THE VERY HIGH AND VERY LOW INDICES. (247) 55 
TABLE b. MEAN HEADLENGTH AND HEADBREADTH OF PEOPLE WITH 
A DIFFERENT HEADINDEX (cf. tab. Xa and XIa p. 185). 







Coeffi- 
Mean + Stan- | cient of 
2 AG I: dard deviation. | Varia- 
tion. 
Length males 85 and more 120 | 18.71 + 5.8 31 
Breadth COR a 10716 15.96) 2.5.9, Mae 
Length IX Temales 85° x 168 1.17.76 + 5.52 | 31.1 
Breadth So, 5 168 | 15.25 + 5.17 | 33.9 
Length males HO RARES 87 | 20.11 + 6.67 | 33.16 
Breadth TO A id 87 | 14.76 + 5.77 | 39.16 
Length females | 75 ,, 3 44 | 19.01 + 4.82 | 25.35 
Breadth Here es 44 | 14.13 + 4.58 | 32.4 
Length ?) males all 933 | 19.34 + 6.58 | 34 
Breadth ?) À 933 | 15.4 + 5.60 | 36.36 
Length ?) | females ke 1426| 18.36 + 5.82 | 31.15 
Breadth 2) Ri 1426 14.8 + 4.85 | 32.77 



very large value. Both these extreme values make the standard devi- 
ation so large. Leaving them out in our calculation ®), then we find 
5 — 4.3, that is smaller than the standard deviation of the whole ma- 
terial. 
For the high index the top of the curve, as well for the length as for 
the breadth of women, is sharper than that of men (curves p. 53). 
When calculating the variability as a coefficient of variation (see 
table b), we find a somewhat other result (1920b). Then we find f.1. 
that in all cases the variability of the headbreadth is somewhat 
larger than that of thehead length and that the variability of the head- 
dimensions of heads with high indices is somewhat larger for women 
than for men, whereas for low indices and for average ones the con- 
trary is the case (table b.). 
1) In a 2nd calculation of a somewhat larger number I found N=174,M +0 = 
17.74 + 5.51. 
2) See 1920b, tab. 5—7. 
3) The minus variation is that of a child of 2 years having staid behind in growth 
and development (1920b, C 331). The plus variation is that ofa child of 1.5 year with 
an extremely broad head (1920b, E 3g). 
