The Tree of Life 69 



in assuming that species are fixed and perfectly isolated from 

 each other. 



Measuring Resemblances 



When we try to picture the relationship of a particular 

 individual to his uncles and his cousins and his aunts, we de- 

 velop a figure which is sometimes spoken of as a " family 

 tree." If we examine the actual individuals which such a 

 tree represents, we find a certain connection between the de- 

 grees of relationship and the degrees of resemblance. It is 

 a familiar fact that twins resemble each other more than do 

 ordinary brothers and sisters, and that brothers and .sisters 

 resemble one another more than do cousins. Careful meas- 

 urements have been made of thousands of boys and girls, 

 men and women, with respect to large numbers of facts like 

 stature, pigmentation, proportions of the head, the limbs, 

 the nose, and so on. Through the comparison of many such 

 measurements it has been possible to find some scale for 

 the degree of resemblance. The mathematical calculations 

 by which such resemblances are measured are rather compli- 

 cated, but for our purpose it is enough to consider degrees 

 of resemblance measured on a percentage scale. On such a 

 scale the resemblance between identical twins would be from 

 90 to 93 per cent. Brothers and sisters would resemble each 

 other 50 per cent. Cousins would resemble each other ap- 

 proximately 20 per cent. 



The Tree of Life 



While we cannot measure in the same way minor de- 

 grees of resemblance, we can apply the same principles to 

 our classifications of plants and animals. Whenever we at- 

 tempt to arrange the various kinds of plants and animals 

 according to the degrees of resemblance which they show, 

 we invariably produce a figure suggesting roughly a " tree " 

 (Figs. 13 and 14) . In such a tree the types that most closely 

 resemble one another are represented by twigs placed close 



