164 
Vol. XXIX, No. 4 
Journal of Agricultural Research 
are present on both surfaces of the 
husks and on the epidermis of the corm. 
In leaf lesions the bacteria are most 
often found in small compact masses, 
sometimes only in a small area near 
the center. While numerous, they are 
not so abundant as in some other bac¬ 
terial diseases. From fresh sections 
they ooze out only moderately and do 
not readily distribute in the mounting 
fluid. Capsules have been demon¬ 
strated on bacteria obtained directly 
from the plant and it is likely that this 
viscid covering tends to reduce the 
distribution of the bacteria both in the 
tissues and in the mounting fluids. 
Fig. 1.—Camera lucida drawing of a section made 
through the base of a corm lesion. The healthy 
cells, full of starch, are separated from the diseased 
cells by cork and stone cells 
Microscopic examination of the le¬ 
sions removed from corms shows a 
wide layer of cork cells at the back, 
10 to 12 cells deep, and just inside the 
cork are found several rows of stone 
cells (fig. 1). These protective layers 
of cork cells and stone cells are de¬ 
veloped soon after infection occurs and 
evidently account for the small size 
and slight depth attained by the lesion. 
The ridge surrounding the lesion is 
formed by the out cropping of these 
cells. In the central part of the in¬ 
fected area are collapsed cells and a 
gummy substance, amber to brown in 
color, according to age. As in the 
leaf lesions, the bacteria are found in 
rather restricted areas near the center 
of the pit and not far below the epi¬ 
dermis. 
These succulent corm lesions are 
subject to invasion by other organisms, 
both bacteria and fungi. Leaf and 
husk lesions are much less liable to 
secondary infections. 
Growing corms which were inocu¬ 
lated with Bacterium marginatum had 
typical lesions 1 to 4 mm. in diameter 
4 weeks later. The epidermis was in¬ 
tact and not visibly sunken though 
some of the cells just below the surface 
were brown and collapsed. Most of 
the interior cells were fairly normal 
in appearance and w 7 ere still full of 
starch. The cork tissue at the back 
of the lesion was w T ell developed but 
the stone tissue was barely discernible. 
ISOLATION 
The organism which causes this dis¬ 
ease has been isolated repeatedly by 
means of agar poured plates from 
leaves, flower stalks, and corms, both 
from fresh and dry material, and also 
from the gummy exudate. 
From either new or old infections on 
leaves and husks there is no difficulty 
in isolating Bacterium marginatum , but 
from any except recent corm lesions 
the plates are likely to have a discour¬ 
aging number of various sorts of or¬ 
ganisms. A corm spot that is still pale 
in color and with unbroken epidermis 
should be selected. From such spots 
with a sterilization of 2 to 3 minutes, 
absolutely pure cultures of Bacterium 
marginatum have been secured. The 
causal bacteria are present, in some at 
least, of the old lesions on corms, as has 
been proved by isolation experiments, 
but they do not often succeed in produc¬ 
ing characteristic colonies in the isola¬ 
tion plates because of the great number 
and vigor of the secondary organisms 
that have invaded the lesions. Various 
sorts of bacterial colonies appeared in 
isolation plates made from old corm 
lesions, but only rarely w ere there any 
yellow colonies. If the yellow colonies 
produced looked at all like Bacterium 
gummisudans, which causes a leaf blight 
of gladioli (3), they w r ere investigated 
more closely. None was ever found 
with the characters of Bacterium gum¬ 
misudans, and inoculation experiments 
(see p. 167) further prove that Bacterium 
gummisudans is not connected with 
these corm lesions. 
In a field where both Bacterium 
marginatum and Bacterium gummisu¬ 
dans occur, both diseases have quite 
often been found on the foliage of the 
