Recent articles have tended to concentrate on the major family history websites, so it is good to start this time with some useful data collected and published locally.

The website of Keynsham & Saltford Local History Society ( already contains some excellent transcriptions of records, including baptism, marriage and burial records from both Anglican and non-conformist churches. The latest addition is Keynsham Cemetery records. These are taken from the memorial where one exists, or from the cemetery burial register. I have found some burials of Keynsham residents where the death was recorded in the Bristol Registration District, perhaps because they died in hospital there. The series also includes school admissions which show address and father’s occupation. All the records are free to access and easily downloadable. A sample from the burial register is shown below.

 Keynsham burial index

New records on Findmypast ( include The Welsh Collection of Welsh parish records. These have images of the original register. I could not find any details of coverage, but they appear to include records back to the sixteenth century, although not nonconformist ones. In view of the movement between our area and south Wales, especially when the Welsh coal mines were at their peak, it could be well worth checking the records for missing events of people in your family.

The collection of electoral registers 1920-1932 on Findmypast is not new data, but the search methods have been changed. Previously the records had been indexed using optical character recognition. The difference is explained in the site blog - these are existing records, which have been improved, meaning you can now search for an exact name. Before you could search for John Bear and be shown a page with someone called John living in Bear Cottage, now we return exact name matches for John Bear. Not only that, but the names are in the search results too, so it's much quicker and easier to find who you are looking for. The coverage of electoral registers is not complete. The collection does not appear to include Bristol records but does include the City of Bath, such as the one shown below.

Bath electoral register

Using the electoral register for 1931, I found my grandfather’s cousin living in Rowton House in Hammersmith, London. There were a number of Rowton Houses, particularly in London, and they provided superior hostel accommodation for working men. Good facilities were provided, and the men had individual cubicles including primitive en-suite facilities in that each cubicle had its own chamber pot. You can read more on the excellent workhouses site at .

Findmypast has recently added some more Wiltshire records – 1602 to 1688. These are Quarter Session Records which were collected and published in 1932 but were made available in CD and download format by Anguline Research Archives ( They are more accessible in their Findmypast form, and I found some interesting information about the sufferings of the Maggs family in Horningsham during the plague.

Another new dataset from Findmypast is of 881,000 clandestine marriage records covering the years 1667 to 1775. These were marriages performed outside the usual requirements of the law, mostly in London. Although the details are interesting, there is a lot of duplication, and it is not always easy to identify your family from the available information.

Finally, in the recent releases from Findmypast are sets of street and other directories from the University of Leicester Special Collection – but much easier and quicker to use. If you have struggled with the Leicester website in the past, you will find the new interface much better.

Ancestry ( has little new recently, but the Obituary Index is worth looking at. The data certainly includes Bath, but I did not see any Bristol entries. The origin of these notices is shown as BMDSonline, which is a site I have mentioned previously, and which displays death notices from local newspapers. BMDSonline has in fact been renamed and is now called Funeral Notices ( While Ancestry only has an index to the site, Funeral Notices itself shows the full text, which can be helpful in identifying other family members. It is free to use, and you can contribute by lighting candles for those recorded.

Family Search shares many of its records with Ancestry and Findmypast, so it is only exceptionally that there is a real advantage to using it. One possible area is English & Welsh wills from 1858 to 1957. The website at only provides an index, but images of the wills themselves can be accessed at LDS Family History Centres, which are locally at Bristol, Weston, Trowbridge. This could be cheaper than paying £10 for a copy of the will from the probate office.

Ancestry vs Findmypast

A frequent question we hear at the Research Room or at family history events is whether Ancestry or Findmypast is the better online source. It is not easy to compare transcriptions and indexes, since you can find errors in both sites. Both sites cover the principal census and civil registration records, while each has certain things in its favour, depending on your area of interest. For example, researchers for Gloucestershire will find the Gloucestershire Wills and Gloucestershire parish records on Ancestry very useful, while Findmypast has Bigland’s eighteenth-century transcriptions of Gloucestershire memorial inscriptions and lots of Wiltshire records. One big attraction of Ancestry is the PCC wills up to 1858, with the full text available.

