Alice Lee 
287 
(6) Find the percentage of cases in which a humerus of under 320 mm. will 
be associated with a femur of over 460 mm. 
Here h = 7-72/23-72 = -32546 ; A; = - 10- 10/15-38 = - -65670. 
Accordingly our system is 
h = -32546, k = -65670, r = - -8421. 
Our Tables give for r = — -80 
I Z .g} f'/^ = -009,0146 ; ^ 2 'jj '^1^ = -006,2334. 
Thus, for /; = -3, k = -65670, d/N - -007,4377. 
Agam h = -4J ^^^^ ^ -006,3352 ; J, ^ '^j dIN = -004,3066. 
Thus for h = -4, k = -65670, djN = -005,1850. Accordingly for 
h = -32546, k = -65670, 
we find d/N = -006,8642. 
We must now repeat this work for r = — -85. 
I I !o| (^1^ = -004,6616 ; J, Z ij} ^^/^ = -002,9477. 
Thus for h = -3, k = -65670, f?/iV = -003,6898, 
1 1 !g| f'/^ = -002,9950 ; J' Z iy} ^^/^ = -001,8483. 
Thus for h = -4, A; = -65670, d/N = -002,3448. Accordingly for 
h = -32546, A; = -65670, 
we find d/N = -003,3474. 
We have accordingly for the given values of h and k : 
r = - -80, r = - -85, 
fZ/A^ = -006,8642, rl/iV = -003,3474, 
or, for r = - -8421, f//A = -003,9006. 
Accordingly the occurrence of individuals with femur and humerus within the 
limits given is about 0-4 per cent, of the male French population. 
Many other illustrations will occur to the biometric reader of cases wherein 
these additional tables are likely to be of service. I have to record my very hearty 
thanks to Dr W. F. Sheppard for the loan of his manuscript tables of the probability 
integral which enabled me to work to a larger number of figures than his published 
tables, which go to fewer figures, would have permitted. 
