8 BULLETIN 395, U. S. DEPARTMENT OF AGRICULTURE. 
meter in diameter. The color of the affected surface may be a pale 
green, almost indistinguishable from the normal, or it may vary to 
light brown or pinkish purple. The true color, however, is frequently 
modified or masked by an abundant development of olivaceous conidi- 
ophores and conidia. These are usually present in greater or less num- 
bers when the infection first becomes macroscopic. When conditions 
especially favor their development, they may be borne in sufficient 
quantities to give a dark-green color to the lesions and to constitute 
their most conspicuous character. On the leaf lamina the fully 
developed lesions are somewhat angular to irregularly circular. Their 
average diameter is 1 to 2 millimeters, though in rare cases they may 
attain a diameter of 5 or 6 millimeters (PI. I, fig. 7, and PI. IV, fig. 1). 
On the petioles and midribs they are usually much longer and nar- 
rower, generally measuring 3 to 8 by one-half to 1 millimeters. The 
color of the blotch proper may vary from light brown to dark brown, 
or it may be pinkish purple. The demarcation remains indistinct 
and the elevation is usually unchanged, except in extreme cases, 
when the lamina above the larger lesions may become convex below 
and correspondingly concave above. When the lesions first appear, 
no evidence of disease is visible upon the upper surfaces of the leaves. 
In later stages, however, the tissues above and immediately around 
the affected areas may lose some or all of their green color and become 
pale yellow or purplish. In extreme cases such areas may die, but 
ordinarily no definite leaf spotting is induced. The lesions may occur 
miscellaneously scattered over the lower surfaces of the leaves and 
upon the petioles, or they may develop in such abundance as to 
become confluent and form irregular patches (PL III, fig. 2). The 
infection appears to occur with equal readiness on the petioles, 
midribs, veins, and leaf lamina. It was demonstrated in the inocula- 
tion experiments that badly diseased leaves may turn yellow and fall, 
but the writer has rarely observed this condition as the result of 
natural scab infection. 
PATHOLOGICAL HISTOLOGY. 
Histological technique. — In Hiring material for histological work, 
several standard fixing agents were employed, viz, Flennning's weak, 
medium, and strong fluids, picro-formal, and chrom-acetic. The 
stains used were Flemming's triple, Haidenhain's iron alum- 
haematoxylin, Durand's haematoxylin-eosin, gentian violet, and 
safranin. 
For most purposes, Flemming's medium fluid and the triple stain 
gave the best results. This combination not only stained the fungus 
well, but gave an excellent differentiation of fungus and host in twig 
and fruit lesions. The cuticle of the host remained practically 
colorless, while the dead epidermal and subepidermal cells took little 
stain, being yellowish green in the finished preparation. In sharp 
