VARIATIONS OF BLUE-EYED GRAS.S. — BAKBOUR. 191 



In one thousand specimens I noticed the following varia- 

 ions : — 



60 specimens possessed only 6 leaves in the perianth. 

 20 " '• " 5 •' 



10 " " " 8 



The number of stamens corresponded to the number of 

 divisions of perianth present. 



The divisions of the perianth absent always belonged to 

 the inner segments of the perianth. 



As regards colour variation of perianth, first of all let me 

 point out that the perianth is, as a rule, blue, and the outer 

 leaves of it on the reverse sides are alw^ays lighter in colour 

 than the reverse sides of the inner ones, the former being very 

 light blue or almost white on the reverse side, while the latter 

 on the same side retain their dark blue colour. Now% in the 

 specimens examined, I found that the blue colour of the perianth 

 is replaced by a distinct violet-purple in one hundred instances, 

 and that this variation occurred not in fading flowers but in 

 those in fresh, full bloom. 



Another point I noticed, which is not a variation, but which 

 I think may be found useful for diagnostic purposes, is the 

 question of venation. Given only one segment of the perianth to 

 look at, the outer leaves of the perianth are always marked 

 by five distinct dark veins of blue, the inner only possesses 

 three such lines. 



The three outer segments of the perianth are always a 

 millimetre or two broader than the inner three leaves, and the 

 notches on the inner three leaves nearly always deeper than 

 those on the outer three. 



In about sixty specimens in a thousand you will find 

 the bristle absent in the outer segment and replaced by a 

 frino;e. 



