Understanding precisely how the GRO indexes are organized can give us key information about individual entries. The marriage index is a whole other kettle of fish, so this answer pertains only to births and deaths. The key pieces of information to consider are:
- The GRO indexes were compiled quarterly (that is, at the end of March, June, September, and December), for events registered in that quarter.
- Each Registration County is divided into a number of Registration Districts. Each of these Registration Districts is subdivided into several Registration Sub-districts. Each sub-district contains several civil parishes. You can find out what sub-district and district a parish is in by using the census and registration reports on Histpop.org. Note that districts and sub-districts did change periodically over the years.
- The GRO organizes its registers by the geography described above: first by Registration County, then by Registration District, and then by Registration Sub-district. Within each sub-district, the entries are organized by date of registration (not by date of birth/death).
- Each sub-district in the register usually starts on a right-hand page (odd numbered).
- Through Jun qtr 1911 there were up to 10 entries per page of the GRO birth and death indexes. From Sep qtr 1911 onward there were 6 entries per page.
The key piece of information here that is not directly apparent is the sub-district. If you know the sub-district, you have significantly narrowed down the number of possible places in a registration district that an event could have occurred. Luckily, with some effort you can work out what the sub-district is from the GRO index page number.
I will use an example to show how one might use this information to their advantage. Say that I'm looking for a Elizabeth Taylor who I know was born in about 1858 in Winthorpe, Nottinghamshire. From the 1861 census population tables I have determined that this parish is in the following jurisdictions:
- Registration County: 31. Nottinghamshire
- Registration District: 442. Newark
- Registration Sub-district: 442(1). North Collingham
- Civil Parish: Winthorpe
Searching for Elizabeth Taylor in the birth index produces at least two very possible entries in September qtr 1858:
- Taylor, Elizabeth, Newark, vol 7b, p 268
- Taylor, Elizabeth, Newark, vol 7b, p 275
Which – if either – of these is the Elizabeth I'm looking for?
Step 1: Extract all the births in this district and quarter, and copy and paste them to a spreadsheet (I am using MS Excel). This can be done easily by searching FreeBMD and entering only the quarter and district name. I like to check for any obvious errors such as duplicate entries, then sort by page number so I end up with something like the following:
Step 2: Look-up the expected page range for this district and quarter. FreeBMD have lists of Page Map-District Index for every district. For Newark Page Map scroll down to September qtr 1858, and check the expected page range. In this case it is from 265-294:
Step 3: Go back to the spreadsheet, and open up a new sheet in the same workbook. In the first column fill in the series of cells with the page numbers from 265 to 294:
Step 4: In the second column, use a formula to make your life much easier. In cell B2, if using Microsoft Excel, enter
=COUNTIF(Data!F:F,A2), and then drag that formula down through all the columns. If using LibreOffice, use the formula
COUNTIF(Data.F:F,"="&A2) instead (thanks @Stephen for this suggestion). Replace "Data" in the formula with whatever the name of the original sheet containing all the data is. In this case, as you can see above, the page numbers are all in column F. What this formula is doing is just counting the number of entries in the "Data" sheet that has a page number in column F that matches the page number being searched for. For example, the formula counts that there are 10 rows in the data that have page number 267.
If you want to, you can do some quick colour coding or conditional formatting to make the numbers stand out:
Step 5: Look for the breaks in the data (the colour coding helps with this). As mentioned before, there are up to 10 entries per page. Each sub-district is usually separated by a blank page (0 count by our formula) or in some cases an incompletely filled page (such as the page with only 3 entries). The data is never perfect – you can see there are few "missing" entries in the data because a few pages are turning up only 9 entries instead of 10.
Using the above results and knowing the number order of the sub-districts from the census report, we can deduce that the page ranges for each of the Newark sub-districts in this quarter are as follows:
Step 6: Where do our entries of interest fit into this? As noted, the Elizabeth Taylors appear on pages 268 and 275. From our page to sub-district mapping, we now know which sub-district each of these birth occurred in:
- Taylor, Elizabeth, Newark, vol 7b, p 268, North Collingham sub-district
- Taylor, Elizabeth, Newark, vol 7b, p 275, Newark sub-district
Remember from the beginning that we said that Winthorpe was in North Collingham sub-district. Thus, we can infer that the birth we want is very likely the one on page 268.
Furthermore, we can infer that the first Elizabeth Taylor's birth was likely registered towards the end of the quarter (around September), since her birth appears on page 268 near the end of the page range for that sub-district. On the other hand, the second Elizabeth born in Newark sub-district was likely registered at the beginning of the quarter (around July), since she appears on the first page for that sub-district. Note that we can only guess at the approximate registration date and not the birth date.
While this may not seem like a lot of new information in this example, I have used this method with great success in narrowing to the correct birth or death entry. It is obviously most useful with common names, but it is also useful with less common names or if you want to narrow down what parishes to first start searching for a baptism or burial.
The above method can be used for births and deaths from 1837 up to 1936. After that date, the registers were organized differently without the breaks between sub-districts, meaning unfortunately this method cannot be used after that date.
It is important to keep in mind that the data being used to generate these page ranges is far from perfect. Both the FreeBMD transcription and the original index have their issues. However, despite this I have found that with a little creativity (such as cross-referencing with other sources) it is usually possible to determine the correct page ranges.