108 



MISSOURI BOTANICAL GARDEN". 



tically all the flowers had opened or were nearly ready for 

 anthesis. All those with young buds were discarded to avoid 

 the possibility of including stalks which had not completed 

 their growth. These were very few in number and the 

 samples taken may be regarded as representative of the 

 species in this habitat. The rather coarse scale of centi- 

 meter units was adopted partly because of the possible error 

 in the collecting of the material referred to above and partly 

 to obviate the apparently needless labor of dealing with 

 much larger correlation tables. 



TABLE A. N0TH08C0HDUM. 



Organ and Source 

 of Material 



Average and 

 Probable Error 



Standard Devi- 

 ation and Prob- 

 able Error 



Coefficient of 

 Variation 



Length of Stalk 

 Southwest Slope 

 Southeast Slope 

 Difference 



Flowers per Inflor- 

 escence 

 Southwest Slope 

 Southeast Slope 

 Difference 



19. 407 ±.162 



17.970 ±.115 



+ 1.437 ±.198 



6.212 ±.056 



5. 423 ±.039 



+ .789 ±.068 



4.545±.114 



3.511 + .081 



+ 1.034 ±.140 



1.584 + .040 



1.217±.028 



+.367 ±.049 



23.421 

 19.54J_ 



+ 3r880 



25.505 



22.435 



+ 3.070 



VARIATION CONSTANTS. 



The means, standard deviations and coefficients of vari- 

 ation, and the probable errors of the first two constants, are 

 given for length of stalk and number of flowers per inflor- 

 escence in Table A. Sheppard's correction for the second 

 moment was used for all the variabilities given in this paper. 



These constants require little discussion in this place. The 

 means, standard deviations and coefficients of variation are 

 all somewhat higher, and as the probable errors of the differ- 

 ences show significantly higher, in the series from the south- 

 west slope than in the other of the two lots . The empirical 

 range is also wider, although N is smaller, in the collection 

 from the southwest slope, where I had estimated that edaphic 

 conditions were more variable than in that from the south- 

 east slope where they had appeared to me more uniform. 



