aoe ce ee So aa 
ee ii bareiaiicis a 2S ics er ee ok ek i eM og a iy 5k = Tes aa 
1913] BURLINGAME—ARAUCARIA BRASILIENSIS 107 
nucleolus and rather scanty chromatin. From this is organized a 
rather long, loose spireme, apparently containing scanty chromatin. 
It segments to form 8 chromosomes, which contract into rather 
small dense oval masses (plate fig. 18). 
Male gametophyte 
The results of this first division are a primary prothallial cell 
and a free nucleus (plate fig. 19). The latter at once divides to 
produce a second prothallial cell, which takes its place over the 
preceding. The next division yields the free tube nucleus and the 
primary spermatogenous cell (plate fig. 21). Meantime the pro- 
thallial cells have divided (plate figs. 20, 21, 22). The primary 
spermatogenous cell now divides, yielding the usual stalk and 
body cell. The stalk cell is very evanescent, usually becoming 
confused with the general cytoplasm very quickly, Plate fig. 25 
shows one of the few preparations in which it could still be distin- 
guished as a distinct cell. Another interesting case was observed 
in which there seemed to be two free and equal cells in the unshed 
pollen grain. The usual condition of a mature pollen grain, with 
15-25 free prothallial nuclei, a recognizable tube nucleus, and a 
single body cell, is shown in plate fig. 27. 
It is in this condition that the pollen is shed in April (usually). 
The potential number of pollen grains is four to eight times as great 
as the actual number formed. It has already been remarked that 
the ordinary number of mother cells is about 1000, while the usual 
number of pollen grains that mature in a sporangium is only 
500-1000. At the time they are shed the pollen grains have a two- 
layered wall, the outer coat of which easily separates from the inner 
(plate figs. 25, 26, 27), but which does not appear to do so naturally 
to form wings. At the time they are shed they contain a great 
many very large starch grains, a few of which are shown in plate 
fig. 27. They are usually so crowded with it that the microtome 
knife scatters the contents all about in cutting sections of shed 
pollen 
Pollination 
The ovulate cones are first recognizable in late April. At this 
time there is no trace of ovules on the scales and the pollen lodges 
somewhere near the free edge of the so-called ligule (text fig. 10). 
