I first came across it in a course on English family history. What the tutor meant was that we should always collect the entire set of vital record details for an ancestor, including the death records, and not just be satisfied with their birth and marriage or census data.
Recently, revisiting a branch of my English family tree, that I had only barely scratched the surface of, I proved why this advice is so valuable.
In the 1851 English census I was able to find my 3 x great-grandmother Mary Ann Westlake, whose maiden name was Legg, married to Thomas Westlake, a Brass Founder and Plumber in the Devon city of Plymouth. From these records for Thomas and Mary Ann I noticed that they were both the same age, having been born in 1818.
When I moved on to find them in the 1861 census and noted that the transcript on one look-up site had Thomas' wife listed as "Clara M Westlake" but as her date of birth was still 1818 I just put this down to a transcription error. Opening the image I could see that the writing was none too clear, giving the transcriber a bit of a job to work out. What it certain, though, is it didn't look anything like Mary Ann!
Popping over to another subscription site and the transcription for their 1851 census was given as "Chrisk W".
At a third site and the same 1851 census I got the transcription returned as "Catherine W". The writing on the census page had challenged the transcribers at all three sites and I can not blame them for their differing attempts to make sense of the entry as I certainly couldn't.
So what had happened to Mary Ann? Had she tired of her name and changed it to something more exotic? Or had she died and Thomas had taken a new wife, who also happened to have been born in the same year as he and the former Mrs Westlake? I decided to do some detective work and search for a death of Mary Ann Westlake from after the 1851 census and before the 1861. What I found was a number of candidates that could have been my great-great-great-grandmother.
So now I approached the problem by seeing if I could find a second marriage for Thomas and here I can testify to the usefulness of the advice to "always kill off your ancestors". You see, by having done just this for Thomas, having found his death in the records and then the listing for his probate, I was able to discover that he had an unusual middle name of "Scoble".
Now I could look for a marriage of Thomas Scoble Westlake and I found just the two in the databases. One was in 1841 to Mary Ann Legg in Stoke Damerel, which is in the Devonport area. The other was to Christian Upcott Harwood in the last quarter of 1859 in Falmouth, Cornwall. I had the name of the second wife!
Though this asked the question, if Thomas and Christian were wed in 1859, then what had happened to Mary Ann? The records show that in the second quarter of 1859 a death was registered in Plymouth for her, allowing Thomas to take a new wife in the fourth quarter! I will need to order a copy of the death certificate to find out if this is the correct person and also what she died of, but the assumption is there.
So who was Christian Upcott Harwood? I had looked for her birth or christening without any luck. Then it struck me that perhaps she too was a widow. I now looked for the marriage of a Christian Upcott, leaving the bride's maiden name blank, to someone called Harwood and I found one to Samuel Peter Harwood in 1841 in Lewisham. Christian was from Plymouth, Devon and he was from Plumstead in Kent.
A death occurred in East Stonehouse, Devon in the year 1858 to one Samuel Harwood and I assume it was his widow who married Thomas Scoble Westlake. Family history is a fascinating subject. Like all topics, however, you need to approach it with a certain discipline or mistakes will creep into you family tree. There are rules and conventions and my advice is to learn as much about them to prevent you "barking up the wrong family tree".