W. R. Macdonell 
213 
a ciitne being suspected, the police might from my formulai give a very fair 
estimate of the stature of the supposed victim for the purpose of identification. 
The form of such an estimate from the cubit, for my first individual, for example, 
would be 62" + 1". 
TABLE 28. 
Showing the difference in cms. between Actual and Calculated Height in 
10 Criminals selected at random. 
Actual 
Statdke Calculated 
FROM 
Average 
StSiturc cms. 
Finger 
Cubit 
Foot 
Finger and 
Cubit 
Finger and 
Foot 
Foot and 
Cubit 
160'3 
168- 3 
170-2 
158-4 
161-9 
165-7 
167-6 
160- 
161-9 
169- 5 
+ -3 
-1-4 
-2-6 
+ 5-4 
+ri 
+ -3 
+ 1-6 
+ 4-5 
+ 8-1 
-5- 
-2-4 
-2-3 
-4-2 
+ 1-3 
+ 3-1 
- -4 
+ -5 
+ 1- 
+ 6-2 
-1-1 
+ -6 
-6-2 
-3-3 
-2- 
+ 2-6 
+ 3-2 
+ 5-8 
+ 3-7 
- -6 
-4-2 
-2-4 
-2-3 
-4-3 
+ 1-2 
+ 3-3 
- -5 
+ -4 
+ -9 
+ 6- 
- -8 
- -2 
-5- 
-3- 
- -5 
+ 1-8 
+ 2-5 
+ 5- 
+ 3-7 
+ 1-9 
-4-6 
-2-1 
-3-8 
-3-9 
- -5 
+ 2-8 
+ -8 
+ 2-6 
+ 1-5 
+ 3-9 
-2- 
- 1-03 
-3-50 
-3-55 
+ -82 
+ 2-45 
+ -98 
+ 2-65 
+ 2-55 
+ 4-25 
-2-95 
Average 
3-03 
2-25 
3-22 
2-21 
2-82 
2-39 
2-47 
Prob. Error 
±3-3 
±2-6 
±2-9 
±2-6 
±2-9 
±2-5 
±2-8 
(22) Summary. I will conclude with a short summary of the results set forth 
in detail in this paper. 
(i) The great value of the new method of calculating standard deviations, 
means, and coefficients of correlation is demonstrated ; it economises labour, and 
leads to results closely approximating to those obtained by more laborious 
methods. 
(ii) Then as regards the criminals ; they are shown to be homogeneous, fairly 
normal in the distribution of the selected cliaracters, markedly different from the 
educated classes in stature, and in size and shape of the head, but in variability 
they agree generally with other classes and races of mankind. 
(iii) Finally as regards the problem of identification, it has been shown that 
certain characters at present in use are not very suitable for the purpose, owing to 
their high correlation, but assuming them to be employed, we have discovered the 
proper order in which they should be entered in an index cabinet — an order very 
different from that determined by Scotland Yard — and have indicated a method 
of calculating uncorrelated characters which would furnish, an ideal system of 
identification. 
