B Ke) BOTANICAL GAZETTE [JANUARY 
with a deep furrow across the top, due to the fact that the exine 
does not cover the entire surface, but is lacking at the apex of the 
spore, so that when the spore contracts in the drying out which 
precedes shedding, the elastic exine springs together until the oppo- 
site sides touch, thus making it look as if the exine covered the entire 
spore. When placed in water or in a nutrient solution, the spore 
immediately begins to swell, and in a few minutes becomes quite 
spherical. In a to per cent solution of cane sugar, or in the juice 
of either fresh or preserved pears, germination takes place at once. 
Within 24 hours the intine begins to protrude, and in 3 or 4 days 
some of the tubes are two or three times as long as the pollen grain. 
In cultures there is a considerable elongation of the pollen tube and 
some increase in the amount of starch, but I have never succeeded 
in finding a division of the generative cell. The beginning of 
germination, as it appears in a 10 per cent sugar solution, is shown in 
figs. 8, 9, and ro. 
The pollen tube is quite characteristic, and easily distinguishes 
Ceratozamia.from any cycads yet described. As in other cycads, 
the brown roof of the pollen chamber, with the nucellar beak in its 
center, is present, but the brown lines due to the haustoria of pollen 
tubes are scarcely visible, and even in abundantly pollinated 
strobili the brown spot itself is seldom more than 1 mm. in diameter. 
That there are haustoria, 2-3 mm. long and lying just beneath the 
surface of the nucellus, is evident from a glance at a section, but they 
do not cause conspicuous brown lines upon the surface. 
The most striking feature of the pollen tube is a series of second- 
ary haustoria developed from various parts of the enlarged basal 
end of the pollen tube (fig. 11). As soon as the pollen grain is shed, 
the primary haustorium, as the familiar haustorium of cycads 
might be called, begins to develop, and with little or no branching 
reaches a length of 1-2 mm., its course lying just beneath the sur- 
face of the nucellus. The secondary haustoria are developed much 
later. They have about the same diameter as the primary haus- 
toria, but are more sinuous in outline and usually branch. Their 
general direction is toward the archegonia, and their development 
is so rapid that long before the division which is to form the ventral 
a 
