75 
in which two sets of sepals of two each are found; the lower set 
are then half an inch or less distant from the flower, and do not 
decussate, as might normally be expected. They form an angle 
of 45 degrees, or even much less, with the upper set. 
Eichler, in his Bliithendiagramme, figures the plan of the 
flowers as though there were two whorls of petals of two petals 
each. By dedoublement, he derives from these two whorls of ap- 
parently four petals each. This compels him to locate the sepals 
between the petals of the outer whorl. Observation, however, 
will show that the sepals are not thus situated, but are beneath 
two of the petals of the outer whorl. If it be permitted to sug- 
gest another plan or arrangement, the following may, perhaps, be 
nearer the truth: The two sepals may be considered as decus- 
sating with two of the petals of the outer whorl; these decussate 
in turn with the remaining two petals of apparently the same 
whorl. The two sets of petals considered as a whole then decus- 
sate with the inner whorl of petals. This rather anomalous 
method of arrangement is considered in good taste in other 
plants, and perhaps will be of equal value here. 
(Qc) (vee) (gay) 
(So) «((80)) 
Ml. 
EXPLANATION OF FIGURES. 
All figures are represented as having four scales at the base 
of the plant; these with the bracts above are indicated in the 
diagrams by double crescentic lines—the leaves by single lines ; 
leaves and bracts of the primary stem in the plate are designated 
consecutively by Roman numerals, those of the secondary stem 
by Arabic numerals. The stems are represented by circles, in 
the diagrams decreasing in size, in the plate lettered consecu- 
tively. The buds giving rise to the branch of next year are 
