This is taken from an e-mail by Russell Ridout to the SoG mailing list - with his permission. He refers to books by Michael Foster - well worth reading. ******** he (Michael Foster) does report that there are instances of either or both parties to a marriage not being indexed. In addition, there are cases of father's names - and witnesses - being indexed instead of the bride or groom and at least one where the bride's name was repeated for the groom. Add to that wrong page numbers/district/misread names/mistyped or miswritten data. Also factor in register copies which never reached the GRO for any number of reasons; transcribers losing concentration and missing a few entries - or not realising that the copy entries that they received were on double sided paper and it makes you extremely grateful to have found any of the events that we need! Given that until very recently all the indexing has been done manually you can understand why the GRO staff had regular breaks, as evidenced by changes in handwriting every half dozen pages or so. Michael mentions that he carried out an experiment with a quill pen and thinks that each clerk probably wrote for three or four hours at a time. Error checking - if it occurred at all - seems to have been rudimentary at best, but having worked in an industry where records were originally kept manually I understand how easy it is to introduce mistakes no matter how careful you are! Whether the modern computerised records are any more reliable remains to be seen. Ideally the records should be completely reindexed from the originals, but there isn't the spare cash for this to be funded by the state. Some records have been reindexed locally but these are in a minority. ******** So - please do not be concerned if you find multiple sources that are inconsistent. Just bear in mind that such records are manually produced and can be in error. Regards Chris