Comparisons are easier when looking at the way the two sites handle and present information, although much may depend on the sort of search you are performing, the amount and reliability of the information you already have, and the name you are searching for. We are also always at the mercy of the sites’ algorithms, which decide what should be displayed when nothing matches completely.

There are some clear differences between Findmypast and Ancestry in their census search options.

Findmypast has better search options if you search a particular year via the A-Z of Record Sets. You can do a search by address or by name, and using search terms such as marital status, house name, or using census reference data such as RG13 4811 119 45. Ancestry is less flexible, and uses the same search screen for all censuses, irrespective of the information provided. Findmypast helpfully tells you the number of successful results as you go. For example, there 2,735 people named Brian in the 1911 census, prompting you to find some way to narrow the selection.

I did a search in the 1901 census for the name Ann Read, born 1834, with the usual allowance of 2 years either way in the date of birth, and for variations in spelling of both forename and surname. Findmypast presented 567 results, while Ancestry offered 117. However, Findmypast included people like workhouse inmates whose name was not shown in full, but where the initial letter of the forename was A, and those with a variation of the surname and a second initial “A”. Findmypast shows 20 results per page, while Ancestry lets you choose how many are displayed between 10, 20 or 50.

The order in which the results were presented was also different. Findmypast first showed exact matches, but arranged by date of birth, so the closest matches came half way down the first page. Ancestry had 3 exact matches of forename, surname and date, while Findmypast had 4. However, this mismatch was due to a different reading of the age in the original, and either answer could have been correct. Allowing for some variation is normally helpful. Ann is often spelt with a terminal A and there are variants such as Anna and Hannah. Read can be commonly spelt at least 3 ways. One of my ancestral families came from near Sherborne in Dorset and spelt their surname Read. A son went to Cramlington in Northumberland to be a coal miner, and the spelling changed to Reid and sometimes Reed, which were presumably more common in the local area.

Although Ancestry makes it easy to find the most used datasets, such as the census, I find the Findmypast “A-Z of Record Sets” more user-friendly than Ancestry’s “Card Catalogue”. The latter tends to be congested with many U.S. datasets and the order of presentation is often unhelpful.

There are also differences in the presentation of the results of a census search. I did a search in the 1861 census for John Lawrence, born 1824 in Kent. Ancestry gave six results, one of them a duplicate, while Findmypast offered four.

This is the information shown on the two websites:




Parish and county

Full address



Full transcription


Shown in full for head of household, not shown for others

Shown for all occupants, but county only

Birth year

Shown for head of household only

Shown for all occupants

Marital status


Shown for all occupants


Not shown


And here are the two pages. Note that Ancestry suggests other relevant records, and these are all good suggestions in this instance. John Lawrence married twice in the same village, which is why there are four suggested marriage records. Findmypast also had a suggested record for the 1851 census, but it was not relevant.


John Lawrence 1861 Ancestry

 John Lawrence 1861 FMP

Ancestry offered corrected spellings for both the surname and birthplace, which Findmypast does not do, and the original page is not very easy to read. It does clearly state that he is a “rail road labourer”.

Finally, a reminder that you should not just accept the first record you find as being the whole story. There may be more to discover.

Baby Dorcas was born on 8th March 1840 at Verwood in Dorset. Her parents were George Read and Caroline Vincent, who did not marry until 1845. This is from her birth certificate:

Dorcas Read birth cert

The old GRO indexes to the birth records, which were transcribed and made available on FreeBMD and most other websites, can be searched under just the forename, and I could limit the results by registration district. The birth was listed under both surnames. A surname is compulsory when using the new GRO indexes, and I found her only listed as Dorcas Read, with no mother’s maiden name. I found nothng under Dorcas Vincent. When she was baptised, the vicar described her as illegitimate and the daughter of Caroline Vincent. The word under Dorcas in the baptism has been transcribed as “Dan”, and this is repeated several times on Ancestry, although I wonder if it is actually “dau”, short for daughter.

 Dorcas Vincent baptism

The 1841 census lists her as Dorcas Vincent, but in 1851, after her parents’ marriage, she is listed as Dorcas Read. The family trees on Ancestry give a good account of Dorcas and her later life, but still miss some of the details.