5 
behind it and is connected with it by: means of a vascular 
bundle in common. In bi-alate Cannas this staminode has disap- 
peared altogether. Admitting that originally Canna (and other 
Marantaceae) possessed six pollen-bearing stamens, the first 
stage of transition must have been a radial splitting into a 
petaloid portion and a polliniferous one, the next: atrophy of 
‘the polliniferous portion and persistency of the petaloid, and 
the final period the abortion also of the three alae in Distémon. 
Only the pollen-stamen forms an exception in so far that its 
fertile portion persists whereas in the five other stamens this 
portion is the first to obliterate. In Fuchsia ') and probably in 
many other obdiplostemonic flowers a similar process takes 
place viz. the radial splitting up of the petals, thus producing 
an additional stamen. 
c) Various flowers show cohesion of the paired petals viz. in 
subsequent stages of perfection amounting to their complete 
coalescence. The flower seems to be dimerous then but is in 
fact psendo-dimerous. On p. 119 of his already cited book on 
the Zingiberaceae Mr. Va.rton states that this anomaly is quite 
common with a cultivated variety adding that in this case the 
lip is not notched but entire. The material sent to Holland 
did not admit of a thorough study of the said phenomenon. 
d) A number of ovaries which arrived disjointed from the 
flowers showed only two cells; most probably they have be- 
longed to the pseudo-dimerous flowers. 
Ill. Zingiber officinale Rose. 
Legit A. Orrens, April 1919, Buitenzorg. 
According to verbal statements by Dr. VaLeron this species 
(and perhaps others also?) has two forms of stems: foliaceous 
and floriferous. The latter are termed ,spikes” by this authority *). 
The peculiarity of the specimens sent to Europe consists of : 
spikes producing ordinary leaves together with flowers. 
1) Linnaean Society’s Journal-Botany, Vol. XXV. 
2) New notes, p. 128. 
