Sunday, May 1, 2016

The Case of the Two Wives: Using Ancestry Suggested Documents to Sort Out Wives

I have had two cases of double wives lately that caused quite a bit of confusion: Both Ancestry and the family trees I was able to connect to indicated that two different wives of two men – George Arrington and Eldon Clide Clement – were one and the same person. Yet there was very strong evidence in both cases that there were definitely two separate wives.

 Case # 1 – George Washington Arrington (7 Oct 1895 (1,2) to 7 Dec 1977 (3,4)).

First wife – Freddie Gertrude Warren (born July 1898 in Tennessee (5), died before 1940).

According to the Collin County, Texas, Marriage Index, 1800-2010 Record, accessed on, they married in 1916.

They had three children:

Modine – born 5 Nov 1918 in Texas and died 29 Jul 1919 in Farmersville, Collin County, Texas from a fall. (6) The informant on the Death Certificate was W. H. Warren, and this would later give me a clue as to which Warren family Freddie belonged to.

James Weldon Arrington, born approximately 1927 in Texas (7), no further information.

Myra Jean Arrington, born approximately 1927 in Texas (same source as for her brother James Weldon), no further information. (Getting this Myra Jean Arrington sorted out from the daughter of George’s brother Eugene Grey Arrington, also named Myra Jean, also born 1927 in Texas (8), was another matter – Ancestry kept suggesting the wrong document (only one) for her.

Second wife – Mattie Niece Bratton, born 18 Jan 1898 in Missouri (9, 10)

Although Ancestry suggested and continues to suggest that Mattie Niece Bratton and Freddie Gertrude Warren were one and the same person, and they were born in the same year, there were some important differences. There were the names, to begin with, and places of birth also differed: Missouri for Mattie and Tennessee for Freddie.

It was the appearance of Jennie Hughes as the sister-in-law of George Washington Arrington that served as indirect proof and eventually led me to the discovery that Mattie Bratton was not only not Freddie Warren, but showed me who her family was – she was the daughter of Addison Lowery Bratton and Cora Charlotte Worley. When I clicked on the name of Jennie Hughes in the 1940 census, one of the suggested documents was the 1900 census, which showed both Jennie and Mattie as daughters of Addison and Cora Bratton. Both the 1920 (12) and 1930 (13) censuses showed Jennie with her then husband, Delbert E. Hughes, and the 1930 (13) census showed her with her son Neal, who appears with her on the 1940 census with George and Mattie Arrington.

Once I had Mattie in the right family, all sorts of information popped up for her, including the California Death Index, which showed that she died 20 Feb 1990 in Anaheim, Orange County, California. (14) It also provided her maiden name/father’s name Bratton and mother’s surname Worley. Moreover, a glance at the family trees on Ancestry revealed that Bratton researchers were aware of her marriage to George Arrington, and even had pictures of him that I had not seen. The 1920 (15) and 1930 (16) censuses revealed that Mattie had been married before, to an Isaac Nevit Barto and had children Lucille Fern Barto and Jack Barto from that marriage.

Meanwhile, I still had no date or place of death for Freddie Gertrude Warren. Searches in the Texas Death Index and Texas Death certificates turned up nothing. A glance at family trees on Ancestry indicated that all agreed that her father was John Bence Warren, but some had her mother as Lula J. Davidson and others gave Mary Tennessee “Tinnie” Hammons, though only Lula shows up on the 1900, 1910, and 1920 censuses. Freddie’s brother Wallace H. Warren also appears on the 1900 and 1910 censuses; since he must have been the informant W. H. Warren on Modine Arrington’s death certificate, this is how I knew which family was Freddie’s.

