I came across such a weakness while examining the 1891 Enumeration of Tennessee Voters on Ancestry, one of the primary replacements for the lost 1890 federal census in Tennessee, for members of the Shelton family. Since I have already found Isaac Z. Shelton in Overton County, Tennessee, in the 1880 federal census and in White County, Tennessee, in the 1900 federal census, I first focused my search in those two locations.
Among the search results were the following males over the age of 21 within my two counties:
- O. L. Shelton in White County
- Jacob M. Shelton in Overton County
- Jake Shelton in Overton County
- Spolwood Shelton in Overton County
Seeing that Isaac is not represented, nor does he appear with his name or various variations, I could have stopped there, but good research practice means looking at the actual images, not just relying on the search results from the database index.
Paging through the images for White County I discovered another Shelton who was somehow missed by the indexers:
- M. L. Shelton
I have so far not uncovered an ancestor by those initials, but I am certainly filing him in my to-be-researched list.
Double-checking the entries from Overton County, I found that "Spolwood" is actually recorded as Spotswood Shelton on the image. The cross bar for the t is located above the vertical stroke, but is consistent with other Ts on the page.
Have I located my ancestor Isaac Z. Shelton yet? No, but I now know that I will likely need to continue examining each image to truly say he is not in the 1891 census substitute.
Without indexes we would spend hours having to page through records. However, indexes can contain transcription errors and/or may not contain all of the information we need. Always look at the actual record.