Family historians, who have used actual parish registers rather than online indexes, may be aware that in some cases registers may be missing or damaged, preventing records being found. What researchers may not be aware of is how unreliable parish records can be.
During my time as a genealogist and a transcriber for the FreeReg website (www.freereg.org.uk), I have noticed many incorrect entries, many of which are purely down to the person responsible for recording the events.
A reoccurring mix up in a Sussex baptism register relates to the wrong name being given to the father and/or his son. On the 29th March 1829, Richard the son of William and Maria Howard was baptised in the parish of Boxgrove, however a note at bottom of page states “in no. 455 there has been an evident mistake. It should have been William son of Richard and Maria Howard.” Later in the same year a similar mistake was identified when Charles Welburn, the vicar, noted that the baptism of James the son of James and Ann Saunders should have read as “George son of James and Ann Saunders.” Both entries the officiating minister was the local curate. The entries themselves had not been amended so it is feasible that other indexes still give the incorrect details. The FreeReg entries do identify this error and list the entries under both the names given.
In a different parish two different spellings of the child’s name were crossed out with the forename of Bircham being written, but again a note at the bottom of the page states that the child’s correct name was Bertram. This could have been caused by the clerk not being able to read the previously written notes, before the entry was added to the register. Despite the fact that events where supposed to be written into the parish registers at the time, it is clear that some parishes had a different system. One case came to light whereby a parish clerk duplicated 10 1817 baptism entries before realising his mistake and crossing them all out with a note stating that “this mistake was made by taking the day book from 1817 thinking it was 1818.”
Laziness also explains why some baptism entries may be missing from the registers. In the case of another baptism a child’s baptism in 1818 was not recorded until 1825 as the entry had been “omitted to be entered at the correct time.” As the child’s younger sibling was baptised at the time this entry appears, it seems that this brought the error to light. Other researchers/transcribers have told me about complaints written in parish registers from a new incumbent about the poor state or missing entries caused by his predecessor.
In some cases entries could be missing for other reasons. In a Staffordshire parish a page from a marriage register was torn out whilst the book lay on the alter table during a service. Perhaps this was done by someone who no longer wanted to be married. This damage whoever, would have affected other marriage entries written on the missing page. Fortunately, in this case, these marriages had already been copied as part of the Bishops Transcripts, but they are still missing from the original register.
Other errors relate to incorrect information being given, for a variety of reasons. A register for St Leonard’s, Shoreditch reportedly states that Thomas Carn, who was buried in January 1583, died at the grand old age of 207 years old, surely an exaggerated age!
In some cases the reasons for errors are easier to understand, as in the case of the baptism of Anne Eliza Boswell in 1870. Her parents are given as Ellen and Henry, but a note at the bottom of the page, dated 2 years later at the time of her sister’s baptism, states that Anne was the illegitimate daughter of Ellen Whittington. Clearly Ellen and Henry were not married to each other.
These examples are just a small sample I have found in just a few months, but the evidence strongly suggest many more, some of which may never be identified, are hidden away in these ancient volumes. That is why I strongly recommend that researchers should always check the original registers/image and the Bishops Transcripts rather than purely relying on transcribed indexes.