Case # 2 – Eldon Clide Clement, born 21 Feb 1911 in Bonham, Fannin County, Texas; died 1 Sep 1968 in Brownwood, Brown County, Texas. (17)

First wife – Deltha Faye Martin, born 11 Apr 1915 in Milano, Milam County, Texas; died 17 Mar 1937 in Telephone, Fannin County, Texas (18)

Second Wife – Mary A. “Annie,” born 1902 in Oklahoma , no further information. (19)

As in the first case, family trees and Ancestry treated (and continue to treat) these two women as though they were the same person. Despite the difference in ages and states of birth, they could in theory have been the same person – perhaps the “Fay Martin” shown on the Texas Birth Index and in the U.S. Social Security Applications and Claim Index, 1936-2007 on as the mother of Franklin Don Clement was actually the “Annie M. Clement” shown on the 1940 census – her name could have been Fay Ann Mary Clement or Fay Ann Martin Clement, right?

But the 1940 census showed a curious thing. It showed that the Annie M. Clement on the 1940 census (and in numerous U.S. City Directories, 1822-1995 on lived in two different counties in 1935: Snyder, Scurry County, Texas for Eldon, and Cross Plains, Callahan County, Texas for Annie M. Five years is not much time to meet, court, get married, and have a four-year-old child (Franklin Don Clement). Although at one point I was convinced that Mary A./Annie M. was not a different person and even deleted her from my Reunion database (because, curiously, a 1930 census popped up that showed a Faye Martin born in 1902 in Oklahoma!), I was now certain that there was no way that these two women could be the same person. When I was able to find a Texas Death Certificate using “Fay Martin” in the search parameters, I was even able to get an Ancestry family tree that gave her correct maiden name: Deltha Fae Martin. She may be the Fay Clement who shows up with spouse E. Clement in the 1933 Fort Worth, Texas, City Directory in the U.S. City Directories, 1822-1995 on

So I think I’ve learned my lesson. All the evidence has to be lined up for and against the hypothesis that what everyone else treats as one person may actually be two different people. Ancestry often suggests the same records for two different people, but eventually you may be able to sort them out using those same suggested records.

1 World War I Draft Card of George Washington Arrington, accessed at

2 Certificate of Death 89492 of George Washington Arrington, Texas Department of Health, Bureau of Vital Statistics, accessed at

3 Certificate of Death 89492 of George Washington Arrington, Texas Department of Health, Bureau of Vital Statistics, accessed at

4 Social Security Death Index, accessed at

5 1900 US Federal Census, District 6, Lewis County, Tennessee, taken 19 Jun 1900, accessed on 26 Jan 2016

6 Both dates and cause of death were found in the Certificate of Death of Modine Arrington, Texas Department of Health, Bureau of Vital Statistics, accessed on, as was the name of the informant, W. H. Warren.

7 1930 US Federal Census, Tulsa, Tulsa County, Oklahoma, taken 8 Apr 1930, accessed on on 14 Feb 2007

8 1940 US Federal Census, Grand Island City, Hall County, Nebraska, taken 10 Apr 1940, accessed on 28 Jan 2016.

9 Social Security Death Index Record, accessed on

10 1940 US Federal Census, Tulsa, Tulsa County, Oklahoma, taken on 4 Apr 1940, accessed on 27 Jan 2016.

11 1900 US Federal Census, North Salem, Linn County, Missouri, taken on 13-14 Jun 1900, accessed on 26 Jan 2016.

12 1920 US Federal Census, Tulsa, Tulsa County, Oklahoma, taken on 19 Jan 1920, accessed on 1 May 2016.

13 1930 US Federal Census, Tulsa, Tulsa County, Oklahoma, taken 22 Apr 1930, accessed 1 May 2016 on

14 California Death Index, 1940-1997, accessed on 26 Jan 2016.

15 1920 US Federal Census, Precinct 12, Tulsa County, Oklahoma, taken on 5 Jan 1920, accessed 27 Jan 2016 on

16 1930 US Federal Census, El Centro Township, Imperial County, California, taken on 4 Apr 1930, accessed 27 Jan 2016 on

17 Certificate of Death 61494 of Eldon Clide Clement, Texas Department of Health, Bureau of Vital Statistics, accessed on 20 Apr 2016 on

18 Standard Certificate of Death 14664 of Deltha Fae Martin Clement, Texas State Department of Health, Bureau of Vital Statistics, accessed on

19 Apr 2016 19 1940 US Federal Census, Vincent, Howard County, Texas, taken on 2 May 1940, accessed on 19 Apr 2016.