The more time I spend around Saint-Gaudens double eagles, the more I find myself drawn to the coins that people seem to talk about the least. Everyone knows the 1907 High Relief. Everyone knows the common dates. Even people who are not really Saint collectors understand the appeal of the design.

The Proofs are different. They are strange coins. They do not really look like the Proof gold that came before them, collectors did not particularly like them when they were new, and today the surviving population is so small that many collectors may go years without seeing a really great one in hand. I have always found that fascinating.

When most collectors hear the word Proof, they picture deeply mirrored fields, sharp contrast and a coin that immediately announces what it is. A Proof Saint-Gaudens double eagle does not necessarily do that. In fact, if you handed an average collector a beautiful 1908 Matte Proof Saint without telling him what it was, I am not sure his first reaction would be that he was looking at a Proof coin. That is part of what makes them so interesting.

The regular Proof Saint-Gaudens series really begins in 1908. Only 101 were struck that year. Instead of the brilliant Proof finish collectors had become accustomed to during the Liberty Head era, the Mint produced a coin with a matte, sandblasted surface. The entire coin has a fine texture to it. Under the right lighting it can be beautiful, almost glowing. Under bad lighting it can look flat and dark.

I can understand why collectors in 1908 may not have known what to make of them. You have to remember what came before. A collector ordering a Proof double eagle was used to receiving something bright and reflective. Then the Mint sends him this large gold coin with a dull, granular surface and tells him this is the new Proof. Today we appreciate it. Back then, apparently not so much.

The Mint continued experimenting. In 1909 they changed the finish again. Only 67 Proof double eagles were struck and the surfaces became smoother, with what we now generally call the Roman Gold or satin finish. The 1910 Proof followed the same general direction and 167 were struck, which is actually the highest mintage of the regular Proof Saint series.

Then in 1911 the Mint went back to the sandblast finish. One hundred were struck. After that the numbers begin getting almost ridiculous. Seventy-four in 1912. Ninety-nine were reportedly struck in 1913, but only 58 were actually distributed. Seventy in 1914. Fifty in 1915. Then it was over.

What is interesting to me is that these mintages were not necessarily low because the Mint was trying to create great rarities. Collectors simply were not buying them. That changes the way I look at the series.

Today we tend to see a mintage of 50 or 60 coins and immediately think of rarity, demand and prestige. At the time, the Mint was basically having trouble selling them. There are old accounts of Proof gold having so little premium that dealers would simply turn coins into banks. There are records of Proof double eagles selling for barely more than face value.

Imagine that. A Proof Saint-Gaudens double eagle selling for twenty dollars and change because nobody particularly wanted it.

It sounds impossible today, but markets change. Collectors change. What is popular in one generation can become almost forgotten in another, and sometimes the coins that nobody cared about turn out to be the ones we wish had been saved.

That is exactly what happened here.

The survival rates are very small. And this is where the population reports start getting interesting. If we only look at the regular Proof series from 1908 through 1915 and leave the controversial 1907 High Relief Proof and Specimen coins out of the discussion, PCGS currently reports 242 graded examples and NGC reports 368. That gives us 610 grading events between the two services.

At first glance that number sounds high. Then you compare it with the original production. Depending on whether you use coins struck or actually distributed for certain years, the entire series only amounted to a few hundred coins to begin with. And we already know that many were returned, melted, spent or otherwise lost.

So there clearly are not 610 different Proof Saints sitting in holders today.

Anyone who deals in certified coins understands what is happening. Population reports are not survival reports. The same coin can be graded more than once. Coins get cracked out. They get upgraded. They cross from one service to another. Sometimes the old certification gets removed from the census and sometimes it does not. Over enough years, the numbers grow.

That is why I think Proof Saints are a perfect example of why we have to be careful when we talk about population reports as if every number represents a different coin. It does not.

Then you look at CAC. CAC reports only 58 stickered Proof Saint-Gaudens double eagles. If you use the combined PCGS and NGC population reports of 610, that works out to roughly 9.5 percent.

In simple terms, only about one out of every ten grading events in the regular Proof Saint series has received CAC approval. Even that percentage probably does not tell the whole story because of all the duplication buried inside the grading populations.

But whichever way you look at it, 58 coins is not a lot. Eight years of Proof production. Eight different dates. Different finishes. Different surface textures. Different levels of rarity. Only 58 CAC approved examples. That is a very small pool of coins.

I also think the 1907 issue makes the story even more interesting, although I would personally keep it separate from the regular Proof series. NGC recognizes certain 1907 High Relief double eagles as Proofs based on the method of manufacture, die marriage, collar, strike and surface characteristics. PCGS takes a more conservative approach and calls comparable pieces Specimens.

Who is right? Maybe both, depending on how you define a Proof.

There is clearly evidence that certain 1907 High Relief coins were made differently and with far more care than ordinary production pieces. What we do not have is a Mint ledger that clearly says these are Proof High Relief double eagles and this is how many we made. That uncertainty is probably exactly why the two grading services took different approaches.

For me, the cleanest Proof Saint series is still 1908 through 1915. Those are the coins that were actually produced as part of the Mint's regular Proof program and offered to collectors.

And when you really study them, the series becomes much more interesting than just a group of low mintage dates. The surfaces themselves tell the story. The Mint was trying to figure out how to make a Proof version of Saint-Gaudens' design.

That was not an easy problem.

The design was sculptural. The relief was different. The broad fields and flowing devices did not necessarily lend themselves to the brilliant mirrored treatment collectors expected.

So the Mint experimented. Matte. Satin. Roman Gold. Sandblast. Different textures from year to year. You can actually see the evolution of the Mint's thinking through the coins.

And once you have looked at enough of them in hand, you start realizing how important originality is. A normal Mint State Saint can take a small mark or light friction and sometimes hide it in the luster. A Matte Proof cannot. Break that original surface texture and the disturbance can jump right out at you. A tiny shiny area surrounded by untouched matte surface can become the first thing your eye sees.

That is why truly great Proof Saints are so impressive. The surface has to survive. The color has to survive. The texture has to survive. And it all had to survive for more than a hundred years on a twenty dollar gold coin at a time when twenty dollars actually meant something.

When I look at a high grade Proof Saint today, that is what I think about. Not just the number on the holder.

How did this coin make it? How did it avoid being spent when nobody cared about Proof Saints? How did it avoid getting turned into a bank? How did it avoid being cleaned when people did not understand the surfaces? How did it survive the great gold melts? How did it somehow stay intact long enough for us to be looking at it today?

That, to me, is where the appeal is.

The irony of the Proof Saint-Gaudens series is that collectors originally rejected the very thing we admire about the coins today. They wanted brilliant Proofs. The Mint gave them something different. They did not like it.

More than a century later, that strange experimental finish, the thing that probably hurt demand when the coins were new, is exactly what gives them their character.

Sometimes the market just needs a hundred years to change its mind.