This past year, I spent a considerable amount of time scanning Mesa City Cemetery documents for FamilySearch.org. As I went through the documents, one by one, I became both impressed and depressed at the relatively huge number of children who had died either in childbirth or before the age of majority. I did not keep any statistics, but the impression I got was that a considerable number of children died with no other record made of their lives.
In looking at many pedigrees from a variety of individuals over the past years, I find that it is not uncommon to show only the direct-line ancestor as an only child in an ancestral family. But what is more important, the dates recorded are neither believable or even possible.
FamilySearch.org's Family Tree program is a valuable tool in analyzing families for omitted children. Because the program purports to be a compilation of many pedigrees, the data should be more accurate than not. But is this the case?
First, it is apparent that during the 1800s and 1700s and earlier, that families with only a single child were likely rare. Any pedigree showing only one child becomes automatically suspect, especially if there is an appreciable gap between the assumed marriage date and the date of the birth of the only recorded child. Notwithstanding this obvious fact, I continually find pedigrees that record only one child per family going back into antiquity. The existence of these "single child" families is an immediate flag that the information is, at best, incomplete and likely inaccurate as well. But even when multiple births are recorded, it is as if reality takes a vacation when it comes to recording family information.
Almost without fail, I can look at a pedigree, even my own, and find either gaps in the birth years of the children or delayed births from a recorded marriage date. The consistency of the dates is also a question. All of these issues are red flags indicating insufficient research and information on that particular family.
For example, choosing one of my own lines at random as reported by FamilySearch.org's Family Tree, I find the following starting back a few generations:
Sidney Tanner (b. 1809, d. 1895) (my Great-great-grandfather)
John Tanner (b. 1778, d. 1850) md. Lydia Stewart (b. 1773, d. 1825)
Lydia Stewart's parents are recorded as:
William Stewart (b. 1740, d. 1826) md. abt. 1768, Amy Hutton (b. 1755, d. 1839)
William Stewart (b. 1700, d. 1753) md. 3 Oct 1682 Margaret Sanderson (b. 1704, d. deceased)
The first child of William Stewart and Margaret Sanderson is listed as Clark Stewart, (b. 1732, d. deceased). The last child is reported as Elizabeth Stewart (b. 1748, d. deceased).
OK, what do you see? A real mess? Yes. The first William Stewart apparently lived to the ripe old age of 86 years and the last reported child, Sidney Stewart (b. 1813, d. 1814) was born when he was 73 years old. At the time of this last recorded birth, Amy Hutton Stewart was 58 years old. Not impossible, just unlikely.
Things really get interesting in the next generation. The second William Stewart was married 18 years before he was born and William Stewart and Margaret Sanderson's first child was born when his father was 32 years old and his mother was 28.
You can easily see that the dates recorded are not accurate. Let's just say that my own dates, in my own data files do not agree with those reported.
I give this as an example of the need to review, with a little simple math, the dates, gaps in birth dates and other inconsistencies the crop up regularly in pedigrees. Especially those pedigrees that are online conglomerations of a lot of different submissions.