200 BOTANICAL GAZETTE [SEPTEMBER 
One further difference remains to be recorded in reference to the 
petals of the flowers of O. Lamarckiana, O. biennis, laeta, and velutina. 
As we look at these flowers on their respective plants, the smaller 
and even the more striking differences of the petals are apt to escape 
notice. ‘These can be seen most clearly by removing a flower of each 
of the forms just named and spreading them out carefully on white 
paper. To do this best and to fasten them to the paper, a coat of 
mucilage is first applied. This allows the petals to be arranged 
without injury, and at once discloses their true size, form, and 
relation to each other. 
In O. Lamarckiana the four large petals do not touch one another 
even at their nearest edges. Their narrowed bases leave wedge- 
shaped openings. The lobes of the petals are equal. 
In O. biennis the form of the petals is at once seen to be different, 
and they do not touch each other. The petals do not run down to 
the base as straight lines, but curve in so as to leave a nearly flask- 
shaped opening. Furthermore, the petals are slightly unequally lobed, 
but always in one direction. 
Laeia is very plainly different from velutina by the conspicuous 
and considerable overlapping of its four petals. These overlapped 
edges of the petals leave inclosed spaces at their base. These petals 
are also seen to be considerably larger than those of velutina. 
The petals of velutina resemble those of O. biennis in general 
form, but are somewhat unequally lobed and larger. The spaces 
between the petals of velutina are larger and more nearly closed by 
the broad part of the petal than in O. biennis. 
The cross-sections of the stamens mentioned in this paper were 
made as follows: Holes were made in cork with a small cork-borer, 
the anthers were placed carefully in them, and the holes were then 
filled with liquid glycerin jelly. After the glycerin had again become 
firm, the piece of cork with the specimens was placed in 96 per cent 
alcohol and allowed to dehydrate and further solidify the glycet™» 
which required two to three days. The sections were then ne 
through both cork and specimens free-hand with a razor, and mount 
for study in glycerin jelly. Such specimens, of course, can be made 
permanent by sealing the cover glass with Canada balsam. 
se Dae 
Pe Se ae, 
