Notes :
1. These tables give a good indication of magnitudes but are not reliable in detail.
The census material on which they are based was collected, deliberately, as opportunity
offered, and the results are incomplete and may contain a few inaccuracies.
2. The sample is haphazard, since I used any sources open to me to augment information,
including occasionally deduction. In Sakaltutan, I was able to work out that some
marriages were between close kin, and I am fairly sure that I missed no marriages
between agnates. Of the marriages on which I have no recorded information about the
kinship relationship of husband and wife, most, therefore, are likely to be between
non-kin, and very few, if any at all, between agnates. For this reason, I have given
the proportion of marriages between various types of kin both as percentages of all
recorded marriages (column a), and as percentages of all marriages on which I have
definite information (column b). For marriages, between agnates, column (a) is reliable;
for other marriages the truth lies between the two figures.
For the cases of girls marrying out of Sakaltutan, and for all marriages in Elbasl,
my information on the kinship network of the individuals was far less complete, and
hence the cases on which I happen to have recorded information are less likely to
represent a biased sample. I have not therefore given both sets of percentages in
these cases.
3. Figures for marriages of women of the two villages who married out are not typical
of figures for all marriages—they form a highly biased sample. To add them on to
the figures for internal marriages would be to bias the overall total. In particular,
the number of marriages with non-kin between villages is obviously likely to be higher
than within the village; and the marriages with agnates between villages are almost
non-existent. The class of marriages between spouses from different villages is already
fully represented by the cases where spouses have married in. The last column is
therefore only for overall totals and is otherwise left blank.
4. Since women normally attempt to use the full span of their fertility, even the
children of one mother may differ widely in age. Since elderly widowers or divorces
frequently re-marry, paternal half-siblings differ even more widely - up to forty
years. Hence marriages between cousins of different generations is not uncommon.
First cousins once removed I have counted as second cousins, and second cousins once
removed as other agnates or kin. The difference of generation may have no effect
on the quality of the relationship.
5. Affines include all cases where a link through an existing marriage was part of
the chain, from cases where people married brothers or sisters-in-law, to cases of
much more remote connection.
6. The category 'kin' includes both cases of known kinship more remote than those
specifically mentioned, and cases where I knew that a kinship tie existed but could
not make a reliable guess at its nature.