50 GENETICS 



above are mixed together, they form a regular 

 series from the inspection of which no one could infer 

 their double origin. (See the heavy line in Figure 27.) 



c. Skew Polygons 



The direction in which variations are tending 

 may sometimes be determined by the statistical 

 method. As an illustration of this may be cited the 

 number of ray florets on 1000 white daisies (Chrys- 

 anthemum leucanthemum) , 500 of which were col- 

 lected at random by the writer from a small patch 

 in a swampy meadow in northern Vermont, while 

 the other 500 were selected in the same random 

 manner upon the same day from a dry hillside pas- 

 ture hardly more than a stone's throw distant. 

 Among these two lots of daisies the number of ray 

 florets varies from twelve to thirty-eight and their 

 frequency polygons, as shown in Figure 28, form 

 what are termed "skew polygons," because the mode 

 in each case lies considerably to one side of the arith- 

 metical mean. 



It will be seen that lot A from the swampy meadow, 

 which in spite of the greater fertility of the soil and 

 tKe unquestionably greater luxuriance of the plants 

 themselves, produced heads with fewer florets, 

 fluctuates around the number 21, while the dry 

 pasture population J5, characterized by blossoms 

 which were in general noticeably smaller, fluctuates 

 around the number 34. 



