Genealogy Chat

and to clarify a bit ... Ancestry had a deal with FreeBMD that allowed it to use FreeBMD's transcription of the index directly (some years ago it could be consulted at Ancestry free of charge). Ancestry's site now says

http://search.ancestry.com/oldsearch/rectype/vital/freebmd/bmd.aspx

Births 1837 – 1915 (transcribed by FreeBMD)
Marriages 1837 – 1915 (transcribed by FreeBMD)
Deaths 1837 – 1915 (transcribed by FreeBMD)

so if you have a correction to make, best to submit it to FreeBMD and Ancestry will catch up when FreeBMD submits batches of corrections to Ancestry.

The "St" problem could have originated with Freebmd or slipped in somehow when Ancestry imported the information. If you cross-check the same births at FreeBMD you will be able to see what it says there (where it may have been corrected).

One I have seen that does originate with Freebmd is a whole quarter of births in the early 1900s in one registry district that are shown at Freebmd and thus Ancestry as taking place in a completely different year from when they really occurred, I forget the details but something like showing as occurring in 1915 when they really occurred in 1904.

Just to clear up one point from the message Inky1 received from Ancestry - Freebmd did not and does not use OCR for transcribing any index. All are transcribed from images by volunteers and the aim is to have every entry double transcribed (ie transcribed by two different people). Where the second transcription differs from the first, both are shown ; if the entry shows in bold type, this means that it has been double transcribed and that both transcribers have entered the same info.

" the OCR technology tries to guess words intelligently" - freebmd transcribers are instructed never to guess (even intelligently) but only to type what they see.

In addition to IGP's comments, the following cut&pastes are from one of my many messages to Ancestry:-

SENT 10Jul2011
Hi,
Am currently searching early 1900 births in the Southwark (Surrey/London) area. It seems that many (or even all?) transcriptions for St. Olave registration district have been entered merely as "St".
Any reason why?

REPLY 17Jul2011
Because these Historical Records have far too much text to be indexed manually, indexes are built using advanced Optical Character Recognition (OCR) technology to automatically recognize text within an image. We believe we are using the very best OCR technology available, but it is still less accurate than human eyes and brains. These limitations make it impossible to achieve the same level of accuracy found in other indexes. In cases where the original microfilm text is unclear, the OCR technology tries to guess words intelligently. It seems to have failed in this case.
Thank you for bringing this to our attention. I have reported this error in the England & Wales, FreeBMD Birth Index, 1837-1915 to our content team to look into. Feedback from you, our valued customer, helps us correct errors and improve the website. Your patience and efforts to assist us in this matter are appreciated.



If Freebmd is linked into Ancestry (?), then I presume that GR/FMP has also used OCR software to generate its BMD lists. Anyone know?

Thanks JoonieCloonie. I'll keep you posted as soon as find out anything further.

Partial refunds were stopped - they were technically illegal, apparently. You now get a full refund if the cert can not be found from the information you supply.

The trick in ordering is only ever to supply information that you are 100% certain is accurate. Many members have fallen into the trap of providing 'extra' details mistakenly thinking they are being helpful. Far from it.

As far as the GRO making mistakes? well yes, of course errors will occur. My own marriage had both our surnames indexed incorrectly, not actually the fault of the GRO - or more precisely their employees - but down to a humans interpretation of the poor handwriting of the minister who filled out the details in the church register.

It is also important not to confuse the GRO Index with the GRO Register. The two are not the same. Again, using my own wedding example, the Register was correct, the Index was not. It is further complicated when third party bodies such as freebmb, etc use people to re-transcribe the GRO index thereby risking the introduction of even more errors.

The GRO Index will never tell you who married who, that can only be found in the Register, and the only way to see that information is to order the cert.

In another example we have, the case of my great grandfather, his surname is recorded completely incorrectly. We originally thought there might be an error at the register office and / or the GRO and / or with a website transcription, but no.

He married my great grandmother when she was 9 months pregnant with my grandfather but in the next parish to where they lived and on checking the actual parish register it now appears he used a false name.....!

You have a plethora. :-)

Full info is always most useful -- the one who married James was Elizabeth Leilia Barrett, per the marriage record.

