This holiday season I’ve been amusing myself with the Ancestry website. It’s a great genealogy resource for people with ancestors from Britain.

Thousands of users have already posted public family trees on the website. I found nearly everything I wanted to know by looking at those trees and double-checking the sources.

Each person on a tree has their own profile with links to records such as census data, marriage certificates, and baptism registers. If you’re really lucky, you’ll find that some users have also uploaded pictures and newspaper clipping from their research outside the website.

I always check the records to see if the facts go well with what I already know. It’s common to find wrong turns such as two men with similar names but different home addresses, wives, and offspring.

I was able to speak to relatives born in the 1940s and I had access to some family papers. So, I was able to confirm the accuracy of my family trees back to the middle of the 19th century. One line went back even further because a very distant cousin, a writer, had posted her own independent research on Ancestry.

I was able to check her data very carefully because these ancestors were gentry and well documented. It’s much more difficult tracing ancestors from poorer backgrounds. Variations in the spelling of names was common and some people were also inconsistent about how old they were.

One woman caught my eye. Her name was Tryphena. She appears on at least 27 public family trees, but her details are often muddled.

Tryphena told census takers that she was born in Newtown, Montgomeryshire. Some website users have confused her with a Tryphena from Newport, Monmouthshire. Both places are in Wales.

On at least one profile her name is given as Sarah Tryphena Wade. I found the record of Sarah Wade’s marriage to George Wilby and the record for the baptism of their firstborn child. In the census a few years later, George and two children are living with someone called Tryphena. Were Sarah and Tryphena the same person?

It was common for people to prefer their middle names. It seemed possible that Sarah always went by Tryphena and gave that name to census takers.

However, in the marriage register it said that Sarah’s father was a farmer and that her residence was Long Acre. Places in Wales did have, and still do have, Welsh and English names. But equally, Long Acre could have been a farm in England.

I considered the possibility that Tryphena might be George’s second wife. I then looked for some official record of her in Montgomeryshire. This was easier said than done, as I kept tripping over the Tryphena who lived in Monmouthshire.

I finally struck gold when I allowed for variations in the spelling of Tryphena. I found a girl of the right age, baptised in Llanllwchaearn near Newtown in Montgomeryshire (the county is now called Powys). This is what the baptism register said:

Trephina, base daughter of Mary Whittaker and reputed father Edward Turner (spinner), of Penygloddfa. Baptised 17th April, 1836. By Charles Wingfield at the parish of Llanllwchaiarn in the county of Montgomery.

I then looked for official evidence that this Tryphena had actually married George in London. I found none. It could be possible that Tryphena was a non-conformist who wasn’t married by the Church of England. I also found no evidence for the Anglican baptism of the second of the two daughters listed with George and Tryphena on the census. Did Tryphena have her baptised in a chapel?

There were Welsh non-conformist chapels in London for Welsh migrants. There’s a strong possibility that Tryphena spoke Welsh and might have wanted to stay connected with other Welsh speakers in London.

I couldn’t find evidence for the death of George’s first wife, Sarah Wade/Wilby, in London. There were other Sarah Wades and Wilbys elsewhere in the country, but it was impossible to know whether they were George’s Sarah. A record of death would have helped establish whether Tryphena was the biological mother of the younger daughter.

One possible version of events is that Tryphena, like many Welsh people, came to London to find work. She met George, who was then a single father to one daughter. They then married in a non-conformist chapel and had one more daughter. Many years later, when the eldest daughter married in an Anglican church, her stepmother Tryphena was a witness at the wedding.

Of course, without documentation, there is no way to know for sure. The Tryphena I found in Wales may not be George’s Tryphena. Or if she was, perhaps they never married, or perhaps both daughters were Sarah’s.

What is clear is that a Tryphena from Montgomeryshire had a long relationship with George Wilby and raised two daughters with him in East London. She was therefore a significant part of their family’s story, as was Sarah Wade of Long Acre.

Identities became muddled on the Ancestry website for all kinds of reasons. The 19th-century handwriting on census forms is difficult to read, so confusing Monmouthshire with Montgomeryshire is understandable. Many of Ancestry’s users are not in the UK and may not be aware of historic county names.

The variations in the spelling of Tryphena’s name made baptism records hard to find. Later, the absence of an entry in a Church of England marriage register obscured the possibility of Tryphena being a second wife.

If you use Ancestry, never take the information on the profiles at face value. Similarly, the wrong documents can be attached to pages that appear helpful at first glance. For example, I thought I found evidence of the Monmouthshire Tryphena’s death, but when I looked at the document she wasn’t named.

I’m quietly confident that my Tryphena was Trephina, daughter of Mary. I’ll never know for sure, but that’s the nature of historical research.