1908] DANFORTH—VARIATION IN DAISY 351 
sequently the probablity is that all German data are based on the 
typical Chrysanthemum Leucanthemum, while all American data are 
doubtless based on var. pinnatifidum. In view of this fact the value 
of comparisons between the two forms may at first seem doubtful, 
but a slight further consideration of the character of numerical 
Variations of the present kind may throw some light on the way 
m which such variations should be 
Tegarded. 
SHULL (5), working with Aster, con- — ae : a. 
cluded that there is no tendency for FORD BETWEEN May 27 AND 
all the flowers of the same plant to _ JUNE ™ 1995 
TABLE III 
fall in the same mode, but that those 
that blossom first have the — es 
greatest 
amount of nourishment and therefore ms I 
show the highest modes. TowER (6) - I 
Tikewise thought that in the daisy the 16 
igher modes are met with early in the 3 3 
season, the lower modes later. In se 5 
other words, it would seem that these i: 78 
authors are inclined not to regard the = 30 
Peal modes as indicative of incipient 24 - 
a My own observations, I think, a 38 
Y substantiate their views in this a iE 
lage A comparison of Tables ag 3 
and TV will show clearly a change = 79 
‘ the predominant mode from 34 in 32 Pe 
in collected between May 34 7 
oo. € 14, to 21 in material col- 36 20 
prin tween July 3 and July 15. 24 = 
hough this comparison may not be 39 3 
aken as certain evidence, inasmuch as a : 
= lots were from different local- : 
» Nevertheless there is a strong suggestion here of a connection 
Ween mode and season. 
"hig pairs of tables, one based on material collected at Norway, 
: oy the other on material collected at Dennis, Mass., ake much 
T evidence of the relation between mode and environment. 