Searching for births for the name Elizabeth Le* Barrett (because I doubted that that one avoided variant spellings) we find:

Births Mar 1870
BARRETT Elizabeth Lebiley Newport, M. 11a 188
Barrett Elizabeth Lehiley Newport, M. 11a 188

The first one seems to have been identified as correct but I would certainly read that as an 'h', making the name Lehiley and much more similar to Leilia.

The 1901 and 1911 censuses call her Elizabeth L Lawrence born abt 1870 in Newport, Monmouthshire. And sure enough, the 1891 census shows the father born in Cornwall, but the rest of the family born in Wales, the children in Newport.

Now ... what was the question? :-D

The origin of the other Elizabeth, I guess.

Well, no middle name, but in 1911 (she shows as Elizibeth at Ancestry) her place of birth is County of Galway, in case that helps!

This thread is making for interesting reading not the least because it has given me some pointers in response to my original post which staretd it off, so thanks to all.
But this gets curiouser and curiousier and may be of general interest in realising how coincidencs can lead you astray.
If you remember I asked if the GRO can get it wrong and was put in my place (rightly as it turned out) for doubting the records office that John Henry Lawrence didn't in fact marry the Irish Elizabeth Burrup, even though he married an Elizabeth from Ireland according to later censuses. The four names on 11a, 388 of the records show John Henry Lawrence, Elizabeth Barrett, Elizabeth Burrup and George Hine, indicating that it must be the Barrett variety of Elizabeth that he married.
BUT NO.
Having delved a little deeper I found the following:
In 1891 John Henry Lawrence (b1866) was living at 49 Robert Street in Newport, with his wife Rosine and their two children. Rosine sadly died aged 22 the same year. Also in 1891, next door at number 48 there was a family of Barretts (John - b1863 and Martha - b1861), from Ireland and their two children. Nobody in the house is called Elizabeth. There is, however, an Elizabeth Barret b1870, living at number 10 Robert Street with her parents William and Elizabeth - but they are from Cornwall, not Ireland (still with me?). And guess what, this Elizabeth marries James W Lawrence, who is John Henry's brother, in December 1892 (11a Page 371). By 1901 they are a married couple living at 46 Williams Street Newport, next door to number 44 Williams Street, which is occupied by - you guessed it - John Henry Lawrence and his new wife (no I didn't believe it either) who is also called Elizabeth b1870, from Ireland, whom he married in June 1892, plus his original two children and by now five more. As far as can see there are definitely two 1892 marriages to two different Elizabeth Barretts, one to John Henry (June 1892) and the other to his brother James William (December 1892). I'm intrigued but I suppose I will have to wait until the marriage certificate for John Henry pops through my door in a couple of weeks, unless any of you good (and patient if you've waded through this) people can shed any light. I'm not making it up, honest. Anyone dare?
Martyn

DC
I don't think that the partial refund arrangement has stopped. But it applies where you do not get a certificate.

Like SC, I have a number of marriage certificates (also births & deaths) that are not my family. But they were paid for on the basis that they might be. And, if not, then at least I then know to look elsewhere.

Surely you get a partial refund? or has this stopped?

that's why I always request a marriage certificate in the name of ONE person


the person I am interested in.


It may be a waste of £9.25 if it is not "my" ancestor ......... but even that rules out one more person in the search!

You are quite right -- when the original handwritten index was typed up, errors did slip in. I have encountered this when looking at the typewritten version of what would obviously have originally been a handwritten list.

No different from when Freebmd transcribes the parts of the index that are still in handwriting. :-)

A further potential source of error.

Many moons ago I made fairly regular trips to what was the London Family Records Centre. Both to view census microfilms and to stand at the desks and thumb through the mighty (heavy) tomes of BM&D's. In my early days there, many of the quarterly books were handwritten. But over time, there were often gaps in the alphabetical sequence of books because they were being revised into a typed format. I don’t know how or where that activity was carried out. But there may well have been further transcription errors twixt the original handwritten and the typed version of the books.

My understanding.

The original marriage record generally stays with the Church or Register Office. And a copy is subsequently sent to the GRO. That is why, when you order a certificate, the whole record that you get seems to be completed by one hand. (ie you do not get an image that includes the actual signatures of the couple or their witnesses.)

This leads to yet another source of error within the GRO system - the occasional ommission of some records from the quarterly lists. Probably happens to Births & Deaths as well.

