^88 



THE GRAMMAR OF SCIENCE 



of the Chiltern Hills. Compare them with another group 

 collected from several fields at the very top of the 

 Chilterns, or with a third series from the Quantocks, and 

 we see at once how local races may differ not only in 

 type, but in variability. Or, again, compare them with 

 two series of the Shirley poppy, a selection from a wild 

 Indian poppy, and we notice how largely the type can 

 differ as we pass even to a closely allied group. 



Poppies, Stigmatic Bands 



We have here illustration of how race, environment, 

 selection correspond to numerical differences in type and 

 variability, differences which can be used for purposes of 

 exact comparison and description. 



The reader who would gain some idea of such measures 

 of type and variability, and the influence upon them of 

 environment, has only to count in some hundreds of in- 

 dividual cases the petals or florets of some simple wild 

 flower, and observe the differences which occur when the 

 group is taken from a field with one soil, aspect or altitude, 

 and then from a second wherein all these things are 

 changed. The arithmetic is quite easy, and the results 

 often surprising for the light they throw, even on the 

 differences between early and late flowering members of 

 the same race. 



So far we have dealt with characters, the intensity of 

 which can be represented by discrete numbers. But if a 

 character, like the length of a bone or the size of an organ, 

 be under consideration, we can proceed in an almost identical 

 fashion. Our measuring instruments are not so fine that 

 we obtain actually " exact " lengths. We always measure 



