Hi Tim,
I will check the index and if the mistake hasn't already been corrected in the current version I will get Steve to alter it. As you would expect with such a massive work as the Roud Index produced by one man there are bound to be many errors such as this. Steve hasn't had time to inspect every version closely which is why researchers like ourselves are looking for them and passing them on to Steve. He is very busy adding to both this and the broadside index and can't break off every time a correction needs altering, but they all eventually get done usually within a month.
The error here is easily understood as 1478 has lines in common with some versions of 218. 218 is quite a common song whereas 1478 has only a handful of known versions and no known broadside as yet has surfaced.
Thanks for posting the 2 versions. I hadn't got both of them. The first is reprinted in the Journals and Reeves' Everlasting Circle of course.