Manchester Memoirs, Vol. It. (1907), No. 10- 



XVI. Some Tables for illustrating Statistical 

 Correlation. 



By A. D. Darbishire, 



Royal College of Science, London. 

 Received January 4th, igoy. Read February 12th, igoj. 



It was not my original intention to preface the 

 description of my new Tables with an account of 

 Weldon's experiment* But I was persuaded that without 

 such an account the meaning of my Tables would not be 

 evident to many. It must be understood, therefore, that 

 except in the matter of presentment the first part of this 

 paper makes no claim to originality. 



The second part contains an account of an interesting 

 extension of the experiment described in the first. 



I. 



Let us begin at the beginning, so far as we can. In 

 the case of a very great number of vital phenomena we 

 are unable to predict exactly what the result of certain 

 events will be. We know that they will fall within 

 certain limits, but where within those limits we cannot 

 tell. We believe that a duck will not produce a duckling 

 with a beak as narrow as a snipe's, but the exact breadth 

 of the beak — measured, let us say, in terms of its length — 

 in a given instance, we cannot foretell. 



If these words ever happen to lie before the eyes of a 

 biologist, the chances are that he will be inclined to ask me, 

 "What does it matter what the length-by-breadth index 

 of a duck's bill is ? " My answer to this interruption is 



* :06. Weldon, W. F. R. ''Inheritance in Animals and Plants," 

 pp. 81 — 109, in " Lectures on the Method of Science." Edited by T. B. 

 Strong. Clarendon Press, Oxford, 1906. 



June 28th, igoy. 



