144 
Journal of Agricultural Research 
Vol. XXI, No. 3 
different colored areas being rectangular rather than round. This mot¬ 
tled appearance and the position of the spots between midrib and margin 
clearly distinguish this disease from the common dying back of the older 
leaves, where the margin is first to succumb and the dead areas are a 
uniform red-brown. 
Infection usually takes place while the leaves are still rolled in the 
bud. In case stomatal infections occur just before the leaf pushes up, 
the whole bud often looks very pale, almost white, and close examination 
shows it to be covered with minute white spots. Sometimes these do 
not spread further, and the leaf as it matures becomes green with a pep¬ 
pering of tiny spots, but more often as it unfolds it remains pale and 
stunted. At other times the infection has progressed so far that by the 
time the folded leaf emerges it is wholly or in part blackened, sometimes 
in spiral bands (PI. 35, B). In such cases the younger leaves usually 
become infected by direct contact, or the disease runs down the petiole 
and kills the young stalk and bud. A non-fatal bud infection is shown 
in Plate 35, A, where blackening has occurred but not to such an extent 
as to kill the shoot. 
From leaf-blade infections (PL 33; 36, C) the bacteria invade the 
petiole, not by way of the vascular system but through the parenchyma, 
chiefly through the channels of very loose tissue which occupy a large 
portion of the interior of the petiole. In the tightly rolled buds infection 
appears to pass directly from one leaf to another, so that when a young 
shoot is cut across near ground level, several petioles may be found to be 
diseased. The buds do not become soft-rotted but usually stand up 
black and dry or are bent over or broken off. Eventually the center rots 
out, leaving the hollow stalk standing with one or two mature leaves. 
This when cut across near the base is found to contain a watery rot* 
The rootstocks have never been found to be diseased. In the autumn 
of 1919 when the plants in the grounds were lifted for winter storage all 
the plants of one large bed that had shown heavy infection were thor¬ 
oughly examined. All stalks were cut across a few inches from the 
ground, and in a large proportion of the clumps one to four stalks were 
found with interiors rotted out, to or below ground level, the lower part 
of the cavity being filled with fluid. In no case, however, was the rot 
found extending into the rootstock, the tissues of which do not seem to 
favor the growth of the organism. In many stalks showing a character¬ 
istic top—that is, standing erect with older leaves intact but with a 
hollow blackened center, the decay had not reached the lower part, so 
that a cross section 1 foot from ground level showed only sound tissue. 
Often when the shoot has escaped early and complete destruction, the 
flower clusters are ruined either by the infection of the young flower buds 
or by the decay of the stem. In the former instance the stem and 
pedicels develop but the buds blacken and die while still rudimentary 
(PI* 34 > A, C). In the latter case the stalk bends or breaks in the infected 
