228 
Journal of A gricultural Research voi. xxvn, no. 4 
DESICCATION 
The majority of young cultures dried on cover glasses were dead inside 
of 24 hours, and all within 3 days. 
The bacteria were easily isolated from dry leaves up to 10 days after 
collection. After 10 days' drying, success in isolation was infrequent. In 
only 1 case out of 14 was the organism recovered from leaves dried as 
long as 2 months. 
Exposure to sunlight for 15 minutes killed 100 per cent of the or¬ 
ganisms. 
RELATION TO HOST TISSUES 
Microscopic examination of the lesions shows that the organism attacks 
the parenchyma. The intercellular spaces and the cavities formed by 
disintegration of cell walls become packed with the bacteria. 
For a study of the early stages, material was secured from plants in¬ 
oculated by spraying with pure cultures of the bacteria suspended in 
water. As soon as infection was evident the tissues were collected and 
fixed. Stained sections of this material show that the bacteria enter by 
way of the stomata (PI. 2, B and E). In the earliest stages the bacteria 
are found only in the stomatal chamber, from which they spread into 
and fill adjoining intercellular spaces. That some chemical change 
occurs at this stage in the cell walls adjoining the masses of bacteria is 
evidenced by their increased affinity for stains. Whether they are later 
dissolved or only crushed by pressure and so changed chemically as to 
be vague and doubtful is not yet determined, but cavities occur and be¬ 
come entirely filled with the parasite. In lesions of considerable size no 
definite cell-wall structure remains in the central space. In some large 
cavities the bacteria are arranged in layers (PI. 2, A). 
COMPARISONS WITH SOME SIMILAR BACTERIAL DISEASES 
The translucency of the lesions and the yellow color of the bacterial 
growth suggested that this leaf spot might be identical with some pre¬ 
viously described bacterial disease on monocotyledons. However, care¬ 
ful comparisons show that Bacterium gummisudans has characters that 
separate it from the several organisms that it, to a certain degree, re¬ 
sembles. Bacterium translucens Jones, Johnson, and Reddy 6 and Bac¬ 
terium translucens var. undulosum Smith, Jones and Reddy,®' 8 causing 
disease on barley and wheat, respectively, are unlike Bacterium gum¬ 
misudans in cultural characters in several media and in the consistency 
of the bacterial growth. Moreover, Bacterium gummisudans does not in¬ 
fect grains. Four varieties of barley were used in the inoculation ex¬ 
periments. Several tests were made, but no infections resulted. Control 
inoculations on gladioli gave typical infections. Aplanobacter rathayi 
EFS. 7 ’ 8 and Aplanobacter agropyri O’Gara 9 (on western wheat grass) 
are nonmotile organisms, while Bacterium gummisudans is a motile or¬ 
ganism. 
6 Jones, I,. R., Johnson, A. G., and Reddy, C. S. bacterial-blight of barley. In Jour. Agr. Re¬ 
search, v. 11, p. 625-644, 2 fig., pi. B (col.), 47-49. 1917. Literature cited, p. 643. 
• Smith, Erwin F., Jones, L. R., and Reddy, C. S. the black chaff of wheat. In Science, n. s., 
v. 50, p. 48. 1919. 
7 Smith, Erwin F. bacteria in relation to plant diseases, v. 3, p. 155-160. Washington, D. C. 
1914. (Carnegie Inst. Wash. Pub. 27.) 
8 - An introduction to bacterial diseases of plants, fig. 12-22. Philadelphia and London. 
1920. 
8 O’Gara, P. J. a bacterial disease of western wheat grass, agropyron smithh. In Phyto¬ 
pathology, v. 6 , p. 341-3S0. pi- 9 - 13 * 1916. 
