64 
the place between the stamen and ala e@. If by way of excep- 
tion there appears a fourth ala (0), we find it between the 
stamen and ala . Eicurer considers all these alae as products 
of the stamen, as a sort of stipules belonging to it but grown 
out so as to constitute independent parts of the flowers (,Sti- 
pulargebilde’’). The stamen consists of two different halves, 
one (the right) being petaloid, the other one (left) being a one- 
celled anther. 7 
The expressions left and right refer to the structure of the 
flower; if the observer is placed in the axis, he has the parts 
of the flowers placed in a spiral line running upwards from 
left to right. 
Ercuuer’s diagram symbolises very distinctly his opinion about 
the morphological significance of alae and labellum: he looks 
on the outer whorl of stamens as quite suppressed, hence the 
three asterisks in the outer circle. Of the inner whorl he finds 
only one stamen s¢ and the labellum /ad., but there is no trace 
of a third stamen. As we have said before, the alae are according 
to his opinion produced by st. 
In stead of registering all the observed cases separately, which 
would render the perusal of the present paper needlessly fati- 
guing, we prefer to pass in rapid review the different parts 
and their deviations. 
Calyx. Sepals 4, in another flower 5. In a third flower there 
was a sepal grown together with a petal to a faleated whole. 
Corolla. Two of the petals are (serially ')) doubled, accordingly 
there are 5 in all. 
A second flower shows the same thing. 
A third flower shows 4 petals. 
Some petals assume a completely petaloid character widely 
differing from the rather tough tapering leaflets which in normal 
cases constitute the corolla. 
1) E1cuize distinguishes two kinds of doubling, serial and collateral, 1. c. p. 5. 
In the above case the doubling is undubitably serial, in the following the difference 
between the two could not always be ascertained. 
