go THE GROWTH OF GROUPS 



among Indian cottons found that the size and colour of the 

 petals were correlated. The petals were either large or 

 small, they were either yellow or white. The large 

 petals were yellow, the small ones were white. In over 

 a hundred thousand plants there was no exception to 

 this rule. A small yellow or a large white flower could 

 not be found. 



But, as Davenport points out, this correlation of char- 

 acters is exceptional, whereas we see the association of 

 uncorrelated characters in every part of the animal and 

 vegetable kingdom. But how can we explain this 

 association ? The case afforded by these two sports is 

 very simple. The occurrence of the four characters in 

 both must have been due to their community of descent 

 from an individual, and this explanation may be extended 

 to larger groups which are marked by a number of asso- 

 ciated but uncorrelated characters and to groups such 

 as are often called species. 



It may be said that this is building on slender founda- 

 tions ; but the key to complex problems is often found in 

 simple instances. 



There seems to be an important difference between 

 the occurrence of many-charactered groups, such as the 

 one described in this chapter, and single-charactered 

 groups such as were mentioned in the last chapter. The 

 former seem to arise on one occasion only, whereas the 

 latter arise on many occasions. For example, if other 

 individuals, having exactly the same four characters as 

 the two sports under discussion, were to be found in the 

 Amritzar district, we might be sure that they were 

 derivatives of the same family group as the others, but 

 we can never be sure of the place of origin of a sport 



