Led astray by the census! Or, being sensible about public records.

My Sherwoods have been a fascinating -- and exasperating -- puzzle. I wrote some about this last time, but there have been new developments, and they might shed light on some of your genealogical problems.

My ancestor, Thomas E. Sherwood -- what IS that middle name? -- seems to have been born in Kentucky between 1790 and 1800, and is in Bartholomew County, Indiana, by 1828, when he is involved in a land sale with Azalad Maskal, a relative of his wife’s previous husband.  Got that?

Elizabeth Evans Maskal Sherwood had three children (including my great-grandmother, Nancy Jane) then died, about 1835. Thomas married again very shortly, to a woman whose name was recorded as Jane Fisher. They had two children and Thomas died three years later.

Some of  Thomas’s children were still young enough to be in the family household as late as 1850, and since that year’s census is the first to name everyone in the household, with age and birthplace, it seemed like a good place to follow up on this group.

I could find just one Sherwood household in the county.  It was headed, the enumerator wrote, by James Sherwood, male, white, age 40, b VA.  Three youngsters were listed, two of whose names I recognized.  The third was a girl, Susan, age 18, also b VA. All had ditto marks for the Sherwood surname.

So who was this James?  He seemed about the right age to be a brother of the deceased Thomas. And even if Thomas lived and died in Kentucky he could have been from Virginia, especially since that state’s counties extended to include Kentucky until 1792.

To me, it appeared from this census record that Thomas’s second wife, Jane, had also died and a Sherwood relative had stepped in, bringing a child with him.  

Then … I found evidence that Jane Sherwood had not died, as she is shown in court records as late as 1857 acting as guardian for Susan FISHER and the two youngest Sherwood children.  

This made me revisit the 1850 census and take a hard look at “James”, the 40-year-old male. It soon became obvious that despite the well-defined handwriting -- no mistake in interpretation -- this was JANE, and Susan [Sherwood] was Susan Fisher.  The enumerator must have asked one of the children, or a neighbor, and misunderstood the reply.  I’ve seen mistakes in census records before, but none like this!

The lesson here is: look at ALL records with a skeptic’s eye.  Seek out additional sources. Go back again and again to that puzzling entry and study it carefully. 

Of course now that I know Jane was guardian of a Fisher child, I am ready to seek a record of her earlier marriage, to the child’s father, and of course, her own birth name.

Later census records

A while ago I was expounding on the various aspects of the United States Federal Population Census records, and had gotten as far as the (mostly) doomed 1890 enumeration. So let's carry on.

The 1900 census has more questions than any other, and so is a great source of information for the genealogist working in that time period.. The form has spaces for indicating each person's month and year of birth – so those dates can be narrowed considerably. There is also a column for the number of children borne by the women, and number still living. "Years married" is also asked. All these help one draw, confirm or discard conclusions about a family and its individual members.

Another helpful section covers citizenship. If you are looking for immigration information, be sure to check this. Individuals are supposed to indicate their year of immigration, number of years in the U.S., and whether naturalized.

The 1910 census does not ask for month or year of birth, just the person's age (which, as you know by now, can vary considerably from one census to the next, often depending on who is giving the information). It does, however, expand on occupation, asking not only for "trade or profession" but also "type of business," and whether the individual is an employee, employer, or "works on own account." It also has a column to indicate whether the person is a veteran, an often overlooked item which could lead to further information sources.

1920 is much the same, though it has space to indicate not just street name, as in the previous census, but the actual address. In addition to birthplaces of the individuals and each of their parents, it has space to indicate "mother tongue." Citizenship, too, is addressed again.

There is an interesting story behind the original indexing of some of these records. William Dollarhide, in his "The Census Book" (Heritage Quest, 2001) discusses the original intent of the Soundex indexes, which were created for 1880 through 1920. They established a code for surnames in an attempt to bring together names which were phonetically similar (like Cain, Cane, Kain, Keen, Koeghn, etc.). The intent was to enable the Social Security system, which would begin making payments in 1937, to locate individuals in the census records and thus prove potential recipients' ages. Every-name indexes were done in this Soundex code for the 1900, 1910 and 1920 censuses, while the 1800 census was indexed only for families with children aged ten years or under.

