J. F. Tocher 
191 
likely to be fi'om Irish or English sources. The presence of neither foreigners, 
Irishmen, Englishmen, nor of brown-haired immigrants from rural districts at 
home (although they might contribute a little) can explain the excess of medium 
hair. None of these groups are likely to have contributed ; it has been proved, 
in short, that they do not. Having considered among others the effect of the 
presence of persons of a non-Scottish origin — the effect of a section of the popu- 
lation proved to be present whose origin is forth of Scotland — and shown it 
to be inappreciable or non-operative, one must conclude that the cause has an 
internal origin and is not derived from an external source. It must be some- 
thing operating within the Scottish population itself. What factor is operating 
within Scotland producing an excess in densely populated areas of the various 
shades of brown hair classed as medium ? 
One or more of at least three factors might possibly operate and provide the 
explanation. 
(A) Darkening among the fair-haired might occur earlier in towns and might 
he more intense. No grounds exist for this explanation. It is purely hypothetical 
and requires investigation. (B) The medium class might be the most fertile. 
Since this class is correlated with density of population, since the lower classes live 
in the densely populated areas, and since it has been shown that the lower classes 
are the most fei'tile, one might conclude that the medium class is the most fertile 
of the fertile lower classes. If true, this would explain the excess. (C) The excess 
might be due to the effect of blending of the fair and dark classes of the population. 
With regard to (A) until observations from towns and rural districts, bearing 
on this, are calculated, the truth or otherwise of the hypothesis cannot be verified. 
The pigmentation survey I'eturns contain no data capable of furnishing the means 
of testing this hypothesis. 
(B) The probability of the medium haired class being the most fertile. Com- 
paring the number of births per 100 families (calculated from the figures of 
the Census Report — the only data at present available to estimate the relative 
fertility in the various divisions of Scotland) with density of population, the value 
of the correlation coefficient was found to be 
r = -782 + -093 ; and ^ thus = 8-44; and = 3-08. 
J^r -C'(c=o) 
That is to say, births per family are greater in number in more densely populated 
areas than in sparsely populated parts*. Of course this does not give the measure 
of true fertility. To get this, one would require to get a return of the number of 
wives for each division, whose ages are within the childbearing range, and compare 
* On the other hand on comparing the number of families per 1000 of the population with density 
of population the correlation was found to be negative (?■= - -6109 ± -ligS). This does not necessarily 
mean that in towns the families are larger. The large population of young men and women employed 
in industries and otherwise and drawn from less densely populated areas contribute largely, if not 
mainly, to the result. 
