1922.I P. C. Mahai«anobis : Analysis of Stature. 9 



I have intentionally made the present analysis very elaborate. 

 A total of only 200 observations did not perhaps merit such close 

 scrutiny. As there was no early prospect of increasing this total 

 considerably, I thought it better to complete even a provisional 

 investigation thoroughly rather than wait indefinitely for a larger 

 sample. But the chief reason which prompted me to make an 

 intensive study of the small available amount of material is this, 

 so far as I am aware no work in this line has been done in India, 

 no Anthropologist in India has ever made any use of the modern 

 statistical calculus associated specially with the name of Karl 

 Pearson and the Biometric School. The present study is intended 

 to illustrate the urgent necessity of the application of statistical 

 methods to Anthropology. The conclusions based on only 200 

 observations cannot of course claim any degree of finality. But 

 these serve to show the kind of results which can be reached 

 by statistical methods and also show the great scope and huge 

 possibilities of statistical methods. 



Remarks on the Application of Statistical Methods. 



Before proceeding to the more systematic part of the work I 

 wish to make a few general observations on the application of 

 Statistical methods. I cannot do better than begin by quoting 

 some remarks of Charles Goring in this connection. 1 



ff Statistical enquiry, all scientific enquiry, is observational in 

 character : that is to say, it is based upon the observation of in- 

 dividual facts. But these facts, in themselves, do not constitute 

 knowledge. Knowledge consists in the discovery of relation- 

 ships revealed by the systematic study, and by the legitimatised 

 weighing of facts." 



" No series of biological or social observations constitutes 

 knowledge in itself. Knowledge lies potential in the facts, but 

 ineffectual for use until their associations with each other have 

 been accurately weighed. It is the weighing of observations 

 which demands for the present enquiry, the employment of statis- 

 tical methods : such methods being merely a regulated mechanism 

 by which the relation between certain order of facts can be precise- 

 ly determined." 



u There is not, as is sometimes imagined, any special theory 

 or hypothesis involved in conclusions revealed by statistics. The 

 science of statistics provides only for the systematised study and 

 legitimatised interpretation of observed facts : such interpretation 

 consisting mainly in one and the same process — the associating or 

 dissociating one set of facts with and from another. Before any 

 association can be legitimately postulated, certain conditions must 

 be fulfilled; evidence must be produced to show that the relation, 

 affirmed to exist, is not a chance or accidental, but a natural asso- 



J Charles Goring, The English Convict, pp. 19-20 1 H.M.S.O. 1913) 



