APPENDIX I. NOTE ON STATISTICAL TERMS, 



In this appendix I have made an attempt to explain, in non- 

 mathematical language, some of the more frequently occurring 

 technical terms of statistical theory. Considerations of space have 

 prevented me from giving concrete illustrations. I hope however 

 that the following pages will serve some useful purpose in helping 

 anthropologists who lack the requisite mathematical training, in 

 taking an intelligent interest in the various technical discussions 

 contained in this paper. I have only attempted to give a general 

 idea of the different terms; the statistician will, 1 hope, forgive 

 me for the consequent lack of precision in many places. 



Let us consider our 200 measurements of Anglo-Indian stature. 

 Almost all individual measurements are different from one another. 

 The existence of variability is patent. The important fact is, 

 however, that this variability of stature is not chaotic in its dis- 

 tribution, but that it is governed by definite laws. 1 



We can classify our material into different groups in accord- 

 ance with size. We find, for example, that there are 2 individu- 

 als whose heights are less than 1465 mm. Between 1465 and 1485, 

 there is only one. Between I4S5 and 1505, there are 4, and so 

 on. Thus with a 20 mm. unit of grouping, we get the following 

 distribution of frequency in each group. (The number of indivi- 

 duals in any group is called the frequency of that group). 



Frequency Groups in units of 20 mm. 



Group 



1445 

 to 

 1465 



mm. 

 mm. 



1465 mm. 

 to 

 1485 



1485 

 to 



1505 



1505 

 to 



1525 



1 



1525 1545 



to to 



1545 1565 



1565 



to 



1585 



10 



1585 



to 

 1605 



1605 1625; 



to to" 



1625 1645 



Number 



2 



1 



1 4 



12 



2$ 32 



Group 



1645 



to 



1665 



21 



1665 



to 



1685 



17 



1685 1705 

 to to 



1705 1725 



1725 

 to 



1745 



1745 

 to 



1765 



F765 1785 



to to 



1785 1805 



1801; 



to" 



1825 







1825 



to 



1845 



I 



1845 



to Total 

 1865 



Number 



21-5 18-5 



10 



5 10 2 



1 200 



These frequency groups are shown graphically in Plate I. 



Let the horizontal A'- axis represent stature. Then, at the 

 middle point of each group, we ran erect vertical Lines propor- 

 tional to the frequency in that group. For example, at 1455, 

 which is the middlepoint of the group 1445-1465. we erect 

 a vertical line whose length is two units, to represent the 

 frequency in that group. At 147.5, the height of the vertical 

 line i-> one unit and so on. If the extremeties of these vertical 



I taring fhe Rngtish < onvict," p. _><>. 



