182 



NOTE ON THE ALPINE DWARFING OF POLYGONUM 



BISTORTA 



By J. Arthur Harris 



In evolutionary writings one frequently meets statements 

 concerning increased variability of a species subjected to new 

 conditions.* In connection with some work on this question, 

 it seemed worth while to consider the variability in individual 

 habitats of a species ranging from lowland to alpine conditions. 

 The familiar P. Bistorta L. of the Rocky Mountain region 

 appeared to be a good subject. In August, 1906, I collected 

 seven small series for a preliminary investigation, in the Pikes 

 Peak and Mount Garfield region of Colorado. The localities 

 need not be described in detail. 



The series are numbered in the order of (as it seemed to me) 

 increasingly alpine conditions.! The diagram shows the length 

 of the flowering stalks. The crosses on the bars show the mean 

 value and length of the bars the range in variation for the several 

 series. 



Expressing the results in the convenient biometric constants, 

 we note from the table that as the conditions become more alpine 

 the average length and the standard deviation, measuring the 

 absolute variability around the mean length, greatly decrease. 

 But when we express variability in relative terms by taking 

 the ratio of the absolute variability to the mean, i.e., Standard 

 Deviation X 100 / Mean = coefficient of variation, we note 

 that the variability is about the same from habitat to habitat. 

 Considering fluctuations due to sampling it is not safe to say 

 that the coefficients of variation differ at all among themselves. 

 But it is noticeable that the coefficients of variation are lowest 

 for lots I and VII which have the maximum and the minimum 

 development of stalk length. These are also unquestionably 



* For a review of some of the pertinent literature see Harris, Amer. Nat. 43 : 

 350-355. 1909- 



t Elevation is the only point upon which quantitative information could be 

 secured. By Alpine I have understood more the ecological condition than the alti- 

 tude merely. In making my estimates I was guided by the general physical sur- 

 roundings and by the appearance of the vegetation. 



