A Morgan comes across the counter on a Tuesday afternoon. 1893-S. The date every collector wants and almost nobody has. The gentleman holding it inherited it from an uncle, and the uncle inherited it from somebody before that, and the story is warm and the coin is heavy and the light catches Liberty's cheek the way it should. I set it on the felt. I get the loupe on it. The devices are sharp. The reeding is even. The weight, on the scale, reads 26.7 grams, which is exactly what it should read.

Then I tap it against the wooden handle of a pencil, and it clicks.

A real silver dollar rings. It sings for a full second, sometimes two, a clean high note that hangs in the air after the coin has gone still. This one clicked, the way a stack of nickels clicks, dead and short. The gentleman watched my face. I told him what I was hearing before I told him what it meant. That is the part of this job that has gotten harder in the last ten years, and it is the reason I am writing this down.

The fakes coming out of overseas workshops now are not the fakes of thirty years ago. Thirty years ago a counterfeit Morgan was obvious. Wrong color, wrong weight, mushy details, a seam you could find with your thumbnail. You did not need a loupe, you needed a glance. The workshops have caught up. They are casting in the right alloy, or close to it. They are hitting the weight. They are copying the dies from real coins, so the devices come up crisp, and the reeding comes up even, and the diameter comes up to the tenth of a millimeter. The three tests that used to catch everything now catch about half.

The Morgan dollar is the favorite target, because the premium coins in the series are worth real money and the common coins are worth enough to move quietly. 1893-S, 1889-CC, 1895 proof. But the trade dollars are getting hit harder, because collectors know them less well and the chop-marked examples give a counterfeiter cover for surface irregularities. And the early world crowns, the Spanish colonial eight reales especially, are showing up in quantities that do not match how many were made. Somebody carried this a long way, the seller will say, and the coin will look like it has been carried, because the workshop tumbled it in a barrel with grit for a week to earn that wear.

So. Three tests, in the order I run them.

Weight first, because it is the cheapest to check and it still catches the lazy fakes. A Morgan is 26.73 grams. A trade dollar is 27.22. A Spanish eight reales runs 27 grams give or take a hair for wear. If the scale reads low by more than a few tenths, you are done, put it in the flip and move on. If it reads right, you have not proven anything yet, you have only ruled out the worst of them.

Ring second. Balance the coin on the tip of your finger, and tap it with something wooden, a pencil, the end of a chopstick. Silver rings. It rings long and high and clean. A cast fake, even a good one, clicks or thuds, because the internal structure of a poured coin is not the internal structure of a struck one. This is the test the workshops cannot fake, because they cannot fake the physics of a struck planchet without actually striking it, and if they are striking it they are spending money they usually will not spend.

Detail last, and this is where the loupe earns its keep. Look at the rim before the field. Look at the smallest letters, the mint mark, the designer's initials. A cast copy softens at the edges of the smallest features, because that is where the mold gives up its detail. The strike tells the story. A real Morgan has a fine ridge where metal was pushed into the die, a fake has a soft slope where metal was poured against a mold. Once you have seen the difference on ten coins, you cannot unsee it.

The gentleman on Tuesday took his coin back and thanked me for being straight with him. The story is still true. His uncle held onto something for a long time. It just was not what everyone thought it was. That happens. The coin goes back in the drawer, the story stays in the family, and the next one that comes across the counter gets the same three tests in the same order. Weight. Ring. Detail. The ring is the tell.