… and those 10 children the offspring of 3 separate (but related!) marital pairs.
I’ve written about this before (see, for example, “Blended Families”): the blended family is nothing new. When a widower married a widow, and one or both parties to the marriage had children from a previous marriage, the resulting new household might contain a complex combination of children from two or three marital pairs. That complexity was not necessarily registered in the census returns.
In some (many?) cases, the widow’s children from her previous marriage will be listed in the census under the surname of her new husband (the children’s stepfather), even though legally, and often practically as well, those children carried the surname of their biological father.
Moreover, since a widower’s children from a first marriage carried the same surname as his children from a subsequent marriage, the census listing will generally not (as in, almost never) distinguish between the children produced by a first marriage and the children produced by a subsequent conjugal relationship.
Here, for example (and I am taking my census geekitude to a new level with a colour-coded representation), is the household of George Vallely and his second wife Mary Moyle in the 1861 Census of Canada:
The first four children (the “greens”) are the children of George Vallely and his first wife wife Anne O’Hanlon. The next four children (the “blues”) are the children of George Vallely’s second wife Mary Moyle and her first husband John Bean. But note that they are listed here with the surname Vallillee (Vallely), even though their surname was Bean. And to further complicate matters (well, whoever said genealogy was supposed to be easy peasy and trouble-free?), the surname Bean was sometimes later rendered as McBane The final two children in this census listing (the “yellows”) are the offspring of the widower George Vallely and the widow Mary Moyle.
So the “greens” were the half-siblings of the “yellows” (they had the same father: George Vallely). And the “blues” were the half-siblings of the “yellows” (they had the same mother: Mary Moyle). But the “greens” and the “blues” were not blood relations at all: they were related only by marriage.
Well, “only by marriage.” The above-listed children of the the three marriages (George Valley to Anne O’Hanlon; John Bean to Mary Moyle; Mary Moyle to George Vallely) were all Roman Catholics, and according to Catholic canon law, they were all related by affinity. It was forbidden for Edward Vallely (son of George Vallely and Anne O’Hanlon), say, to marry Jane Bean (daughter of John Bean and Mary Moyle), say, even though they had no blood relation at all, because one of Edward Vallely’s parents (his father, George Vallely) had married one of Jane Bean’s parents (her mother, Mary Moyle).
In sum: I think it’s fair to assume that the Canadian census returns will not adequately capture the nuances of Roman Catholic canon law. Also: don’t assume that every child given the same surname in a census return is the offspring of the same marriage.