122 
Journal of Agricultural Research 
Vol. XXV, No. 3 
in a hanging-drop preparation of a suspension of actively motile crown- 
gall organisms in sterile distilled water in a manner similar to that 
described by Pfeffer (4). In a few minutes the bacteria were observed 
collecting in the tube. At the end of a half hour the motile organisms 
were thickly concentrated in the tube, while no active ones were observed 
outside. This experiment was repeated eight times with tubes of different 
diameters. In addition, it was tried with the positions of the bacterial 
suspension and the tomato sap reversed. In every case the motile bac¬ 
teria collected in the liquid expressed from the tomato stem. It appears, 
therefore, that this tomato sap exerts a positively chemotactic stimulus 
upon the crowngall organism, which alone may be sufficient to account 
for the entrance of the bacteria into a puncture. 
How long one of these individual punctures might serve as an infec¬ 
tion court seemed an important point. One hundred and fifty punctures 
were made into tomato stems, and at intervals of time bacteria were 
applied to the punctures with a camel’s-hair brush. The plants remained 
in an open greenhouse throughout the experiment. The results, recorded 
at the end of three weeks, appear in Table I. 
Tabi,^ I .—A summary of results of experiments on the length of time that a puncture 
may serve as an infection court for B. tumefaciens 
Number 
of punc¬ 
tures. 
Time between puncture and application of organism. 
10 
5 minutes. 
10 
2 hours. 
10 
4 hours. 
20 
1 day... 
20 
2 days. 
to 
2% days. 
10 
■x days. 
10 
4 days. 
IO 
5 days. 
Infection after 3 weeks’ 
incubation. 
Number. 
Per cent. 
IO 
IOO 
IO 
IOO 
IO 
IOO 
19 
95 
12 
60 
9 
90 
2 
20 
2 
20 
O 
O 
Under more favorable conditions, when the plants were placed in a 
moist chamber for 24 hours following the application of the bacteria, 5 
infections were secured from 10 punctures that were six days old and 
2 infections from 10 punctures that were seven days old. Eight and 
nine-day-old punctures produced no galls. A repetition of this experi¬ 
ment gave similar results. 
Although it is evident that conditions are important in determining 
the length of time a puncture may serve as an infection court, these 
experiments show that infection is very readily accomplished within the 
first day, and that it may under favoring conditions take place when the 
application of the organisms is delayed until as long as seven days after 
wounding the tissue. These results appear to be incompatible with the 
conception (< 5 , P>ijo) that the bacteria first establish themselves in wounded 
cells. It seems unlikely that in the case of such delayed infection any 
cell walls which might have been punctured at the time of the original 
wound would still offer open channels for the penetration of the bacteria 
into the living cells. Before seven days had elapsed the injured cells 
probably would either have died or healed the ruptures, provided they 
were capable of the latter. 
