July 21,1923 
Crowngall Organism and Its Host Tissue 
121 
December 15, 1920. On February 21, 1921, those that had received the 
inoculation at the top showed well-developed tumors. No other trace 
of gall could be found on any of the plants. Pure cultures of the organism 
were reisolated from both the soil and the agar. These readily infected 
other tomato plants when introduced into them. It appears, therefore, 
that infection did not occur in the tomatoes from organisms in the soil 
or agar in the absence of wounds. 
Ten inoculations were made with a hypodermic syringe into the hollow 
pith cavities of castor bean petioles and stems. Varying amounts of a 
suspension of the organism ranging from 0.2 cc. to 1 cc. were injected. 
On October 4, 1922, three months later, these inoculations were examined. 
Well-developed galls were found at the punctures, but nowhere else, 
even though in some cases the path of the inoculum as it flowed down 
the pith cavity was traceable by a more or less dark stain. Plates were 
poured from this discolored region. The organism appeared in abun¬ 
dance and its identity was subsequently established by infection of 
tomatoes. Repetition of these experiments gave similar results. This 
work simply confirms the conclusions of earlier writers, namely, that 
under ordinary circumstances wounds are necessary for crown-gall 
infection. 
Since wounds appear to be necessary for infection, it seemed desirable 
to inquire into the relation of size or character of the wound to gall 
development. This seemed especially pertinent in its possible relation 
to the associated water-soaking of the tissues. Glass and quartz rods 
were drawn out to make needles of different sizes ranging from 30 n to 
385 fi in diameter. A series of punctures were made with these needles 
through masses of the gall organism into tomato stems. After 25 days 
galls had developed, the diameters of which bore a more or less direct 
relationship to the diameters of the punctures (PI. 1, A to E). Since the 
larger punctures released more liquid into the intercellular spaces, it 
appears that the size of the gall may be roughly proportional to the 
volume of tissue in which the intercellular spaces are flooded with liquid. 
A repetition of this experiment gave similar results. 
The experiments outlined above and others of like import raised the 
question whether the inoculum must be inserted into the injured tissue 
or if the organisms might enter from the surface through the wounds. 
Consequently, 50 inoculations were made by puncturing through masses 
of bacteria which had been applied to tomato stems with a camel’s-hair 
brush. In a parallel series, 50 punctures were first made with a sterile 
needle, and the organism was applied with a brush a few minutes later. 
Another series in which 50 sterile punctures were made in a similar 
manner, but without the application of any bacteria, served as controls. 
In this experiment not more than 10 punctures were made on a single 
plant. After five weeks galls had developed at every puncture where 
the organisms were applied while none had appeared on the controls. 
The developing galls showed no difference in relation to the two methods 
of inoculation. 
The forces which govern this entry of the organism into the tissue are 
not definitely understood. The bacteria might conceivably be influenced 
by any or all of such factors as the collapse of drying tissue, negative 
pressure, sap rise, and motility with or without a chemotactic stimulus. 
To test the possible relation of chemotaxis, very small capillary glass 
tubes were filled with expressed tomato stem sap. These were placed 