At least one other site has some original Church record books online. And you see the actual signatures/marks. I don't know if any Register Office books are online?

the son's birth certificate will definitely tell you Elizabeth's surname, unless she fibbed! and that is the one and only way to find the names of a child's parents, again, unless they concealed something (or a father was not named).

The couple may not have married. But have you tried just searching for William Speller marriages to Elizabeths around the right time and place? If you have cent for a birth cert though there is no point in guessing, just wait and see.

When you get your certificate, do go back to your original thread

http://www.genesreunited.co.uk/boards/board/ancestors/thread/1334154

(google found that for me) and add any questions you have there!

Thank you,
I have looked at all those names and just couldn't make out why so many same numbers, thank you for explanation, I thought it was too good to be true, all the cenuses show that his wife was Elizabeth, and I have a copy of her death cert. but of course no maiden name, I have sent off for one of their son's birth cert. that may show something, I waste so much time and money on the Speller family, always thought it was an unusual name, how wrong can you be. Margaret x :-S

This is the same answer.

Did you look at Freebmd?

Marriages Dec 1848
BOATMAN John Dunmow 12 173
PATMORE Charles Dunmow 12 173
Perry Charlotte Dunmow 12 173
SEARLE Elizabeth Dunmow 12 173
Skingle William Dunmow 12 173
Speller Martha Dunmow 12 173
Speller William Dunmow 12 173
Waller Naomi Dunmow 12 173

If the GRO says John Boatman married Elizabeth Searle, then you can be very sure that John Boatman married Elizabeth Searle because they are looking at the *marriage certificate*.

These names come out of the original register page.
Usually there are four names on a page - two marriages consisting of one bride and one groom each.
Sometimes there are only two, sometimes there are as many as 10.
On that page, there are 8 (barring mistranscriptions).

But what we are all looking at is the GRO index, where the names from those pages were transferred into an alphabetical index for each quarter of the year, all districts combined.
In 1911, that index started to show the spouse's surname. Before that, it is just a long list of names in alphabetical order, pages and pages and pages, everybody who married in England and Wales in that quarter.

Without the magic of search engines, you would have had to find 'John Smith' who married in Hackney in June quarter 1895, and was recorded on page 1 of volume 1. Then you would have had to read through the *entire index* to find the other names that were recorded on page 1 of volume 1.

What Freebmd (and now the commercial sites) has done is painstakingly transcribe all those names in on all those pages and enter them into a computerized, searchable database.
You can search it to find all the names on a particular page.
Or you can search it to find a name that appears on the same page as another name.

What that *will not* tell you, before 1911, unless you have the luck of finding a two-person page, is who married whom.

So all you know is that william Speller and Elizabeth Searle were recorded on the same page of the marriage register -- along with six other people.

And they did not marry each other -- because the GRO people have looked at the actual marriage certificate for Elizabeth Searle and she married John Boatman.

From a search, it seems that no William Speller ever married an Elizabeth Searle. Did you have another reason to think there was a couple by that name? It is possible there is and one of the parties has been mistranscribed. Mistranscriptions are legion, so if you have an idea of the time and place, it is always worth checking into.

I am also having trouble finding a marriage certificate from GRO \i was asking for Willian Speller married to Elizabeth Searle, they sent me a certificate for a John Boatman married to Elizabeth Searle, The page No: Vol No: and quarter all matched with the each other, but the line No: was different after checking on several sites I asked them to check it, well the answer was in a nut shell you are wrong and we are right !!!! Where do I go to from here. :-S Margaret

You are very welcome and now you know where to look next time, for free :-)

Getting the birth certificate of a known child as Elizabeth says is always the best (and really the only) way to know you have the right couple.

ElizabethK - thanks for that suggestion; I guess it would - I'll research the Barrett bit and then have a look.
Martyn

Just seen your post JoonieCloonie and this looks very interesting as my cousin (my mother's sister's daughter) tells me that her mum used to tell her John Henry Lawrence married an Elizabeth, but she thought she was a Barrett (Barrett-Burrup, easily confused), but as I couldn't find a Barrett I assumed that must be wrong, but it seems you may have hit upon something, which I have missed. So thanks so much while I go and check it all out. I'll be back as they say and in the meantime thanks to all who have offered their advice so far