210 
Table of the Gaussian " Tail " Functions 
Without arguing this point out here, we may illustrate the use of the Table 
(p. 214) of ■\{r's by taking two of Parsons' frequency distributions for females; we 
will cut them off at the points suggested, and then investigate the total popu- 
lations of females which result. Our author pools for these distributions right 
and left bones. 
Taking the diameter of " head of femur " for the females, we have 
Diameter in mm. ... 
36 
37 
38 
39 
40 
41 
43 
43 
44 
Frequency ... 
1 
1 
3 
8 
14 
12 
18 
12 
These are exactly the bones the Dwight process gives as female. We find 
S== 2-8851, 
d =2-6159 (measured from 44-5). 
Hence f, = T^/d" = ■4<216. 
Whence by interpolation from the table 
A' = -782, 1/^2 = -864, ^/r3= 1-278, 
leading to o-=2-260, /i = l-767, 
and Mean = 42-73 mm., iV"=88-2. 
Parsons gives for R. femur. Mean = 43, 
L. femur. Mean =42, 
and the total number of bones dealt with 55 + 48 = 103 (Tables, loc. cit. pp. 249 — 
251). In his frequency distribution (p. 256) he only records 85 female bones, 
which give a mean of 42-54 and a standard deviation of 2 078 mm. These values 
are clearly not widely divergent from those we have found above by supposing 
all bones under 45 to be female. 
To test the matter further the 105* female bones of which the head was 
measured by Parsons were taken out. They provide the distribution : 
Diameter in mm. ... 
36 
37 
38 
S9 
40 
41 
42 
43 
44 
45 
46 
47 
48 
Frequency ... 
1 
1 
4 
10 
18 
16 
24 
13 
14 
3 
1 
These give Mean = 42-47 
s.D. = 1-996] 
* It is not possible to say whether he has omitted two queried measurements. He has not omitted 
bones he queries in breadth of lower articulation. 
