428 BOTANICAL GAZETTE [JUNE 
material. The presence or absence of pollen, used by several 
candinavian authors as a specific character, is considered by the 
authority just quoted to be highly variable within the species, 
unless the observations of other men were erroneous. 
ACHENES.—These are generally regarded as affording the most 
critical taxonomic characters, that is, in size, form, color, beak 
length, pappus color, ribbing, and tubercles. Actually size is no 
more exempt from the laws of fluctuation in Taraxacum than in 
any plant, while form is subject to considerable modifications by 
pressure during development. HANDEL-MAzetTt1, however, states 
that uniform culture conditions have less effect upon achene form 
than upon leaf and bract characters. Colors vary from dark brown 
to clear green in T. officinale according to SCHORBATOW, who states 
that the two extremes are fixed in inheritance, but whether he 
includes 7. laevigatwm under the other species is not clear. In 
Stork’s cultures, chosen from parents varying as widely as possible 
in achene as in other characters, the progeny were either red or 
gray fruited. Color does not develop until the outer cells die, and 
in red fruits is due to homogentisic acid which is absent in gray 
ones (10). Freezing may prevent its appearance in T. laevigatum 
(17). Beak length is notoriously interfered with by injury, whether 
by freezing or mowing. Pappus color is said by Hanpet-MazetTIr 
to be due to diffraction phenomena and not pigment. The small 
species group characterized by definite pappus color is therefore 
likely to be a distinct one. Ribbing is due to the number of 
stereome bundles beneath the epidermis. 
Summarizing, it is seen that while bracts and achenes are 
conceded to afford the critical characters, the leaves are actually 
given great weight in spite of their variability and the uncertainty 
concerning its nature. There is also not a little evidence for 
bract variability. Pappus color seems dependable, and achene 
color ought to be generally so, barring environmental extremes. 
Observations 
MeETHOD.—Several years of desultory observation in connection 
with other phases of this work having failed to throw the necessary 
light upon the problem of leaf variation, it seemed best to conduct 
