years should be taller than the half-blood child, and then develop more slowly than the latter. This peculiarity is most striking in the growth of the tribes of medium stature, as in this case the difference in the statures of adults is so considerable. Unfortunately, we do not know if the same difference prevails at the time of birth; but even if this were the case the difference in the rate of growth would remain mysterious. The various phenomena described here merely emphasize the fact that the effect of intermixture is a most complicated one, and that it acts upon physiological and anatomical qualities alike. We observe in the mixed race that the fertility and the laws of growth are affected, that the variability of the race is increased, and that the resultant stature of the mixed race exceeds that of both parents.
One of the most striking characteristics of the Indian face is its great breadth as compared with that of the whites. It is therefore of peculiar interest to compare this measurement among the full-blood Indian, the half bloods, and the whites. The curves on
Fig. 5. Breadth of Face. Sioux.
Fig. 4 show the result of this inquiry. Among adult students of American colleges we find an average breadth of face (between the zygomatic arches) of 140 millimetres, while the average value among Indians is nearly 150 millimetres. The facial measurements of the half bloods are intermediate, the average value being nearer the typical Indian measurement and remote from the white measurement. We find in these curves also the peculiarity observed before—that the half blood is more variable than the pure race. This fact is expressed in the greater flatness of the curve.
It will be noticed that the central portion of the curve illustrating the distribution of the measurements of breadth of face of half bloods is markedly irregular, particularly that it shows a depression in its central portion. This might seem accidental, but it will be seen that in Figs. 5 and 6, where the same measurements for the Sioux and Ojibwas are given, the same phenomenon appears. We see in all these curves that the measurements which are near those of the parental races appear more frequently in the
