Dear Caroline, >A search of the parish registers showed that no-one of that name, indeed no >STAVERTs at all were born in either Selkirk or Ashkirk before the late >1730s. The parish registers showed no baptisms: they do not record all births. The lack of baptisms may be a strong indicator of lack of births, but it is not definitive proof, and cannot lead to such a definitive statement. >In all the Borders parishes, there was only 1 Andrew STAVERT born in 1730. >He was born in the parish of Lyne near Peebles, the son of John STAVERT, who >was a tailor. Beware that for some Borders parishes the registers do not go back to 1730, so there are no records of baptisms for them (though people would almost certainly have still been being born). The picture is as I said incomplete. And again even those parishes that do have baptism registers at the period do not reliably record all births. You cannot conclude that there was only 1 Andrew Stavert born in the region in that year: you have found only evidence of 1 such child being baptised, but that is not the same thing. It might seem to some people that I'm arguing semantics here. But through my own research (family, demographic and wider) I've seen far too many examples of people whose births are not recorded reliably to ever rely on the parish registers as a reliable record of all births in any area. It's no use thinking that this won't happen to your family research: the chances are that it will, especially as you go further back in time. And it's a sad fact that the further back you go with family history the harder it is to establish links. The uncertainties become greater, the records run out on you, and proof goes out the window. >What do you think? Do I have enough evidence to make a connection? You can have a working hypothesis based on the research that you've done, but you cannot prove things beyond doubt. And unless you take account of the incompleteness of parish registers (even where they do survive), your working hypothesis may be based on dodgy ground. >I would suggest that to a large degree the rarity of the surname will >determine the amount of evidence you need before you can accept a link >between two people. I am afraid that I would have to disagree with this, in a context where the records are so incomplete. I'm doing a one-name study of the surname Cavers which is an incredibly rare surname in Scotland, let alone elsewhere. At least it has a nice Borders origin :) I'm finding vast numbers of Cavers people missing at various times from the registers (as I find with other more common surnames) and the chances of linking people up any reliably are not any greater than with an unusual name. Just because a surname is unusual doesn't mean that the vital reference won't be missing from the registers. If anything it might seem easier to link up a more unusual surname, since there is often not a choice of references to choose from (i.e. you're not battling over many Andrew Stavert baptisms for example). But just because you might have only 1 reference to choose from, even promising in many ways, doesn't mean it's the right one. The reference you really need might not be in the surviving registers, and the chances of this happening are probably greater than you think. By the way I might have some new Stavert references for you. I'll check! (and email directly if so) Viv Dunstan Fife, Scotland http://www.vivdunstan.clara.net/