152 
ARCHIVOS DO MUSEU NACIONAL — VOL. XXyl 
flowers bears very little fruit. So much so, that though many trees pro- 
duce from a 100.000 to almost a million flowers they often yield less 
than 500 fruits. < 
In order to ascertain the cause of this and having no practical litte- 
rature at hand I made observations on as many trees as possible. Some 
of these were kept under constant observation and. others observed whene- 
ver opportunity offered. They included trees growing in different loca- 
lities, under varying.conditions, and belonging to several varieties. The 
inflorescences wereíchosen from different parts of the tree and in various 
stages of developméíit., 
Examinatiotí 1 was macro and microscopicaL Among technical pro¬ 
cesses employed may be referred the use of phenol as a clarifying agent 
for opaque tissues, with excellent results and the culture of pollen in 
capillary tubes, in accordance with the indications of A. Lutz. 
T|he inflorescences are composed of primary and secondary ramifi- 
cations up to the fifth degree. Their dimensions vary from 6,5 to 36,5 
or even 42 cm., with medans from 17 to 20 cm., and 29 to 34. They are 
conical or sub cylindric and either pale green or more or less deeply 
rose colored, according to the varieties. They produce from 100 to 3000 
flowers. 
Anatomic examination of the axis of the inflorescences reveals the 
presence of schizogenous resiniferous canais in the medulla and garlands 
of hemicylinders of bast. ! 
Schizogenous canais were also found in the peduncle, disk and other 
floral organs (petals and sepals). 
Various anomalies, all of which tend to the development of vegeta- 
tive instead of reproductive organs were found in these inflorescences. 
The stamina vary in number from 1 to 5; generally only one is well- 
developped, the others degenerate in varying degrees, from simple re- 
duction in size with pollen formation, through intermediate stages with 
formation of two thecae devoid of pollen, to a more or less club shaped 
undifferentiated mass at the edge of the disk. 
The pollen, which, as known, is very small, sub-spherical, slightly tri¬ 
angular in outline and has three pores, is very scare in most of the sta¬ 
mina . | 
Pollen cultures were made on agar in Petri dishes and later in 
capillary tubes, cut to the size of a microscope slide, shut at one or both 
extremities. 
The media employed were water with 5; 10; 12, 5; 15; 20; 25: 
30; 33 and 40 % saccharose, or of glycose or the liquid resulting 
from the trituration of flowers and of styles and stigmata, also slices cí 
fruits. 
We were unable to obtain artificial germination and production :: 
pollen tubes and found only one or two spontaneously germinatinr :n 
the stigmata, 
j i ■ 
