F. H. WHITEHEAD 



211 



The floras, however, state that the bracts of C. atrovirens dire always completely 

 herbaceous. This was found to be incorrect. When the bracts were examined under 

 the microscope, in every case there was at least one row of colourless cells at the tip 

 of the bract and in some cases there was a scarious tip almost as large as the usual C. 

 pumilum tip. In a few cases the bract tips of C. atrovirens are more scarious than the 

 least scarious-tipped C. pumilum. This character, as with the previous characters, is 

 not an absolute criterion but it does more nearly approach absoluteness than any of the 

 other characters. It was found that the means for groups A, B and C were all significantly 

 different as may be seen from the data below. 



Group 

 AandB 

 BandC 

 AandC 



Standard difference 

 between means 

 2-0 

 1-7 

 1-9 



Actual difference 

 between means 

 27-80 

 3-62 

 3142 



25- 

 20- 

 15- 

 10-- 

 5-- 



Group 1 



EZt 



as-- 



ZO" 



UJ 



or 



lu JO 

 a 



Group 



25-. 

 20-- 

 15 

 10 



Group 3 



900 



lOOO 



llOO I200 

 SEED SIZE 



(300 



