Its been a while since I've been on here, so I decided to share some quick pointers to identifying the two reverses of the 1939 nickels.
The reverse of 1938 was used for a short time in early 1939 in Philly, with estimates ranging from 5 to 10 percent of the mintage. Nagengast estimates 10 percent, and I estimate a number closer to 5 percent from circulation finds. This would be a mintage of somewhere between 6 and 12 million coins, all minted close to the start of 1939.
As for Denver coins, Nagengast estimates around 25 percent of the mintage of the 39-d are reverse of 38, and PCGS coinfacts estimates it's the other way around, with around 75 percent being reverse of 38. I haven't seen enough 39-d coins in circulation to estimate a number myself. Being the second smallest mintage of the series, being able to tell the difference between the two hubs on circulated coins is quite important.
The 39-s is estimated to be around half from the Reverse of 38, and the other half from the Reverse of 1940.
A lot has been written about the two hubs, and how to tell them apart. All is well if you are looking at uncirculated or AU coins, that still have step detail, but when you are looking at circulated coins, often witout much detail left, if you just have the steps to tell the two hubs apart, you will not be able to tell the two apart without a lot of guesswork.
I've found the best way to tell is to totally ignore the steps, and concentrate on some small details that do not wear away as quickly as the steps. The letter S in E Pluribus Unum.
The first picture is a coin with the reverse of 1938

Notice on the S in Pluribus, the top of the letter is quite rounded with out a serif.
Second picture is a coin with the reverse of 1940

On the second coin, a Reverse of 1940, notice the top of the S is rounded and HAS a serif. Now you have a second pickup point to sort the two hubs of 1939.