mixed race than the intermediate measurements. It is true that the number of observed cases may seem rather small to draw this deduction with absolute certainty; but I have noticed that all tabulations of face and head measurements which include more than
five hundred individuals give very regular curves except in the case of half bloods, so that I believe I am justified in interpreting the phenomenon illustrated in Fig. 4 as a real one, and that it is not due to the small number of measurements. The correctness of this view can be proved definitely by an appropriate grouping of
the available material according to the following point of view: The breadth of face and the breadth of head of man are closely correlated. The broader the head, the broader the face. Irregularities in the distribution of the measurement of the face will.