https://govtrackus.s3.amazonaws.com/legislink/pdf/stat/1/STATUTE-1-Pg283.pdf), but surely these were codified somewhere not long after.

 

So, you have an academic or research background, too?

I raised the same "unsourced" objection the first few times I saw the chart posted, until someone actually came up with the secondary source the chart itself was copied from. But I don't think that source specifies its primary sources, and I still haven't seen any explanation of the anomalous weights listed for some issues (like the dimes I called out previously).

The best I can say is that so far I haven't seen any reliable sources contradicting the tolerances listed in that table. I've seen legislation confirming some of them (like the statutes you quoted), but not all of them.

 

I checked some of them last night and for a few examples the tolerance worked out to fairly exact percentages, for example 4%. Others seem wonky - why does the raw weight tolerance stay the same while the standard weight changes?

I was surprised to learn that during the Bronze Age (3000-1200 BC), Mesopotamia had the capability to weigh things with an accuracy to 0.14g (1/60th shekel). https://www.penn.museum/sites/expedition/hanging-in-the-balance-2/

 

Or any primary sources?

You would need in some cases to go back to the legislation actually passed to see what was then removed from the consolidated current statutes.

For example, the 1851 trime:

https://www.govinfo.gov/content/pkg/STATUTE-9/pdf/STATUTE-9-Pg591.pdf



One sentence of 3 clauses, one sentence of 2 clauses, vs. the 300 page bills of today.

 

12 3/8 grain = 0.801886511g
1/2 grain = 0.0323995g

Thus the commonly quoted

0.8g +/- 0.03g
0.750 +/- 0.005 fine

But don't miss the 1 pennyweight or 1.55517g per thousand pieces

 

Reading more thoroughly these posts by @Burton Strauss III and @jrg79, I trust the chart more. Numbers for early coins are in multiples of grains and fractions of grains. A grain is 0.06479891 grams. 12.375 grains rounds to .802 grams, the original weight of the 3c. The tolerance, half a grain, is .032 rounded. So I'm pondering why legislation would specify a 3/8 grain fraction in the coin weight. Anybody know? My hunch is that it's something to do with the specification of the alloy in relation to the coin value. The seated dollar is 412.50 grains, and the tolerance 1.5 grains, so the 3c fraction of that is exactly 12.375 grains. So the odd 3c fraction derives from the half-grain in the dollar spec.

I'm wondering how they would measure whether a 3c was below 11 and 7/8 grains. Did they really have weights in 1/8 grain increments, or a 12 grain weight and a pointer on the balance scale showing how far light or heavy? I couldn't find an image of an actual assay scale (or the weights) from the early mint.

It's interesting how this stuff presumably goes back to Mesopotamia, where the measuring system was based on grains of barley, and they may have used physical grains

 

Also speculating here, but they wouldn't need to have weights in 1/8 grain increments. They could have just picked a sample of 8 coins and weighed them against 95 whole grains.

More generally, in reading these old documents, I'm struck by how even up through the mid-19th century, there was this strong tendency to think in terms of fractions (and preferably fractions with nice even denominators) rather than decimally, the opposite of how most people tend to think in 2024. The older specifications were full of things like "one part in 144," or "four and two-fifths" when probably most of us would write it out (anachronistically, IMO) as 4.80.

Which I don't think is intrinsically any better or worse, but if you're using an old balance there are clear advantages to using "nice" fractions like 3/8 or 2/3 for specifications; as per my point above, determining if something weighed 13.3127 grains on your scale would take more work.

 

Woah, that means than more than 20% of the specified weight (2.5 g for Barber dimes) was lost due to wearing down!

So finally going back to my original question: given this, my own experience weighing coins in my collection, and the relatively lower tolerances in the primary sources I've managed to find, I'll keep in mind that little old copper coins could be significantly underweight due primarily to wearing down through circulation.

 

It's easier to work in fractions when you have a balance scale.

Put the pans on with the unequal arms, say 1:4 and balance with shims.

Put a grain weight on one side. Put small bits of say lead on the other side until it's level.

Melt the lead bits and stamp it 1/4 (grain)

 

I think it's proportional to the surface area of the coin, and the smaller the coin, the larger its surface area relative to its weight (the "square-cube law"). So a slick dime might be 20% underweight, but a half-dollar worn to the same degree might be only 5-10% underweight.

Remember, though, as @GDJMSP said above, the weight doesn't really start to drop much until you're down into very low grades (Good or below). As a rule of thumb, if the coin's rims are still distinct, it probably hasn't lost much weight.

 

On copper coins no they didn't because the metal value, while somewhat close to the face value, had enough wiggle room in it that it provided the profit upon which the mint operated. You half cent might have between a quarter and 3/8ths wourth of a cents worth of copper in it and your weight would have to be significantly off to lose money on them. Plus rolling a sheet of metal to a specified thickness wasn't too difficult and since they knew the diameter of the blanks they would be punching out it was fairly simple matter to control the sheet thickness to get blanks within the correct weight range.

Now silver and gold were another matter. The value of the metal was very close to the face value and the higher value of the metal meant that tolerances had to be much closer. Enough so that you couldn't just depend on the thickness of the sheet of metal. So in the early mint and until the first could decades of the 20th century every blank WAS individually weighed. In the early mint they had an entire team called adjusters that would take each blank and weigh it. If it was too light it was condemned back to the melting pot. If it was within tolerance they passed it on to the next department. But if it was overweight they would gently use a file to reduced its weight until it was within tolerance and then pass it on to the next step in the coining process. In the early years they would file the face of the blanks but that often left marks that were still visible after the coins were struck (known as adjustment marks). Later they stopped filing the faces and used the files on the edge of the blanks where subsequent upsetting and striking would wipe out the marks. In the latter part of the 19th century automatic weighting machines were used that could weigh the blanks and sort them much faster than the adjusters could. This was an automatic weighing machine from the US Mint.



The picture above the machine shows a group of such machines in use. After the machine sorted them they only had to worry about adjusting the heavy ones. Eventually they stopped adjusting at all and simply condemned the heavies back to the melting pot as well. As the price of silver kept falling in the early 20th century they finally stopped weighing the silver coins as not being worthwhile

Gold coins were handled about the same way as the silver in the 19th century but by the 20th the procedure changed somewhat, The coins were still weighed and sorted, but the lights and heavies were not sent back to the melting pot. Instead they would put the proper face value in each bag, but they would selectively combine lights, heavies, and normal coins so that the total weight of the coins in the BAG were within tolerance.

The original chart in the almanac was compiled by Tom DeLory. Much of the information can be found in the legislation as mentioned earlier, for those not mentioned, Tom got the information from the US Mint.

Three pieces from each deposit of gold or silver, not just three pieces from each year.

They were more accustomed to the use of fractions or X parts in so many up until the government adopted the metric system in 1866. After that amounts are specified in grams and decimal fractions of a gram.

It is surprising how precise they were able to make balances back them. They had balances that could weigh several hundred pounds of metal to within a quarter grain.