THE AMERICAN NATURALIST [Vol. XL V 



but an accumulation of plain, unadorned facts available to any 

 one's inspection, it seems useless to try to bolster either of them 

 up by the dialectic methods of a lawyer's appeal to the jury. 



Raymond Pearl. 



June 21, 1911. 



OX THE FORMATION OF CORRELATION AND CON- 

 TINGENCY TABLES WIIEX THE X EMBER OF 

 COMBINATIONS IS LARGE 



ful tool, the correlation eoeflieienl. haw appeared. The first 1 ex- 

 plains and illustrates a convenient method of carrying out the 

 arithmetical routine of calculation, while the second, by Pro- 

 fessor Jennings, 2 describes a method for obtaining the coefficient 



the tables themselves symmetrical. 



The purpose of this note is to point out a method of prepar- 

 ing correlation tables where the number of combinations is 

 large. Such tables are not infrequently needed. Suppose, for 



