Dear Genea-Man: Why can't I find my ancestor in the 1900 census?

Dear Genea-Man: Why can't I find my ancestor in the 1900 census? I know his full name, his date of birth, his parents names, and some of his sibling names.

Genea-Man responds: Even though you know all of those facts, it is possible that the census enumerator didn't get them right - he may have misspelled the names, made mistakes on the dates, or mixed up your family with another family. We don't know who gave the information to the enumerator - was it one of the parents, or one of the children, or the housekeeper, or a neighbor?

Even if the enumerator got most of it correct, what did the indexer see on the census page years later? Was the enumerator's handwriting clear? Did the ink on the page fade, or was the page damaged before it was indexed? Even worse, what if the indexer had a bad day or intentionally messed up? The latter is probably far-fetched...

Based on my own research, and opinions shared by others with me, it appears that perhaps 75% to 85% of all names were enumerated and indexed fairly well. Another 5% to 15% are poorly enumerated or indexed, but are able to be found using the indexes and creative searches. The last 5% to 15% were either not enumerated, or the records are not available or readable, or were enumerated and indexed so badly that they can't be found.

Searchers need to understand that there are differences between the two most popular online census search sites:

1) http://www.ancestry.com (a personal-use subscription site, or free at libraries with Ancestry Library Edition) has all of the US Census records available online. They have a Head of Household index for 1790 through 1840 and an Every-Name index for 1850 to 1930. Ancestry lets you put in a birth year and select an age range (plus or minus 0 years, 1 year, 2 years, 5 years, 10 years or 20 years). Ancestry permits wild cards in the given name and surname fields (a minimum of 3 letters then an asterisk), and permits an exact search (a check in the search box) or a Soundex search (uncheck the box).

2) The other census provider is HeritageQuestOnline which has all of the US census images, but not all names are indexed on HQO. They have a Head of Household index for 1790 to 1820, 1860 to 1880 and 1900 to 1920 only. They have a Head of Household index for the 1930 census for some states, but not all states. The spouses and children in a family are not indexed on HQO. HQO requires a full surname - no wild cards and no soundex-type search can be made. HQO lets you specify age ranges in 10 year increments (e.g., 0 to 10, 11-20, 21-30, etc.). You can get a maximum of 1,000 results on HQO - if there are more then you have to restrict your search to a smaller locality (e.g., All States to State to County to Town).

My list of "tricks to try" for elusive ancestors in the census records are to:

a) Write out the surname in longhand script. For each letter, identify other letters that are often mistaken for that letter (e.g., L and S, R and K, m and n, b and l, etc.). You can then combine letters to create alternate surnames (for example: Seaver could be Leaver, Seaner, Searer, Scaver, Seuver, Seaven, Saever, etc.) You could leave out a vowel (e.g., Sever, Lever, Seavr, etc.) or add a letter (e.g. Severs, Seavers, Seavern, Seavere, etc.) This will result in a list of names to try in the search box (especially on HeritageQuestOnline).

b) Sound out the surname. How would your ancestor have pronounced it? If he pronounced it that way, how would you have written it down? What if the enumerator was a different nationality than the householder? How would a Norwegian spell the New England name Seaver ("Sea-vah") in Wisconsin? This also results in a list of names to try.

c) Can you limit your search to a specific township, county or state? If so, search only in that locality. This reduces the number of results to a more manageable number. If you don't find your target, then expand your search to the next higher governmental division - Town to County to State to All States.

d) Do a search with only the surname. If you know a birth year and birth state, search using those items also. I always use a date range of +/- 2 years for children and +/- 5 years for adults. If those don't work, I expand the range to the next highest choice.

e) Do a search with only a given name and no surname, but with a birth year (and range) and birth state.

f) Do a search without a given name or a surname, but with a birth year (and range) and birth state. You may get too many results in a large city, so this works best if you are searching a small county or if the birth place is not a nearby state.

g) If you know the given names of the children, pick one of the names that is uncommon (e.g., I would use Josephine rather than Mary if that was a choice) without a surname, and use the birth year (with a range of +/- 2 years).

h) Put the given name or middle name initials in the given name field - some enumerators used only initials. You will get results with the initial as either the given or middle name.

i) Did the surname have a prefix like De, Mc, Mac, O, Van, Von, on the surname? The indexer might have listed it by the last part of the surname (e.g., "Knew" for "McKnew").

j) Did the enumerator or indexer leave out a letter in the name? I recently looked for a "Crosby" and it was enumerated and indexed as "Crsby." Try alternative spellings without each letter.

k) Did the enumerator or indexer misspell the given name or abbreviate it (for example Jno instead of John, Geo for George, Goerge for George, Cahrles for Charles, etc.).

l) Was a middle name used as a first name by the householder? This is fairly common.

m) Did the enumerator transpose the householder names (i.e., first name last, last name first) for the head of household?

n) For a blended family, were the children enumerated with the head of household's surname? I've seen this several times.

o) If you are using Ancestry.com, then use the wild card search capability. You can put the first 3 letters of each name in the search box and then an asterisk (like Isa* and Sea* for Isaac Seaver) and see a list of candidates. The results list will often show the spelling variation used by the enumerator or indexer. You can also put a wild-card "?" in place of one letter - for example, input Thom?son and get Thompson and Thomason. This only works with the first three letters known.

p) On Ancestry.com, you can input additional names on the 1880 to 1930 census searches. You can input one or all of the father's given name, the mother's given name, the spouse's given name, and their birth places. This can be very effective especially for uncommon names. Of course, if one of those is wrong in the records, you won't find the target.

q) Ancestry.com lets you search with an "Exact search" box checked or unchecked. If the box is left unchecked, you will get a long list of results, with the best matches first. I usually check the box and use the other "tricks".

r) Sometimes "less information" works better than "more information." You don't know how they were enumerated. If you know the full name (e.g.) "Frederick Walton Seaver" then you might search for "Fre* Sea*" or "Wal* Sea*" in a county rather than put the full name into the search boxes and the town/county names. If there are too many results, then reduce them with a birth year and range and/or a birthplace, if known.

s) Have you checked all the available indexes? The Ancestry index and HeritageQuestOnline index were done at different times by different people, and there are many differences between the search terms and the results.

t) As a last resort, and if you are fairly sure where they resided in the census year, you can search line-by-line either online or on microfilm. City directories may help pinpoint a ward or enumeration district in a large city.

While you think that 100 or more results is "just too many to search through," the reality is that you can usually sort through them quickly. Ancestry.com will provide the spouse's name or the parents names (for children) for 1880 to 1930 searches. It sure beats going page by page, whether on a microfilm reel or online.

The census records are like a haystack. You are searching for a few needles in that haystack. The indexes currently available online are tools we use to find those needles. In many cases, they work wonderfully - we can usually find the actual census image online in minutes rather than weeks (as we did pre-2002 with microfilms).