162 DIMORPHISM IN ERANTHEMUM. 
viscous; with the lenzthening of the corolla-tube the stigma ultimately 
rests between the anthers, and if unfertilized previously by insect or 
of fertilization. The intermediate sized flowers have the co 
to less than half the size of the preceding ; tube being 5 lines long i 
^ 
he limb 6 lines across. This form is occasionally self-fertile, and I 
find it fairly fertile by application of pollen from both the larger and 
smaller sized flowers. The smallest sized and closed lowers are per- 
fectly self-fertile. The corollas in this case are from 2 to 8 lines long 
and never open, but are naturally forced upwards and cast off by the in- 
and sterile flowers; smaller, but open and occasionally tertile flowers; 
and minute closed and perfectly self-fertile flowers. I may also remark 
that in both of the above species, the large and fully developed flowers, 
as a rule, are produced almost exclusively in the cold season. Lat 
Lo suspect that this will also explain the pent 
and comparative rarity of the larger-sized flowers as remark by s 
Kurz In indigenous habitats. Few, I believe none, of the species 9 
Eranthemum exhibit any floral diurnal movements, and even when ferti- 
lized they retain their beauty for two or three days. Anyhow, m the 
nothing whatever to do with the circumstance of his having “ only oe 
s T flowers." I attribute it largely to ie 
spheric conditions. In the cold season, as I have said, the large flow 
exclusively to the production of minute, closed, and self-fertile fonat 
ie r.—The large flowers have a s l de: 
PE: reading limb 12 lines broad. These flowers arè ©. 
2, with a spread h 
easionally naturally fertilized, as I have observed perfectly develo 
i 