In pre-computer days, we genealogists used to have to determine the code for our surname of interest (initial letter plus three numbers), find and search the corresponding microfilm, then use that information to find the census record itself.

There are other census records besides those designated "population." Another time.

1880 and all that

Back to the census records, or more precisely, the U.S. Federal Population Census records. Now that we can bring up the filmed images on our computers so easily, it helps to take a fresh look and be sure we know what we can expect to glean from them.

In previous blog entries I have written about the censuses taken from 1790 through 1870. Now it's time to take a look at the best-known one, compiled in 1880. According to William Dollarhide, in The Census Book, the US population that year was over 50 million, in 38 states and 10 territories – a far cry from the earliest census year, when just under 4 million residents lived in 14 states (the original colonies plus newborn Vermont).

Administrative changes improved the quantity and quality of information gathered in 1880, for which we genealogists may be truly thankful. For the first time, the relationship to the head of household was given for each person in that household. It does save a lot of guessing! An individual with a different surname may be listed as an in-law, thus providing the probable birth name of the wife. Older children with a different surname may be from the mother's previous marriage, even if they are listed as children rather than stepchildren of the head. Birthplaces of each person's parents are also listed – a useful clue (though occasionally incorrect). For the first time the street name and number are supposed to be recorded for each household in cities and towns, which adds a bit of history.

An aside: Familysearch.org has long had a very user-friendly every-name index to the 1880 census, with much detail. It is tempting to look at that, perhaps print it out, and consider the job done. But it is always useful to study the actual image of the page – there may be additional gems of information lurking!

After the wealth of detail found in the 1880 census, we come to a black hole. More than 99 percent of the 1890 census was destroyed as a result of fire in 1921. All that remains are tiny fragments of the original and a portion of a special schedule of Union veterans and their widows.. Unless you are very lucky, your census searches will be blocked between 1880 and 1900, quite long enough for young women to have been born and married, thus changing their surnames. Ancestry.com does have the fragments of the population schedule, but not the special listing of the Union veterans and widows (last time I looked).

Three states and two territories did take censuses in 1885, with federal assistance: Colorado, Dakota Territory, Florida, Nebraska and New Mexico Territory. Ancestry.com has these, and they can help fill in some gaps.

Another special set of records which may be use are the "mortality schedules," taken at the same time as the population census for 1850 through 1880. They list the names of persons who died within twelve months before the official census day. All list the name, age, marital status, and cause of death, and some list the page and family numbers from the population schedules, which enables the researcher to trace the deceased back to his/her family.

Again, I am only touching on the highlights of all these extremely useful records, but, even so, will have to defer discussion of the post-1880 population schedules for another time. Happy hunting!

More Counting Noses

Last time I discussed the pre-1850 U.S Federal Population Census records, which, despite their seemingly meager data, can provide useful clues for the genealogist. It is wise not to overlook these handwritten documents which have been copied and recopied, indexed and microfilmed. It is the filmed images of those second-hand copies we are looking at on Heritage Quest or Ancestry.com, and while it is important to allow for the possibility of errors, one must be grateful that even these sparse records still exist in some form.

By 1850, more information becomes available, though sadly Heritage Quest has not indexed 1830, 1840, or 1850. Try Ancestry.com at the library – I have found that computers are almost always available here in Sonoma, and, if not, time limits will give you a turn eventually. Another avenue is usgenweb.org. This free site will lead you directly to your state of interest, and the counties within it. Check to see if a particular county's census has been indexed online by some individual or organization. If so, make a note of the location details and page number, then browse HQ's images, which are available for all extant censuses. (Be aware that page numbering can be confusing, as there are often two sets of numbers. Also, there may be two sides to a "page," meaning two images have the same number.) Even if an index contains extensive information, try to find a copy of the filmed image – there may be more!

William Dollarhide's The Census Book also lists published census indexes for various states. Some include added information that the indexers found in other sources. Locate these online, by borrowing them on Inter-Library Loan, or in the Family History Library catalog, which may list filmed copies you can borrow to view at the local Family History Center.

