22 Dr Duckworth, On the Anthropometric data collected 
with that of other groups. For this purpose, I have a number of 
data mostly prepared by myself from the original measurements. 
As the data for the cephalic index are most numerous, they will be 
considered first. For this purpose I have collected them as shewn 
in Table V, where the values of the standard deviation are 
arranged in their sequence of increase. 
The first conclusion to be drawn from Table V is that the 
Maldive and Minikoi men agree generally in presenting a higher 
degree of variability than the other groups brought into comparison 
with them. 
But the Minikoi men are not always thus associated with the 
Maldive islanders. For in two important characters, viz. the 
stature and the width of the head, the Minikoi men are dissociated 
from those of the Maldive group. Moreover the Minikoi men are 
in these two respects more homogeneous than their neighbours. 
Thirdly, this homogeneity as regards stature and cephalic 
breadth is in strong contrast with the great variability in respect 
of the nasal index shewn by the men of Minikoi. 
Kvidently the conclusions already formed as to the mixed 
character of these island populations find confirmation in Table V. 
8. Before passing from the strict consideration of such 
numerical data, it is convenient to notice the values of the 
coefficient of correlation for certain pairs of dimensions. They 
are shewn in Table VI. | 
TasBLe VI. 
Characters compared No. of individuals | ‘‘r” | p.n. of ‘‘r” 
Maldives and Minikoi together 
1. Cephalic length and breadth 68 484 + 0564 
2. Cephalic length: Cranial height 68 *244 + 074 
3. Cephalic breadth: Cranial height 68 227 += 075 
4. Cephalic index and length 68 — -520 + 053 
5. Cephalic index and breadth 68 “475 + 056 
6. Nasal length and width 69 528 +051 
7. Cephalic index: Nasal index 68 0104 + “068 
Minikot men only 
8. Nasal length and width 20 281 = oso 
The data presented in Table VI do not differ markedly from 
those based upon measurements of very different origin. I have 
few records available for comparison, but the values set out in 
Table VII are not without interest. Yet they do not seem to 
enhance the value of “r” as a discriminating agency. 
