12 THE AMERICAN MIDLAND NATURALIST 
and that the latter proportionately diminish in number and 
often disappear entirely not only when cultivated but also wild, 
it is probably worth while considering them at least in part as 
real petals. The flowers of the plant are then to be considered 
as really rather asepalous than apetalous. Such double] flowers 
have been found as intimated even in their native haunts in 
perfectly wild conditions. 
Color variations are even more common. A hillside of plants 
of H. acuta was seen during the last season where the flowers 
ranged from perfectly pure white through lavender, pink, rose, 
purple, all with or without darker margins to light blue dark 
blue and even the darkest violet, all the plants being indiserim- 
inately scattered over the hillside facing the sun. I could only 
account for this wonderful show of color variation from the fact 
that the previous fall the area had been burnt over. A :imilar 
condition seemed to cause all or nearly all of the plants of Viola 
popultfolia, our common blue violet, to have blotched and streaked 
petals on their flowers. These violets were found in this condition 
only where the leaves were burnt away the year before. Some 
of these were transferred to a garden and bloomed the second 
season with similarly blotched flowers, blue and white streaked. 
A plant of Hepatica was found in which all the bracts were 
very close to the flower parts, in fact almost on the torus of the 
flower. These bracts were distinctly three-lobed and seemed to 
approximate the shape of the true hepatica leaves in being very 
broad, whereas the ordinary bracts are usually oval and quite 
entire or ovate. In the flowers of this plant the outer ‘‘sepals”’ 
colored in the typical plant were here green and herbaceous on 
the margins. This would seem to indicate a tendency on the 
part of the ‘“‘sepals’”’ to become bractlike or become real green 
sepals. As there are in most plants of Hepatica two more or less 
complete whorls of the so-called ‘“‘sepals’’ it were perhaps not 
incongruous to consider that these plants have both real sepals 
and real petals in more or less complete whorls usually alike, 
butin such cases atavistically tending to differentiate themselves, 
the outer gradually changing themselves under normal con- 
ditions into bracts gradually, and the inner similarly showing 
by the change of stamens into them that there is no break in the 
number of floral envelopes. 
Another anomalous condition of inflorescence in this plant 
