A couple of queries have arisen regard my update analysis, so I'll try to add a bit of background here. After each update, I compare the new "breakdown by year" with the previous one, and report the major increases. The breakdown figures add up to the total "unique records". Records that have been transcribed more than once will show up as one record if they meet FreeBMD's matching criteria for a "unique record", otherwise they will be double (or more) counted. At the start of 2003, there were several years that had extremely large increases in the January update (which nobody commented on) followed by large decreases in the February update (which generated a fair bit of discussion). I think this was caused by some double keyed records not being "matched" in the January figures, and so being double counted. For example, I announced 860,346 "new" 1902 marriages in January, and a reduction of 307,209 in 1902 marriages in February. The actual figure of 553,137 "unique records" in the February figures corresponds to more than 100% of the expected total. (There are 552,408 unique records in the latest update, so some corrections appear to have been made.) In due course, every year will exceed 100%, but only to the extent that interpretations differ. Happy searching/transcribing Peter Dauncey