So I'm starting out at the furthest away person I've got, seeing if I can get further. This is Donald Stuart who was born somewhere around the late 18th Century or very early in the 19th Century - I have little info on him, the only absolute solid data I have is his name on his son William's marriage & death registration entries. His wife, William's mother, has even less - her first name isn't given on William's death registration, but I can probably still be sure she was called May Stuart (maiden surname is also Stuart). From those two register entries I can also say May Stuart died before 1859 & Donald Stuart died between 1859 & 1905. What would've been really helpful is I could have found William and Donald living in the same household in the 1851 census (and even better if May is alive then too), as that would give me other potential siblings and definite (well mostly so) relationships between the family members. Unfortunately, I'm not that lucky :/ I'm fairly sure I'd got William tracked down in 1851 as a farm servant in someone else's household (mostly based on age plus birthplace), which means I never have him and his father or mother in the same household (with the relationship stated in the census).
I had at one point found a household in 1841, that definitely had a Donald and a William of about the right age (born 1831ish so 10ish) and no obvious mother. There's no actual family relationship information on the 1841 census, tho - and it later turned out that I'd misled myself and this family (who I tracked the other members of through to 1901) weren't my family. I found this totally by chance when I tried to find the death registration for Donald, and there were two possibilities in the right time period. And on looking at both of them one entry was clearly the Donald I'd been tracking but that had the wrong wife's name! The other one had the right name as first wife, followed by a second wife. So - I looked for this potential family, and found Donald plus second wife Christina Cruickshank in 1851, 1861 and 1871 with a bunch of kids. One of the children's names is the name of the informant on that death registration entry, too. So I'm fairly sure I have the man and family from the death certificate.
But are they my family? That's where it gets more handwavy. Donald Stuart/Stewart seems a common name for that time and place, as does May/Mary/Marjory Stuart/Stewart - looking for marriages between those names in the right sort of place and the right sort of time, I found 3 possibles without trying very hard. And as this is all pre civil registration of these events (which started in 1855 in Scotland) there isn't actually all that much information recorded. And I can't be sure I've actually got the right area for the marriage. William is pretty sure throughout all the censuses I've found him in (1861-1901 for definite, 1851 probably) that he was born in Kirkmichael, Banffshire in ~1831. (As an aside, if you've never looked into this sort of thing I think you'd be surprised how many people's dates of birth & places of birth change from census to census!) So I was looking for Donald and May in Kirkmichael, and surrounding counties. However, if this Donald I've found is the right one then he was born in Strathdon, Aberdeen. So should I be looking there? That might fling up even more possibles. Or they could have got married in whatever parish May lived in/was born in/grew up in. Which I know nothing about. I tried looking for May's death (it's got to be post 1831 and pre 1859), but found nothing plausible enough. I also tried looking for William's baptism, in the hope that the entry might say something like "Donald & May Stuart of the parish of X". But no joy there, nothing plausible in the right time period for Kirkmichael. Of course he might not've been baptised, or might have been baptised somewhere else (his parents' home parish?) or even born somewhere else and just moved to Kirkmichael at a young enough age that he thought he'd been born there.
So I think what it boils down to is that whilst I might have the right Donald Stuart's death registration I don't have enough connecting data to be sure, and the names are common enough that this may be a different Donald Stuart (who married a May Stuart), not my ancestor. So Donald is still a dead end, as is his wife May.
Next step (which I've actually started, but this entry's long enough as it is) is looking at William Stewart (b.1831)'s wife's family - she is Helen Candy (b.1833) and I have got at least one generation further back on her side, in that I know her maternal grandparents' names (so that's one set of my great-great-great-great-great-grandparen