Having drawn a blank on both Theodore Kelly (see ‘The wrong Theodore’) and John Kelly (see ‘Thank you Thomas Cromwell’), I have looked again at Susan Kelly. Susan married Jeremiah Cronin in Michigan in the late 1850s, just prior to the establishment of a central recording system for the State. She stayed in Marshall, Calhoun County, Michigan for the rest of her life. This helps enormously as Michigan records contain a lot of useful information. Her husband Jeremiah died there in 1889 from Bright’s Disease.
The 1900 Census shows Susan living at 413 North Madison Street. She is the head of the household. In the house with her are her children Frank (or Francis), Jeremiah and Susan. Their ages match the information on the 1880 census. Twin grandsons William and Charles Duffield (age 11) are also present. Their mother was Mary Cronin. She married Edward Duffield in April 1888 and died in September 1892.
Susan’s date of immigration is shown as 1847. Strangely, under month and year of birth, the entry reads “don’t know”. The 1880 census records her as being 43, and therefore born in 1837. The 1860 census backs up this information. (I have not found them on the 1870 Census.)
Her death certificate provides more information, supplied by her son, Jeremiah. She died on 14th October 1907. The main cause of death was typhoid fever. She was born, according to Jeremiah, in May 1837, but the exact date was unknown. She married at 18 (so in 1855) and had given birth to nine children, of which five were still living. The name of her mother was unknown. Her father’s name is difficult to read. It appears to say William. (See picture) I would be interested in other possible interpretations. 
It seems slightly odd that her year of birth was known in 1860, 1880 and again in 1907, but not in 1900. Perhaps a family debate followed the 1900 census during which Susan’s memories of her earliest days were disseminated among the family. This does not prove that Susan was, in fact, born in 1837, but there is clear evidence that she believed this to be the case. My Grandfather, Patrick, believed that he was born in 1884 (as demonstrated on his army enlistment papers and his marriage certificate) but the correct date was 1883.
The information provided on the census returns and the death certificate will now prompt another search of the immigration records to see if I can find out more about her siblings.