478 
AMERICAN JOURNAL OF BOTANY 
[Vol. 8. 
the hills of Vermont, six miles from a town. When this possible source of 
error was suspected, ragweed pollen was purposely obtained from Michigan, 
from two parts of New York state, and from Connecticut. It does not 
seem probable that the bacteria and molds carried by pollen can be so 
constant as to cause similar enzyme action in each instance. 
h. The reactions are too rapid to be due to bacteria. With the in- 
hibiting action of antiseptics the time required for bacteria to develop in 
sufficient numbers to produce similar changes would be much longer. All 
the enzyme reactions recorded have occurred within 24 hours, and several 
have been almost instantaneous. 
c. Slices of wood in water over night are not in any degree sterile, yet 
bacteria which have free access do not destroy the middle lamellae, but 
pollen grains do. Pollen grains taken from unopened anthers and put 
into sterile Petri dishes are not likely to have peculiar bacteria, absent from 
the immediate environment. Besides, examination of the pollen contami- 
nation showed only a few omnipresent common forms of bacteria. 
d. Pollen solutions filtered through a Berkefeld filter gave the enzyme 
action of diastase on starch, and blood fibrin digestion. 
e. It is probable that the ground pollen added something to the milk 
which stimulated the growth of bacteria already in the milk, and that it 
was these which caused coagulation rather than the bacteria introduced by 
the pollen. The reason for this belief is that in all the plates poured from 
milk to which pollen had been added Bacillus fluorescens liquefaciens was 
the dominant type. The plates after standing a few days were a bright 
apple-green from the fluorescent growth. 
On other plates poured later from the pollen extracts only, not once did 
this form appear. In the latter it was often not until the third or fourth 
day that colonies of molds occurred. Doubtless there are resistant forms 
of spores on the pollen which endure the heat of the autoclave and develop 
under favorable conditions on the agar plates, but these can hardly account 
for digestions which occur during twenty-four hours. 
The Chemistry of Pollen 
While many kinds of pollen have been examined for certain special 
constituents such as starch, nitrogen, phosphoric acid, etc., only eight kinds 
of pollen, as far as I have been able to ascertain, have been analyzed with 
any degree of completeness. Czapek (1905) discusses topically the occur- 
rence and distribution of the principal constituents of plants; if a substance 
has been reported present in pollen he mentions the fact. These scattered 
references afford a valuable index to the original literature of the earlier 
analyses. 
According to Heyl (1919 a, p. 672) the walls of the pollen grain constitute 
65 percent of the structure. Biourge (1892, p. 75) distinguishes four sub- 
stances in the wall or envelope of pollen grains: cutin, cellulose, pectic 
