-------------------------------------------------------
| | | | Mainland |Insular
Period. |Principal|Large |Small | rural | rural
| towns. |towns.|towns.|districts.|districts.
-------------------------------------------------------
1855-1861| -- | -- | -- | 105.6 | 106.6
1862-1871| -- | -- | -- | 105.9 | 105.6
1872-1881| 105.0 | 105.6| 106.1| 105.3 | 108.0
1882-1891| 105.1 | 105.6| 105.5| 105.5 | 108.7
1892-1901| 104.7 | 104.6| 104.9| 105.2 | 107.1
-------------------------------------------------------
Average | 104.9 | 105.3| 105.5| 105.5 | 107.2
-------------------------------------------------------

[Footnote 38: Massachusetts _Census_, 103.1; Reg. 1891-1900, 105.6.
Vermont _Census_, 108.1; Reg. 1890-1896, 105.9. Connecticut _Census_,
103.9; Reg. 1887-1891, 107.2. Rhode Island _Census_, 103.8; Reg.
1854-1901, 104.9.]

[Footnote 39: Lewis and Lewis, op. cit., p. 128.]

This would seem to bear out the theory that masculinity is affected by
consanguineous marriage, for consanguineous marriage is more frequent
in rural districts, and especially in insular rural districts. But
unless consanguineous marriages can directly be shown to produce an
excess of male births greater than the normal, such indirect evidence
is valueless.

In the genealogical material previously considered, we have a sampling
of the American population throughout its whole history, but the data