J. F. Tocher 
193 
fair-haired and dark-haired people living in sparsely populated regions. The 
chances of conjugal union of persons of the same colour class, if the mating occurs 
at random or is pangamic, are greater than if they lived all together as one group 
in a densely populated town. In the past, more unions between persons of the 
dark-haired class (for instance, in the west coast) were likely, on the assumption that 
mating occurred purely at random, to occur than between them as a class and the 
fair-haired class. Similarly, isolated groups of the fair-haired class would have more 
unions among themselves than with the smaller dark-haired groups. On the other 
hand, however, wherever towns sprang up, the different classes would be brought 
more in contact with one another and the chances of union among all classes with 
one another would be greater. But does mating actually occur purely at random ? 
That is to say, taking the character here considered, hair colour, does the fair-haired 
class, for instance, select mates indiscriminately from the other classes or do they 
tend to mate moi'e with members of their own class ? Similarly, taking eye colour, 
what is the nature of the mating ? Pearson* has shown that, for certain measur- 
able characters, like tends to mate with like ; that is, assortative or homogamic 
mating exists. For eye colour he has shown that both homogamic and preferential 
mating exist. Can one say with respect to hair colour whether the mating is 
homogamic, preferential or pangamic ? In the past, with isolated groups and with 
the clan system in vogue, endogatnic mating would certainly exist and be a power- 
ful factor in determining the prevailing colour characters. Thus one would expect 
at the present day to find a section of the population in the Highlands with 
characters distinctly different from another section, and this, one finds, is the case. 
Different race or clan groups have married within the race or clan and retained 
the ancestral characters. But endogamic mating can now no longer be a powerful 
factor, except in isolated cases, since greater intermixture and greater dispersal of 
the population now occur than was ever possible in the past. Retaining this form 
as possibly contributing, and remembering that mating of unlikes (conjugal union 
of say a member of the jet black class with a member of the fair-haired class) is also 
quite possible, the five possible forms emerge, namely : 
Homogamic = like with like ; 
Endogamic = members of the same clan ; 
Preferential = preference for a certain colom* ; 
Heterogamic = mating of unlikes ; and 
Pangamic = random. 
Now while it has been shown that inheritance of eye colour is more of the 
exclusive form than of the blended form, is it more likely that hair colour (except 
perhaps red hair which has been already noticed) is a case of blended rather than 
of exclusive inheritance ? As yet there are no statistics from which the intensity 
of blending can be directly proved or disproved. One can only advance the theory 
that blended inheritance prevails largely in hair colour, and see whether it explains 
the excess of medium hair in densely populated centres. Blended inheritance in 
* Pearson and Lee: Biometrika, Vol. ii. pp. 357 — 462; and pp. 481 — 498; and many others. 
Biometrika vi 25 
