I have found that it's worth noting both dates. Some census takers followed the rules as they were written, and some didn't. It may be that the people they were talking to didn't understand what was asked too...who knows. But researchers need to remember both scenarios are possibilities: age according to the actual census date, and age on the date the census taker came. It's like today. Not everybody follows the rules--and maybe it didn't get explained to them correctly so that they understood what was wanted. I'm just glad we've got the records that we do. Karen