30 BULLETIN 395, U. S. DEPARTMENT OF AGRICULTURE. 
ments. However, the pathogenicity of other leaf strains was clearly proved in sub- 
sequent experiments. 
On the leaves inoculated only on their upper surfaces no primary infection de- 
veloped, with the exception of four lesions which appeared, closely grouped, at the 
margin of one leaf. This infection clearly occurred on the lower surface, and was 
probably due to chance inoculation, such as might occur should a drop of the sporif- 
erous suspension accumulate at the margin of the leaf and extend over the edge. 
While great care was taken to avoid mischance of this kind, as is attested by the 
fact that only one other such case was observed throughout the season's work, it is 
obviously impracticable to secure uniformly perfect results from extensive series of 
experiments of this type. In September and October occasional secondary infections 
appeared upon the under surfaces of the leaves of these plants. 
Abundant disease developed on the lower surfaces of the leaves which were inocu- 
lated on these surfaces only. Its development paralleled that on the corresponding 
sprayed plants. 
On the four controls no disease developed, with the exception of occasional late 
secondary infections upon the leaves and young wood of the plant which was placed 
about a foot from one of the most severely infected trees of the series. This infection 
did not become evident until October. The fact that the next control plant, about 
a foot farther from the source of infection, developed no disease is of significance in 
relation to methods of spore dissemination, which will be discussed later. 
Experiment 3, June 22 3 1914. 
Source of inoculation. — Cladosporium carpophilum: Twig strain, from a 14-day-old 
culture. 
Methods. — The inoculations were made upon potted Chili trees of the type pre- 
viously described. Only the upper surfaces of the leaves of one tree were inoculated 
by spray, while only the lower surfaces of the leaves of another were similarly treated. 
A third plant was sprayed with sterile distilled water as a control. The subsequent 
treatment of the experimental plants paralleled that of the preceding series. 
Results. — The leaves inoculated upon their upper surfaces developed no infection, 
with the exception of several closely aggregated lesions on the lower surface near the 
margin of one leaf. This infection clearly occurred upon the lower surface and was 
doubtless of similar origin to the parallel case discussed under experiment 2. On 
the leaves inoculated upon their lower surfaces, abundant infection was noted on 
July 17, 25 days after inoculation. The lesions had probably been visible several 
days. On July 24 the lesions had increased materially in number and size, scores 
of infections frequently occurring on individual leaves. By the end of September 
most of the worst affected leaves had fallen. No infection had become visible upon 
the control plant when it was last examined on October 18. 
Experiment 4, July 2, 1914. 
Sources of inoculation. — Cladosporium carpophilum: (a) Twig strain, (b) leaf strain; 
both from 10-day-old cultures. 
Methods. — These inoculations were made upon potted Chili trees of the type used 
in the preceding experiment. One tree was inoculated only upon the upper leaf 
surfaces with the twig strain of the fungus, another was similarly treated only upon 
the lower leaf surfaces, and a third on both surfaces of the leaver . Very young and 
mature leaves were separately marked and noted. Another plant was inoculated 
with spray from the leaf strain of the organism, while two others were sprayed with 
sterile distilled water as controls. The experimental plants were subsequently 
accorded treatment similar to that of the preceding series. 
Results. — "When the last notes were made on October 18, no infection had become 
visible upon the leaves inoculated only upon their upper surfaces. Those inoculated 
upon the lower surfaces or upon both surfaces developed abundant infection of the 
