After months of research I had finally figured out the entire family (give or take) of my great great grandfather. My conclusions are sound - at least I think they are. After a good bit of searching I finally found the "Bernhardt" family mis-enumerated as "Bernhalt". Unfortunately, in my excitement I failed to notice the ages are completely wrong for the children.
The family up to 1860 should include Franz b 1833, Anna Catharina (known as Catherine) b 1836, Anna Marie Elizabeth b 1857, George Rudolph b 1859, and George b 1860. I know the family had these children based on the death certificates of these children and the home addresses on the death certificates compared to known city directory entries and other documentation for Franz.
Here is mom and dad at the bottom of one page:
And the children at the top of the next page:
Franz/Francis is two years off in birth, but birthplace and occupation are correct. The name is not concerning as I have seen other places where it was written wrong. Catherine's age is a bit off but the birthplace is correct. The children's ages, however, are significantly off, given that they are just children. Two years off on age for an adult - well, that's one thing. But it's rather difficult to confuse a 4 month old with a 3 year old.
Of further interest is the next family: A family that seemingly lives in the same dwelling but is considered a different household; a family that has a one year old child, Caroline, listed first; a family with two members who are employed as tailor's.
My initial thought is respondent error: if these two households lived in the same dwelling then it's certainly possible that the respondent was not a member of the immediate Bernhardt family, if family at all. Could these two tailor's have been employed by Franz?
Facts not in dispute: Franz was a tailor. He was from Saxony. Catherine was from Hanover. They had children with these names. This ward in Philadelphia is the area where Franz and Anna Catharina resided subsequently.
It's certainly possible that this is merely a family with many similarities to mine. But is that the most reasonable to conclusion to draw? Or is the simplest conclusion actually respondent error?