PROFESSOR. K. PEARSOX OX TflE IXFLUEXCE OF XATTTRAL 
3 G 
d ? 
Mean stature, mi . 
Variation of stature, o-j. 
Mean femur, m -2 . 
Variation of femur, (r., . 
Mean humerus m 3 . 
Variation of humerus, 0-3 . 
Correlation, femur and humerus, r .23 . . 
,, stature and humerus, Vis 
,, stature and femur, ri 2 . . 
166-26 cms. 
5 - 50 ,. 
45-23 „ 
2-37 „ 
33-01 „ 
1-54 „ 
-842 
-809 
•811 
154-02 cms. 
5-45 „ ; 
41-57 „ 
2-26 „ 
29-77 „ 
1-53 „ 
•872 
•771 
•805 
Now let us select from among French males a group having the same variability, 
correlation, and mean size of humerus and femur as French females, and let us ask 
how this would alter the variability ( 5 ^^), mean size (M^) of stature in French 
males, and also the correlation between stature and humerus (r^g) and stature and 
femur (1*13). 
We have at once from the second column— 
whence we find. 
—3-C)G. 3-24, 
i'. = 2-26, . 53 = 1 - 53 , 
Piz — ’ 8 / 2 , 
O2 
1 
’’2.3’’13 
•447, 
'23 
^’]3 ~ '^*2302 
1 - ^ 23 = 
= -433. 
From (Ixv.) we deduce, 
Mi= 166-2G + 1-037//2 + r546/,'3. 
This formula gives the stature of any group of males selected from the French, 
and having their femur and humeius respectively and centims. longer than the 
average. 
For the special selection referred to above, = — 3‘GG and Ag = — 3'24, hence 
Ml = 1GG-2G — 3-79 — S'Ol = 157'4G. 
This exam23le shows us that if we selected French men with the same femur and 
humerus as French women, it would l)e the selection of the humerus which would 
contribute mostly to the reduction of stature—a somewhat singular result. Further, 
such a selected group of Frencli men would l)e still some 3 ^ centims. taller than the 
average of French women (instead of about 12^ centims.). Probably had we 
selected the tibia as well, the greater portion of this remaining advantage in height 
would have disa})peai’ed. 
