Personally, I'd go with the certificate. I'm sure the detail of the procedure has changed over the years, but now the date on the registrar's certificate would reflect the date on the form that the doctor completes with the cause of death. I guess it could be that someone dying late in the evening is recorded as having died the following day by the time the certification was done ... but I can't see how the registration date could be before the actual date of death. Memory plays tricks on even close relatives: I know someone who several times quoted one date for the death of their spouse, when it actually occurred the previous afternoon ... and had wrongly completed it on the form for the church's book and garden of remembrance until I spotted it in time! I guess if I hadn't, then the church records would have remained wrong. Karen On 23 June 2011 12:14, Chris Howes <descent@wildplaces.co.uk> wrote: > Has anyone any thoughts on a date discrepancy, as to why it might occur or > which should be taken as the more accurate? > > I have a record of a death of Archie Dunkley in Long Hanborough. The death > certificate states 3 August 1918, informant his daughter on the 5th; he > died > of cancer at only 46. The gravestone and probate both state that he died on > 4 August. I'm taking a guess that the gravestone is most likely to be > correct, backed up by the probate record, though of course the death > certificate I guess should be the definitive record but I can see where a > mistake might be made. Informant a 14 year old daughter named Lillian, a > registrar in Witney making a mistake ... > > Would this have required travel by Lillian to Witney? What was required to > register a death in those days; the doctor pronouncing death helping? > Someone came to you, you sent a letter, what? >