VARIABILITY OF A SINGLE CHARACTER 109 
number of other characters that can be measured, weighed, 
counted, or in any other way accurately determined. Thus the 
following is an actual distribution as to weight of ears, in which 
the character is measured in terms of ounces instead of inches. 
Here the problem deals with a different unit of value, so that 
lV’ now stands for ounces, while / stands for frequency, as be- 
fore. The mode in this distribution is 9 
ounces and the mean 8.807 ounces, the a Nis 
derivation of which is left for the student ix Ounces 
or reader by the methods already outlined. = — 
If now we should pick an ear that is ‘ a 
8.514 inches long and that weighs 8.807 2 3 
ounces, it would be fyprcal both as to 3 8 
length and weight. So in the same way - ie 
other characters could be determined, 4 ; 
such as circumference, number of rows, 7 28 
and in some cases even color, and any ear 8 2 
that was ‘on the type” as to each char- = 
acter could be fairly called a typical ear. i 35 
A typical individual is, therefore, one 12 22 
that is typical, or average, as to all the 13 14 
characters that are considered important. : ; 
Practical experience will show that there 16 ‘ 
are very few typical individuals in any = 280 
species, provided very many characters 
are considered. 
Variability or deviation from type. But the average gives us 
only partial information concerning the character we are studying. 
It gives us no indication of the spread or range of the distribution, 
as to how many of the population have deviated from the type, or 
how extensive was the deviation; that is, the average gives us no 
measure of variability, and it is such a measure that we now seek. 
Average deviation. Referring to the original distribution of 
length of ears, let us consider how much each group of ears 
deviates from the mean or average length, which is 8.514—. 
Mean = 8.807 ounces. 
