94 
The drawing by the native draughtsman Darmosoxpiro is, 
however meritorious it be, misleading for a layman as it seems 
to represent one flower, with supernumerary perianth-leaves 
and stamens. 
In reality we have to do with an inflorescence like that of Iris. 
When examining the flower we find that the perianth is repres- 
ented by 3 + 3 but that there are e.g. 5 stamens and 5 styles. 
This number is caused by two of the three having collaterally 
split up, both stamens and pistils. Of the cells of the ovary 
only one was found to have gone through the process. 
For completeness’ sake we communicate also of two inflores- 
cences full details: 
I. a. immature fruit: 
number of cells 3, 4, 2+ 1 very dilated, 4, 3, 
. open flower: 
sepals 3, pet. 4, stam. 7, branches of style 7, ovary-cells 5, 
c. buds: 
$°, Sepaig: 3,5 4 <3, : poe & ‘ 3, 
rt ee Ry ee oe eee ‘ ersten Oi indistinct 
3°: ” 3, ” 3, ” 4, ” ” ” 6, ” 3 
Another specimen showed: 
Il. a. immature fruit: 
1 dissepiment split up, conseq. producing a 4th cell, 
. flower withered: 
sepals 3, petals 3, stamens 4, styles 4, ovary-cells 4, 
c. flowerbuds: | 
1°: sepals 3, petals 3, stamens 4, styles 4, ovary-cells 3, 
Popo y iB, svat Lal iS eceren ten oft cme 
3°: ” 3, ” 3, ” 3, ” 3, ue 3. 
The two latter cases of II seem to show that the fission 
occurs in a rather advanced stage of development. 
From the above facts it appears that (only with exception 
of the petals in 1) it is the sexual organs wich tend to colla- 
teral fission which in one of the cases is so sharply accentuated 
as to produce 7 stamens and 7 styles in the same flower (1°) ’)- 
oO 
a 
1) Cfr. MassaLoneo on Iris (*Penzie III, p. 329), 
