According to the same tree as above, John Henry married a Betsy Marrow in Romney Marsh in Sep qtr 1873 as Henry Santer. Only one bride showing on FreeBMD though, Rosena Ovenden.
The same tree has them having a son James Thomas Santer born 1872, but he doesn't show up on FreeBMD or GRO, but he does show up as being baptised at Kilndown, Kent on 21 Jul 1872, s/o John Henry Santer and Betsey Marrow.
Ian C
Thank you! That's a really helpful tip - I've been doing some digging into this, and think I've uncovered at least part of the mystery. First of all, I think that tree lists the marriages erroneously - they relate to a different Henry Santer in a different part of Kent.
While I couldn't find any birth record for a James Thomas Santer, I could find a birth record for a James Thomas Barrow born on 1872 in Cranbrook (in the census, his birthplace appears as Goudhurst or Kilndown, which matches the baptism record). So I think 'Marrow' is a transcription error and the child's mother was actually Betsy Barrow, and James was her illegitimate child (on the GRO, he appears with the dash indicating no father listed).
I know that occasionally the natural father's name would be written on the baptism record or similar, so maybe that's what happened here and that's why there's a baptism record for a James Thomas Santer? It still doesn't help answer the mystery of what happened to John but it helps a lot.
On his own marriage record, James records his father's name as 'John Barrow' and indicates he's deceased at the time (1893), and of unknown occupation. So it looks like James was (very understandably) hedging the truth of his own background, which he may not have been fully aware of himself (his mother died a few years after he was born).
ETA: From checking the census, John Santer and Betsy Barrow were literally living on the same street when the 1871 census was taken, which strongly supports this theory.