104 INHERITANCE IN ANIMALS 



throw gave only one successful die, the number of 

 successes in the preceding first throw was never more 

 than 6. 



You see, then, that first results, which deviate from the 

 average in either direction, are generally followed by 

 second results, which deviate from their average in the 

 same direction. I want to show you how we can measure 

 the relation between first and second throws a little more 

 closely; and for this purpose we will consider only the 

 mean values of the entries in every row and in every 

 column. It would be perfectly easy to consider the whole 

 series of numbers as they stand, and students of heredity 

 are obliged to consider quite similar series ; but to do so 

 would take too long. 



In the diagram fig. i, every cross shows you the mean 

 value of all the entries in a line of the table, every circle 

 shows you the mean value of all the entries in a column. 

 Every circle therefore shows you the average character 

 of those second throws which followed first throws of 

 a certain known character ; every cross shows you the 

 average character of those first throws which preceded 

 second throws of known character. You see that each 

 series, whether of circles or of crosses, lies fairly well upon 

 one of the two dotted lines in the diagram. Be good 

 enough to look first at the series of circles, and to begin 

 by looking at the middle of it ; that is, consider the 

 average results of those second throws which have 

 followed first throws yielding six dice with more than 

 three points, or first throws of average character. You 

 see the position of the circle in the middle column shows 

 that the second throws which followed these average first 

 throws were themselves on the whole of average character, 

 containing six successful dice. If you look at the cross 

 in line 6, you find in a similar way that the first results, 