If you have searched 1850 census records, you know that it is the first one to list all persons in each household, with their ages as of the census day (in theory), and their places of birth. There are other details, but these are the ones we usually go for first. Relationships are not shown, as they are from 1880 on, so be wary of assuming the oldest male and female in a household are necessarily husband and wife. It is quite likely, but further proof should be sought. Assuming for the moment that they are a married couple, look at the age of the oldest child to estimate the probable date of marriage (if the "wife" is 21 and the next oldest individual is 15, there must have been a previous marriage, or the mother has died and these are both children of the head of household. Look at his age to see if this is plausible.

Listing of birthplaces gives a picture of the family members' origins and wanderings. The parents may have been born in different states, and their children in yet other states. Some people just have a hard time settling down! With all names listed, locating an unmarried brother-in-law or widowed parent in a household, with his or her birthplace, may provide additional clues for research. Of course, these may not be relatives, but make the assumption to see if it can be proven.

1860 and 1870 censuses have pretty much the same information – all three have space for listing the value of the householder's real estate, occupation, literacy, and whether married in the year preceding. (This can be an indirect way of determining whether there IS a marriage) Occupations and indication of literacy can help distinguish between two individuals with the same name. The fellow who teaches school in 1850 is not likely to be the person "unable to write or write" listed in 1860.

Next time I'll discuss the 1880 census, everyone's favorite (mine, at least), and later enumerations.

Counting noses

The ease of pursuing family history has certainly improved in the last few years. Computers and the Internet are increasing the amount of information we have at our fingertips and the means we have of analyzing and storing it all.

A case in point is that mainstay of American genealogical research, the U.S. Federal Population Census, more often simply referred to as "the census."

It used to be one consulted whatever printed indexes were available, usually at some far-away library, made note of the likely-looking references, then sent for the microfilm and spent hours looking at it, then writing out the findings. I still have pages and pages of these handwritten notes, some as difficult to read as the original images! (And often missing some data because I didn't realize its importance at the time.)

Now, from home, we can go to the local public library's website, enter our library card number, and gain access to full indexes and images of 11 of the censuses, partial index to one (1930), and browse four others – all offered by Heritage Quest through the local library system's subscription. Ancestry.com has even more census records, recently made available in-house at local libraries. [Since they were indexed separately, it is wise to consult both sources.]

With this abundance – almost an overabundance – of data at our beck and call, we might do well to revisit the unique features of these decennial records and make sure we squeezing every bit of information out of them. I find myself going back time and again, to check on one more piece of data from a census page, in order to enlarge my picture of an ancestor or relative, or to provide a possible clue for additional research.

A reminder before I continue: census records were done on a county-by-county basis, and over time boundaries changed, new counties (and states) were created and, in some cases, early county designations were abandoned or revised. Be sure you are looking in the right place for that ancestor! I find William Dollarhide's Map Guide to the U.S. Federal Censuses, 1790-1920 extremely useful for confirming county boundaries at the time of each census enumeration. Another good source is Everton's Handybook, which has current state maps with counties, but also gives dates and details of various jurisdictional changes that have occurred over time.

The Constitution decreed that a count of the population be taken beginning in 1790. There were various means of counting (or excluding) certain residents, which we all remember from our school civic classes (don't we?), but for the most part, free heads of households were named and they and their family members sorted by gender into broad age ranges. The lists up through 1840 are certainly not as fact-filled as later records, but should not be ignored. They give a picture of the family size, the number of males and females, and a general idea of their ages, which can be narrowed down by studying a series of census years. Sometimes the presence of an older relative can be inferred as well.

Because the counts were taken by census takers usually (but not always) proceeding from one dwelling to the next, friends and relatives may be found clustered together. The presence of individuals with differing surnames but in close proximity may indicate a married daughter remaining near her parents' or brother's home.

A quirk which misleads some genealogists at first was the practice, in 1820, of counting free white males in overlapping categories: ages 16-18, and 16-26. It is important to make note of this – if only one person is indicated in each of these boxes, it is the same person. Don't count him twice.

There are numerous other details about these early census records, too many to recount here. For more information, I recommend another book by Dollarhide: The Census Book. He discusses each count in detail, with notes on availability, and gives some intriguing background.

Next time I'll write about the later census years – the ones that record all those helpful extra details, like names of all persons in the household, their ages, place of birth, occupations and relationships.