Given the mintage quoted in Coin Yearbook is over 3 million, clearly at least two or three dozen dies were used and given the short life of dies in this year, possibly more. That leaves lots of scope for different die identifiers to be noted. The one I have in my trays is a high 47 for example - easily seen due to a flaw through the hair to the cheekbone level with the eye. A quick visual check online showed the three varieties listed by Colin, plus the same dies as mine in LCA 151. Baldwins 70 had one that appears to be a wide 7 and another example had a weak diagonal to the 4.
I think a repaired existing die is the more likely cause if a 1 was used. If so, a worn/broken punch would be ideal for partially punching in a digit.